MPlayer(1)                     The Movie Player                     MPlayer(1)



NAME
      mplayer  - movie player
      mencoder - movie encoder

SYNOPSIS
      mplayer [options] [file|URL|playlist|-]
      mplayer [options] file1 [specific options] [file2] [specific options]
      mplayer [options] {group of files and options} [group-specific options]
      mplayer [br]://[title][/device] [options]
      mplayer [dvd|dvdnav]://[title|[start_title]-end_title][/device]
      [options]
      mplayer vcd://track[/device]
      mplayer tv://[channel][/input_id] [options]
      mplayer radio://[channel|frequency][/capture] [options]
      mplayer pvr:// [options]
      mplayer dvb://[card_number@]channel [options]
      mplayer mf://[filemask|@listfile] [-mf options] [options]
      mplayer [cdda|cddb]://track[-endtrack][:speed][/device] [options]
      mplayer cue://file[:track] [options]
      mplayer
      [file|mms[t]|http|http_proxy|rt[s]p|ftp|udp|unsv|icyx|noicyx|smb]://
      [user:pass@]URL[:port] [options]
      mplayer sdp://file [options]
      mplayer mpst://host[:port]/URL [options]
      mplayer tivo://host/[list|llist|fsid] [options]
      gmplayer [options] [file|URL|playlist] [-skin skin]
      mencoder [options] file [file|URL|-] [-o file | file://file |
      smb://[user:pass@]host/filepath]
      mencoder [options] file1 [specific options] [file2] [specific options]

DESCRIPTION
      mplayer  is  a movie player for Linux (runs on many other platforms and
      CPU architectures, see the documentation).   It  plays  most  MPEG/VOB,
      AVI, ASF/WMA/WMV, RM, QT/MOV/MP4, Ogg/OGM, MKV, VIVO, FLI, NuppelVideo,
      yuv4mpeg, FILM and RoQ files,  supported  by  many  native  and  binary
      codecs.   You  can watch VCD, SVCD, DVD, Blu-ray, 3ivx, DivX 3/4/5, WMV
      and even H.264 movies, too.

      MPlayer supports a wide range of video and audio  output  drivers.   It
      works  with  X11,  Xv, DGA, OpenGL, SVGAlib, fbdev, AAlib, libcaca, Di‐
      rectFB, Quartz, Mac OS X CoreVideo, but you can also use GGI, SDL  (and
      all  their  drivers), VESA (on every VESA-compatible card, even without
      X11), some low-level card-specific drivers (for Matrox, 3dfx  and  ATI)
      and  some  hardware MPEG decoder boards, such as the Siemens DVB, Haup‐
      pauge PVR (IVTV), DXR2 and DXR3/Hollywood+.  Most of them support soft‐
      ware or hardware scaling, so you can enjoy movies in fullscreen mode.

      MPlayer  has an onscreen display (OSD) for status information, nice big
      antialiased shaded subtitles and visual feedback for keyboard controls.
      European/ISO8859-1,2 (Hungarian, English, Czech, etc), Cyrillic and Ko‐
      rean fonts are supported along with 12 subtitle formats (MicroDVD, Sub‐
      Rip,  OGM, SubViewer, Sami, VPlayer, RT, SSA, AQTitle, JACOsub, PJS and
      our own: MPsub) and DVD subtitles (SPU streams, VOBsub and Closed  Cap‐
      tions).

      mencoder  (MPlayer's Movie Encoder) is a simple movie encoder, designed
      to encode MPlayer-playable movies (see above) to other MPlayer-playable
      formats  (see  below).   It  encodes  to MPEG-4 (DivX/Xvid), one of the
      libavcodec codecs and PCM/MP3/VBRMP3 audio in 1, 2 or 3  passes.   Fur‐
      thermore  it  has  stream  copying  abilities, a powerful filter system
      (crop, expand, flip, postprocess, rotate, scale, noise, RGB/YUV conver‐
      sion) and more.

      gmplayer  is MPlayer with a graphical user interface.  Besides some own
      options (stored in gui.conf), it has the same options as MPlayer,  how‐
      ever  some  MPlayer options will be stored in gui.conf so that they can
      be chosen independently from MPlayer. (See GUI CONFIGURATION  FILE  be‐
      low.)

      Usage  examples  to  get you started quickly can be found at the end of
      this man page.

      Also see the HTML documentation!

INTERACTIVE CONTROL
      MPlayer has a fully configurable, command-driven  control  layer  which
      allows you to control MPlayer using keyboard, mouse, joystick or remote
      control (with LIRC).  See the -input option for ways to customize it.

      keyboard control
             LEFT and RIGHT
                  Seek backward/forward 10 seconds.
             UP and DOWN
                  Seek forward/backward 1 minute.
             PGUP and PGDWN
                  Seek forward/backward 10 minutes.
             [ and ]
                  Decrease/increase current playback speed by 10%.
             { and }
                  Halve/double current playback speed.
             BACKSPACE
                  Reset playback speed to normal.
             < and >
                  Go backward/forward in the playlist.
             ENTER
                  Go forward in the playlist, even over the end.
             HOME and END
                  next/previous playtree entry in the parent list
             INS and DEL (ASX playlist only)
                  next/previous alternative source.
             p / SPACE
                  Pause (pressing again unpauses).
             .
                  Step forward.  Pressing once will pause movie,  every  con‐
                  secutive  press  will play one frame and then go into pause
                  mode again (any other key unpauses).
             q / ESC
                  Stop playing and quit.
             U
                  Stop playing (and quit if -idle is not used).
             + and -
                  Adjust audio delay by +/- 0.1 seconds.
             / and *
                  Decrease/increase volume.
             9 and 0
                  Decrease/increase volume.
             ( and )
                  Adjust audio balance in favor of left/right channel.
             m
                  Mute sound.
             _ (MPEG-TS, AVI and libavformat only)
                  Cycle through the available video tracks.
             # (DVD, Blu-ray, MPEG, Matroska, AVI and libavformat only)
                  Cycle through the available audio tracks.
             TAB (MPEG-TS and libavformat only)
                  Cycle through the available programs.
             f
                  Toggle fullscreen (also see -fs).
             T
                  Toggle stay-on-top (also see -ontop).
             w and e
                  Decrease/increase pan-and-scan range.
             o
                  Toggle OSD states: none / seek / seek  +  timer  /  seek  +
                  timer + total time.
             d
                  Toggle  frame  dropping  states: none / skip display / skip
                  decoding (see -framedrop and -hardframedrop).
             v
                  Toggle subtitle visibility.
             j and J
                  Cycle through the available subtitles.
             y and g
                  Step forward/backward in the subtitle list.
             F
                  Toggle displaying "forced subtitles".
             a
                  Toggle subtitle alignment: top / middle / bottom.
             x and z
                  Adjust subtitle delay by +/- 0.1 seconds.
             c (-capture only)
                  Start/stop capturing the primary stream.
             r and t
                  Move subtitles up/down.
             i (-edlout mode only)
                  Set start or end of an EDL skip and write  it  out  to  the
                  given file.
             s (-vf screenshot only)
                  Take a screenshot.
             S (-vf screenshot only)
                  Start/stop taking screenshots.
             I
                  Show filename on the OSD.
             P
                  Show  progression  bar,  elapsed time and total duration on
                  the OSD.
             ! and @
                  Seek to the beginning of the previous/next chapter.
             D (-vo xvmc, -vo vdpau, -vf yadif, -vf kerndeint only)
                  Activate/deactivate deinterlacer.
             A    Cycle through the available DVD angles.

             (The following keys are valid only when using a hardware  accel‐
             erated  video  output  (xv, (x)vidix, (x)mga, etc), the software
             equalizer (-vf eq or -vf eq2) or hue filter (-vf hue).)

             1 and 2
                  Adjust contrast.
             3 and 4
                  Adjust brightness.
             5 and 6
                  Adjust hue.
             7 and 8
                  Adjust saturation.

             (The following keys are valid only  when  using  the  quartz  or
             corevideo video output driver.)

             command + 0
                  Resize movie window to half its original size.
             command + 1
                  Resize movie window to its original size.
             command + 2
                  Resize movie window to double its original size.
             command + f
                  Toggle fullscreen (also see -fs).
             command + [ and command + ]
                  Set movie window alpha.

             (The following keys are valid only when using the sdl video out‐
             put driver.)

             c
                  Cycle through available fullscreen modes.
             n
                  Restore original mode.

             (The following keys are valid if you have a keyboard with multi‐
             media keys.)

             PAUSE
                  Pause.
             STOP
                  Stop playing and quit.
             PREVIOUS and NEXT
                  Seek backward/forward 1 minute.

             (The  following  keys  are only valid if you compiled with TV or
             DVB input support and will take precedence over the keys defined
             above.)

             h and k
                  Select previous/next channel.
             n
                  Change norm.
             u
                  Change channel list.

             (The  following  keys are only valid if you compiled with dvdnav
             support: They are used to navigate the menus.)

             keypad 8
                  Select button up.
             keypad 2
                  Select button down.
             keypad 4
                  Select button left.
             keypad 6
                  Select button right.
             keypad 5
                  Return to main menu.
             keypad 7
                  Return to nearest menu (the order of preference  is:  chap‐
                  ter->title->root).
             keypad ENTER
                  Confirm choice.

             (The  following  keys  are used for controlling TV teletext. The
             data may come from either an analog TV source or an MPEG  trans‐
             port stream.)

             X
                  Switch teletext on/off.
             Q and W
                  Go to next/prev teletext page.

      mouse control
             button 3 and button 4
                  Seek backward/forward 1 minute.
             button 5 and button 6
                  Decrease/increase volume.

      joystick control
             left and right
                  Seek backward/forward 10 seconds.
             up and down
                  Seek forward/backward 1 minute.
             button 1
                  Pause.
             button 2
                  Toggle  OSD  states:  none  /  seek / seek + timer / seek +
                  timer + total time.
             button 3 and button 4
                  Decrease/increase volume.

USAGE
      Every 'flag' option has a 'noflag' counterpart, e.g.  the  opposite  of
      the -fs option is -nofs.

      If  an option is marked as (XXX only), it will only work in combination
      with the XXX option or if XXX is compiled in.

      NOTE: The suboption parser (used for example for  -ao  pcm  suboptions)
      supports a special kind of string-escaping intended for use with exter‐
      nal GUIs.
      It has the following format:
      %n%string_of_length_n
      EXAMPLES:
      mplayer -ao pcm:file=%10%C:test.wav test.avi
      Or in a script:
      mplayer -ao pcm:file=%`expr length "$NAME"`%"$NAME" test.avi

CONFIGURATION FILES
      You can put all of the options in configuration  files  which  will  be
      read every time MPlayer/MEncoder is run.  The system-wide configuration
      file 'mplayer.conf' is in  your  configuration  directory  (e.g.  /etc/
      mplayer or /usr/local/etc/mplayer), the user specific one is '~/.mplay‐
      er/config'.  The configuration file for MEncoder is 'mencoder.conf'  in
      your  configuration  directory  (e.g.  /etc/mplayer  or /usr/local/etc/
      mplayer), the user specific one  is  '~/.mplayer/mencoder.conf'.   User
      specific  options  override  system-wide  options (in case of gmplayer,
      gui.conf options override user specific options) and options  given  on
      the  command  line override all.  The syntax of the configuration files
      is 'option=<value>', everything after a '#' is  considered  a  comment.
      Options  that  work  without  values  can be enabled by setting them to
      'yes' or '1' or 'true' and disabled by setting them to 'no' or  '0'  or
      'false'.  Even suboptions can be specified in this way.

      You  can  also write file-specific configuration files.  If you wish to
      have a configuration file for a file called 'movie.avi', create a  file
      named  'movie.avi.conf' with the file-specific options in it and put it
      in ~/.mplayer/.  You can also put the configuration file  in  the  same
      directory  as  the  file  to  be  played,  as  long  as  you  give  the
      -use-filedir-conf option (either on the command line or in your  global
      config  file).   If  a file-specific configuration file is found in the
      same directory, no file-specific configuration is loaded from ~/.mplay‐
      er.   In  addition, the -use-filedir-conf option enables directory-spe‐
      cific configuration files.  For this, MPlayer first  tries  to  load  a
      mplayer.conf  from the same directory as the file played and then tries
      to load any file-specific configuration.

      EXAMPLE MPLAYER CONFIGURATION FILE:

      # Use Matrox driver by default.
      vo=xmga
      # I love practicing handstands while watching videos.
      flip=yes
      # Decode/encode multiple files from PNG,
      # start with mf://filemask
      mf=type=png:fps=25
      # Eerie negative images are cool.
      vf=eq2=1.0:-0.8
      # OSD progress bar vertical alignment
      progbar-align=50

      EXAMPLE MENCODER CONFIGURATION FILE:

      # Make MEncoder output to a default filename.
      o=encoded.avi
      # The next 4 lines allow mencoder tv:// to start capturing immediately.
      oac=pcm=yes
      ovc=lavc=yes
      lavcopts=vcodec=mjpeg
      tv=driver=v4l2:input=1:width=768:height=576:device=/dev/video0:audiorate=48000
      # more complex default encoding option set
      lavcopts=vcodec=mpeg4:autoaspect=1
      lameopts=aq=2:vbr=4
      ovc=lavc=1
      oac=lavc=1
      passlogfile=pass1stats.log
      noautoexpand=1
      subfont-autoscale=3
      subfont-osd-scale=6
      subfont-text-scale=4
      subalign=2
      subpos=96
      spuaa=20

      GUI CONFIGURATION FILE

      GUI's own options are (MPlayer option  names  in  parentheses):  ao_al‐
      sa_device  (alsa:device=)  (ALSA only), ao_alsa_mixer (mixer) (ALSA on‐
      ly), ao_alsa_mixer_channel (mixer-channel) (ALSA  only),  ao_esd_device
      (esd:)  (ESD  only),  ao_extra_stereo  (af extrastereo) (default: 1.0),
      ao_extra_stereo_coefficient  (af  extrastereo=),  ao_oss_device  (oss:)
      (OSS only), ao_oss_mixer (mixer) (OSS only), ao_oss_mixer_channel (mix‐
      er-channel) (OSS only), ao_sdl_subdriver (sdl:) (SDL only), ao_surround
      (unused),  ao_volnorm  (af  volnorm),  autosync  (enable/disable),  au‐
      tosync_size (autosync), cache (enable/disable), cache_size (cache), en‐
      able_audio_equ  (af equalizer), equ_band_00 ... equ_band_59, (af equal‐
      izer=), equ_channel_1 ... equ_channel_6 (af channels=), gui_main_pos_x,
      gui_main_pos_y,   gui_save_pos   (yes/no),   gui_tv_digital   (yes/no),
      gui_video_out_pos_x, gui_video_out_pos_y, playbar (enable/disable), re‐
      play_gain     (enable/disable),    replay_gain_adjustment    (-30..10),
      show_videowin (yes/no), vf_lavc (vf lavc) (DXR3 only), vf_pp  (vf  pp),
      vo_dxr3_device (unused) (DXR3 only).

      MPlayer  options  stored  in gui.conf (GUI option names, MPlayer option
      names in parentheses) are: a_afm (afm), ao_driver (ao), ass_bottom_mar‐
      gin  (ass-bottom-margin)  (ASS  only),  ass_enabled  (ass)  (ASS only),
      ass_top_margin (ass-top-margin) (ASS only),  ass_use_margins  (ass-use-
      margins)  (ASS  only), cdrom_device (cdrom-device), dvd_device (dvd-de‐
      vice), font_autoscale (subfont-autoscale)  (FreeType  only),  font_blur
      (subfont-blur)   (FreeType   only),  font_encoding  (subfont-encoding),
      font_factor (ffactor), font_name (font),  font_osd_scale  (subfont-osd-
      scale) (FreeType only), font_outline (subfont-outline) (FreeType only),
      font_text_scale (subfont-text-scale) (FreeType only), gui_skin  (skin),
      idle  (idle), load_fullscreen (fs), osd_level (osdlevel), playlist_sup‐
      port     (allow-dangerous-playlist-parsing),     softvol     (softvol),
      stopxscreensaver  (stop-xscreensaver),  sub_auto_load (autosub), sub_cp
      (subcp)  (iconv  only),  sub_overlap  (overlapsub),  sub_pos  (subpos),
      sub_unicode  (unicode),  sub_utf8  (utf8),  v_flip  (flip), v_framedrop
      (framedrop), v_idx (idx), v_ni (ni),  v_vfm  (vfm),  vf_autoq  (autoq),
      vo_direct_render  (panscan),  vo_doublebuffering  (dr), vo_driver (vo),
      vo_panscan (double).

PROFILES
      To ease working with different configurations profiles can  be  defined
      in  the  configuration  files.   A profile starts with its name between
      square brackets, e.g. '[my-profile]'.  All following  options  will  be
      part of the profile.  A description (shown by -profile help) can be de‐
      fined with the profile-desc option.  To end the profile, start  another
      one or use the profile name 'default' to continue with normal options.

      EXAMPLE MPLAYER PROFILE:


      [protocol.dvd]
      profile-desc="profile for dvd:// streams"
      vf=pp=hb/vb/dr/al/fd
      alang=en

      [protocol.dvdnav]
      profile-desc="profile for dvdnav:// streams"
      profile=protocol.dvd
      mouse-movements=yes
      nocache=yes

      [extension.flv]
      profile-desc="profile for .flv files"
      flip=yes

      [vo.pnm]
      outdir=/tmp

      [ao.alsa]
      device=spdif

      EXAMPLE MENCODER PROFILE:


      [mpeg4]
      profile-desc="MPEG4 encoding"
      ovc=lacv=yes
      lavcopts=vcodec=mpeg4:vbitrate=1200

      [mpeg4-hq]
      profile-desc="HQ MPEG4 encoding"
      profile=mpeg4
      lavcopts=mbd=2:trell=yes:v4mv=yes

GENERAL OPTIONS
      -codecpath <dir>
             Specify a directory for binary codecs.

      -codecs-file <filename> (also see -afm, -ac, -vfm, -vc)
             Override the standard search path and use the specified file in‐
             stead of the builtin codecs.conf.

      -include <configuration file> (also see -gui-include)
             Specify configuration file to be parsed after the default ones.

      -list-options
             Prints all available options.

      -msgcharset <charset>
             Convert console messages to the  specified  character  set  (de‐
             fault: autodetect).  Text will be in the encoding specified with
             the --charset configure option.  Set this to "noconv" to disable
             conversion (for e.g. iconv problems).
             NOTE:  The  option  takes  effect after command line parsing has
             finished.  The MPLAYER_CHARSET environment variable can help you
             get rid of the first lines of garbled output.

      -msgcolor
             Enable  colorful  console  output on terminals that support ANSI
             color.

      -msglevel <all=<level>:<module>=<level>:...>
             Control verbosity directly for each module.   The  'all'  module
             changes  the  verbosity of all the modules not explicitly speci‐
             fied on the command line.  See '-msglevel help' for  a  list  of
             all modules.
             NOTE:  Some  messages  are  printed  before  the command line is
             parsed and are therefore not affected by -msglevel.  To  control
             these  messages  you have to use the MPLAYER_VERBOSE environment
             variable, see its description below for details.
             Available levels:
                -1   complete silence
                 0   fatal messages only
                 1   error messages
                 2   warning messages
                 3   short hints
                 4   informational messages
                 5   status messages (default)
                 6   verbose messages
                 7   debug level 2
                 8   debug level 3
                 9   debug level 4

      -msgmodule
             Prepend module name in front of each console message.

      -noconfig <options>
             Do not parse selected configuration files.
             NOTE: If -include or -use-filedir-conf options are specified  at
             the command line, they will be honoured.

             Available options are:
                all
                     all configuration files
                gui (GUI only)
                     GUI configuration file
                system
                     system configuration file
                user
                     user configuration file

      -quiet
             Make  console  output  less verbose; in particular, prevents the
             status line (i.e. A:   0.7 V:   0.6 A-V:  0.068 ...) from  being
             displayed.  Particularly useful on slow terminals or broken ones
             which do not properly handle carriage return (i.e. \r).

      -priority <prio> (Windows and OS/2 only)
             Set process priority for MPlayer  according  to  the  predefined
             priorities available under Windows and OS/2.  Possible values of
             <prio>:
                idle|belownormal|normal|abovenormal|high|realtime

             WARNING: Using realtime priority can cause system lockup.

      -profile <profile1,profile2,...>
             Use the given profile(s), -profile help displays a list  of  the
             defined profiles.

      -really-quiet (also see -quiet)
             Display  even  less output and status messages than with -quiet.
             Also suppresses the GUI error message boxes.

      -show-profile <profile>
             Show the description and content of a profile.

      -use-filedir-conf
             Look for a file-specific configuration file in the same directo‐
             ry as the file that is being played.
             WARNING: May be dangerous if playing from untrusted media.

      -v
             Increment  verbosity  level,  one level for each -v found on the
             command line.

PLAYER OPTIONS (MPLAYER ONLY)
      -autoq <quality> (use with -vf [s]pp)
             Dynamically changes the level of postprocessing depending on the
             available  spare  CPU  time.  The number you specify will be the
             maximum level used.  Usually you can use some big  number.   You
             have  to  use  -vf [s]pp without parameters in order for this to
             work.

      -autosync <factor>
             Gradually adjusts the A/V sync based  on  audio  delay  measure‐
             ments.   Specifying  -autosync  0, the default, will cause frame
             timing to be based entirely on audio delay measurements.  Speci‐
             fying  -autosync  1 will do the same, but will subtly change the
             A/V correction algorithm.  An uneven video framerate in a  movie
             which  plays  fine  with -nosound can often be helped by setting
             this to an integer value greater than 1.  The higher the  value,
             the  closer the timing will be to -nosound.  Try -autosync 30 to
             smooth out problems with sound drivers which do not implement  a
             perfect  audio delay measurement.  With this value, if large A/V
             sync offsets occur, they will only take about 1 or 2 seconds  to
             settle  out.   This delay in reaction time to sudden A/V offsets
             should be the only side-effect of turning this  option  on,  for
             all sound drivers.

      -benchmark
             Prints  some  statistics  on CPU usage and dropped frames at the
             end of playback.  Use in combination with -nosound and -vo  null
             for benchmarking only the video codec.
             NOTE:  With  this option MPlayer will also ignore frame duration
             when playing only video (you can think of that as infinite fps).

      -colorkey <number>
             Changes the colorkey to an RGB value of your  choice.   0x000000
             is  black  and 0xffffff is white.  Only supported by the cvidix,
             fbdev, svga, vesa, winvidix, xmga, xvidix, xover,  xv  (see  -vo
             xv:ck), xvmc (see -vo xv:ck) and directx video output drivers.

      -nocolorkey
             Disables  colorkeying.  Only supported by the cvidix, fbdev, sv‐
             ga, vesa, winvidix, xmga, xvidix, xover,  xv  (see  -vo  xv:ck),
             xvmc (see -vo xv:ck) and directx video output drivers.

      -correct-pts (EXPERIMENTAL)
             Switches  MPlayer  to  an experimental mode where timestamps for
             video frames are calculated differently and video filters  which
             add  new  frames  or modify timestamps of existing ones are sup‐
             ported.  The more accurate timestamps can be visible for example
             when  playing subtitles timed to scene changes with the -ass op‐
             tion.  Without -correct-pts the subtitle timing  will  typically
             be off by some frames.  This option does not work correctly with
             some demuxers and codecs.

      -crash-debug (DEBUG CODE)
             Automatically attaches gdb upon crash or SIGTRAP.  Support  must
             be compiled in by configuring with --enable-crash-debug.

      -doubleclick-time
             Time in milliseconds to recognize two consecutive button presses
             as a double-click (default: 300).  Set to 0 to let your  window‐
             ing system decide what a double-click is (-vo directx only).
             NOTE:  You  will  get  slightly different behaviour depending on
             whether you bind MOUSE_BTN0_DBL or MOUSE_BTN0-MOUSE_BTN0_DBL.

      -edlout <filename>
             Creates a new file and writes edit decision list  (EDL)  records
             to  it.  During playback, the user hits 'i' to mark the start or
             end of a skip block.  This provides a starting point from  which
             the user can fine-tune EDL entries later.  See http://www.mplay‐
             erhq.hu/DOCS/HTML/en/edl.html for details.

      -edl-backward-delay <number>
             When using EDL during playback and jumping backwards it is  pos‐
             sible  to  end  up in the middle of an EDL record.  In that case
             MPlayer will seek further backwards to the start position of the
             EDL  record and then immediately skip the scene specified in the
             EDL record.  To avoid this kind of behavior, MPlayer jumps to  a
             fixed  time  interval  before the start of the EDL record.  This
             parameter allows you to specify that time  interval  in  seconds
             (default: 2 seconds).

      -edl-start-pts
             Adjust  positions  in  EDL  records  according to playing file's
             start time.  Some formats, especially MPEG TS usually start with
             non-zero PTS values and when producing EDL file with -edlout op‐
             tion, EDL records contain absolute values that are correct  only
             for  this  particular file.  If re-encoded into a different for‐
             mat, this EDL file no longer applies.  Specifying -edl-start-pts
             will automatically adjust EDL positions according to start time:
             when producing EDL file, it will substract start time from every
             EDL record, when playing with EDL file, it will add file's start
             time to every EDL position.

      -noedl-start-pts
             Disable adjusting EDL positions.

      -enqueue (GUI only)
             Enqueue files given on the command line in the playlist  instead
             of playing them immediately.

      -fixed-vo
             Enforces  a  fixed video system for multiple files (one (un)ini‐
             tialization for all files).  Therefore only one window  will  be
             opened  for  all  files.   Currently  the  following drivers are
             fixed-vo compliant: gl, gl_tiled,  mga,  svga,  x11,  xmga,  xv,
             xvidix and dfbmga.

      -framedrop  (also  see  -hardframedrop,  experimental  without  -nocor‐
      rect-pts)
             Skip displaying some frames to maintain A/V sync  on  slow  sys‐
             tems.   Video  filters  are  not applied to such frames.  For B-
             frames even decoding is skipped completely.

      -(no)gui
             Enable or disable the GUI interface (default depends  on  binary
             name).   Only  works  as the first argument on the command line.
             Does not work as a config-file option.

      -gui-include <GUI configuration file> (also see -include) (GUI only)
             Specify a GUI configuration file to be parsed after the  default
             gui.conf.

      -h, -help, --help
             Show short summary of options.

      -hardframedrop (experimental without -nocorrect-pts)
             More  intense  frame dropping (breaks decoding).  Leads to image
             distortion!  Note that especially the libmpeg2 decoder may crash
             with this, so consider using "-vc ffmpeg12,".

      -heartbeat-cmd
             Command  that  is  executed every 30 seconds during playback via
             system() - i.e. using the shell.

             NOTE: MPlayer uses this command without any checking, it is your
             responsibility  to  ensure  it  does not cause security problems
             (e.g. make sure to use full paths if "." is in your path like on
             Windows).   It also only works when playing video (i.e. not with
             -novideo but works with -vo null).

             This can be "misused" to disable screensavers that do  not  sup‐
             port  the  proper  X  API (also see -stop-xscreensaver).  If you
             think this is too complicated, ask the author of the screensaver
             program to support the proper X APIs.

             EXAMPLE   for  xscreensaver:  mplayer  -heartbeat-cmd  "xscreen‐
             saver-command -deactivate" file

             EXAMPLE   for   GNOME   screensaver:   mplayer    -heartbeat-cmd
             "gnome-screensaver-command -p" file

      -heartbeat-interval
             Specify how often the -heartbeat-cmd should be executed, in sec‐
             onds between executions (default: 30.0).

      -identify
             Shorthand for -msglevel identify=6.  Show file parameters in  an
             easily  parseable format.  Also prints more detailed information
             about subtitle and audio track languages and IDs.  For  example,
             for  a  DVD or Blu-ray it will list the chapters and time length
             of each title, as well as a disk ID.  Combine this with  -frames
             0 to suppress all video output.  The wrapper script TOOLS/miden‐
             tify.sh suppresses the  other  MPlayer  output  and  (hopefully)
             shellescapes the filenames.  You can get less information by us‐
             ing -msglevel identify=4.

      -idle (also see -slave)
             Makes MPlayer wait idly instead of quitting  when  there  is  no
             file  to play.  Mostly useful in slave mode where MPlayer can be
             controlled through input commands.
             For gmplayer -idle is the default, -noidle will quit the GUI af‐
             ter all files have been played.

      -input <commands>
             This  option can be used to configure certain parts of the input
             system.  Paths are relative to ~/.mplayer/.
             NOTE: Autorepeat is currently only supported by joysticks.

             Available commands are:

                conf=<filename>
                     Specify input configuration file other than the  default
                     ~/.mplayer/input.conf.  ~/.mplayer/<filename> is assumed
                     if no full path is given.
                ar-dev=<device>
                     Device to be used for Apple IR Remote  (default  is  au‐
                     todetected, Linux only).
                ar-delay
                     Delay  in  milliseconds  before we start to autorepeat a
                     key (0 to disable).
                ar-rate
                     Number of key presses to generate per second on  autore‐
                     peat.
                (no)default-bindings
                     Use the key bindings that MPlayer ships with by default.
                keylist
                     Prints all keys that can be bound to commands.
                cmdlist
                     Prints all commands that can be bound to keys.
                js-dev
                     Specifies  the joystick device to use (default: /dev/in‐
                     put/js0).
                file=<filename>
                     Read commands from the given file.  Mostly useful with a
                     FIFO.
                     NOTE:  When  the given file is a FIFO MPlayer opens both
                     ends so you can do several 'echo "seek  10"  >  mp_pipe'
                     and the pipe will stay valid.

      -key-fifo-size <2-65000>
             Specify  the  size of the FIFO that buffers key events (default:
             7).  A FIFO of size n can buffer (n-1) events.   If  it  is  too
             small  some  events  may be lost.  If it is too big, MPlayer may
             seem to hang while it processes the buffered events.  To get the
             same  behavior as before this option was introduced, set it to 2
             for Linux or 1024 for Windows.  For small value you should  dis‐
             able  double-clicks by setting -doubleclick-time to 0 so they do
             not compete with regular events for buffer space.

      -lircconf <filename> (LIRC only)
             Specifies a configuration file for LIRC (default: ~/.lircrc).

      -list-properties
             Print a list of the available properties.

      -loop <number>
             Loops movie playback <number>  times.   0  means  forever.   Use
             -loop 0 to automatically reconnect to live streaming URLs.

      -menu (OSD menu only)
             Turn on OSD menu support.

      -menu-cfg <filename> (OSD menu only)
             Use an alternative menu.conf.

      -menu-chroot <path> (OSD menu only)
             Chroot the file selection menu to a specific location.

             EXAMPLE:
                -menu-chroot /home
                     Will restrict the file selection menu to /home and down‐
                     ward  (i.e.  no  access  to  /  will  be  possible,  but
                     /home/user_name will).

      -menu-keepdir (OSD menu only)
             File browser starts from the last known location instead of cur‐
             rent directory.

      -menu-root <value> (OSD menu only)
             Specify the main menu.

      -menu-startup (OSD menu only)
             Display the main menu at MPlayer startup.

      -mouse-movements
             Permit MPlayer to receive pointer events reported by  the  video
             output  driver.   Necessary  to select the buttons in DVD menus.
             Supported for X11-based VOs (x11, xv, xvmc,  etc)  and  the  gl,
             gl_tiled, direct3d and corevideo VOs.

      -noar  Turns off AppleIR remote support.

      -noconsolecontrols
             Prevent  MPlayer  from  reading  key events from standard input.
             Useful when reading data from standard input.  This is automati‐
             cally  enabled  when  - is found on the command line.  There are
             situations where you have to set it manually, e.g. if  you  open
             /dev/stdin  (or  the  equivalent on your system), use stdin in a
             playlist or intend to read from stdin later on via the  loadfile
             or loadlist slave commands.

      -nojoystick
             Turns off joystick support.

      -nolirc
             Turns off LIRC support.

      -nomouseinput
             Disable  mouse button press/release input (mozplayerxp's context
             menu relies on this option).

      -rtc (RTC only)
             Turns on usage of the Linux RTC (realtime clock -  /dev/rtc)  as
             timing  mechanism.   This wakes up the process every 1/1024 sec‐
             onds to check the current time.  Useless with modern Linux  ker‐
             nels  configured  for  desktop  use  as they already wake up the
             process with similar accuracy when using normal timed sleep.

      -pausing <0-3> (MPlayer only)
             Specifies  the  default  pausing  behaviour  of  commands,  i.e.
             whether  MPlayer will continue playback or stay paused after the
             command has finished.  See DOCS/tech/slave.txt for  further  de‐
             tails.
                0    resume
                1    pause (pausing)
                2    keep the paused / playing status (pausing_keep)
                3    toggle the paused / playing status (pausing_toggle)
                4    pause   without   frame   step   (experimental)   (paus‐
                     ing_keep_force)

      -playing-msg <string>
             Print out a string before starting playback.  The following  ex‐
             pansions are supported:

                ${NAME}
                     Expand to the value of the property NAME.

                ?(NAME:TEXT)
                     Expand TEXT only if the property NAME is available.

                ?(!NAME:TEXT)
                     Expand TEXT only if the property NAME is not available.

      -playlist <filename>
             Play  files  according to a playlist file (ASX, Winamp, SMIL, or
             one-file-per-line format).
             WARNING: The way MPlayer parses and uses playlist files  is  not
             safe  against  maliciously  constructed  files.   Such files may
             trigger harmful actions.  This has been the case for all MPlayer
             versions,  but  unfortunately  this fact was not well documented
             earlier, and some people have even misguidedly  recommended  use
             of  -playlist with untrusted sources.  Do NOT use -playlist with
             random internet sources or files you don't trust!
             NOTE: This option is considered an entry so options found  after
             it will apply only to the elements of this playlist.
             FIXME: This needs to be clarified and documented thoroughly.

      -allow-dangerous-playlist-parsing
             This enables parsing any file as a playlist if e.g. a server ad‐
             vertises a file as  playlist.   Only  enable  if  you  know  all
             servers  involved  are  trustworthy.  MPlayer's playlist code is
             not designed to handle malicious playlist files.

      -rtc-device <device>
             Use the specified device for RTC timing.

      -shuffle
             Play files in random order.

      -skin <name> (GUI only)
             Loads a skin from the directory given as parameter below the de‐
             fault  skin directories, ~/.mplayer/skins/ and /usr/local/share/
             mplayer/skins/.

             EXAMPLE:
                -skin fittyfene
                     Tries ~/.mplayer/skins/fittyfene and afterwards /usr/lo‐
                     cal/share/mplayer/skins/fittyfene.

      -slave (also see -input)
             Switches  on slave mode, in which MPlayer works as a backend for
             other programs.  Instead of intercepting keyboard events, MPlay‐
             er will read commands separated by a newline (\n) from stdin.
             NOTE:  See  -input  cmdlist  for  a  list  of slave commands and
             DOCS/tech/slave.txt for their description.  Also,  this  is  not
             intended to disable other inputs, e.g. via the video window, use
             some other method like -input  nodefault-bindings:conf=/dev/null
             for that.

      -softsleep
             Time  frames  by repeatedly checking the current time instead of
             asking the kernel to wake up MPlayer at the correct time.   Use‐
             ful  if  your  kernel timing is imprecise and you cannot use the
             RTC either.  Comes at the price of higher CPU consumption.

      -sstep <sec>
             Skip <sec> seconds after every frame.  The normal  framerate  of
             the  movie  is  kept, so playback is accelerated.  Since MPlayer
             can only seek to the next keyframe this may be inexact.

      -udp-ip <ip>
             Sets the destination address for datagrams sent by the -udp-mas‐
             ter.   Setting  it to a broadcast address allows multiple slaves
             having the same broadcast address to sync  to  the  master  (de‐
             fault: 127.0.0.1).

      -udp-master
             Send a datagram to -udp-ip on -udp-port just before playing each
             frame.  The datagram indicates  the  master's  position  in  the
             file.

      -udp-port <port>
             Sets the destination port for datagrams sent by the -udp-master,
             and the port a -udp-slave listens on (default: 23867).

      -udp-seek-threshold <sec>
             When the master seeks, the slave has to decide whether  to  seek
             as  well,  or to catch up by decoding frames without pausing be‐
             tween frames.  If the master is more  than  <sec>  seconds  away
             from  the slave, the slave seeks.  Otherwise, it "runs" to catch
             up or waits for the master.  This should almost always  be  left
             at its default setting of 1 second.

      -udp-slave
             Listen on -udp-port and match the master's position.

DEMUXER/STREAM OPTIONS
      -a52drc <level>
             Select  the  Dynamic  Range  Compression  level  for  AC-3 audio
             streams.  <level> is a float value ranging from 0 to 1, where  0
             means  no  compression  and  1 (which is the default) means full
             compression (make loud passages more  silent  and  vice  versa).
             Values  up  to 2 are also accepted, but are purely experimental.
             This option only shows an effect if the AC-3 stream contains the
             required range compression information.

      -aid <ID> (also see -alang)
             Select  audio channel (MPEG: 0-31, AVI/OGM: 1-99, ASF/RM: 0-127,
             VOB(AC-3):  128-159,  VOB(LPCM):  160-191,   MPEG-TS   17-8190).
             MPlayer  prints the available audio IDs when run in verbose (-v)
             mode.  When playing an MPEG-TS stream, MPlayer/MEncoder will use
             the first program (if present) with the chosen audio stream.

      -ausid <ID> (also see -alang)
             Select  audio  substream  channel.  Currently the valid range is
             0x55..0x75 and applies only to MPEG-TS when handled by  the  na‐
             tive  demuxer  (not by libavformat).  The format type may not be
             correctly identified because of how this  information  (or  lack
             thereof)  is embedded in the stream, but it will demux correctly
             the audio streams when multiple substreams are present.  MPlayer
             prints the available substream IDs when run with -identify.

      -alang <language code[,language code,...]> (also see -aid)
             Specify  a  priority  list of audio languages to use.  Different
             container formats employ different language codes.  DVDs use ISO
             639-1  two  letter language codes, Matroska, MPEG-TS and NUT use
             ISO 639-2 three letter language codes while OGM uses a free-form
             identifier.   MPlayer prints the available languages when run in
             verbose (-v) mode.

             EXAMPLE:
                mplayer dvd://1 -alang hu,en
                     Chooses the Hungarian language track on a DVD and  falls
                     back on English if Hungarian is not available.
                mplayer -alang jpn example.mkv
                     Plays a Matroska file in Japanese.

      -audio-demuxer <[+]name> (-audiofile only)
             Force  audio  demuxer type for -audiofile.  Use a '+' before the
             name to force it, this will skip some checks!  Give the  demuxer
             name  as  printed by -audio-demuxer help.  For backward compati‐
             bility it also accepts the demuxer ID as defined in  libmpdemux/
             demuxer.h.   -audio-demuxer  audio  or  -audio-demuxer 17 forces
             MP3.

      -audiofile <filename>
             Play audio from an external file (WAV, MP3 or Ogg Vorbis)  while
             viewing a movie.

      -audiofile-cache <kBytes>
             Enables  caching  for  the  stream used by -audiofile, using the
             specified amount of memory.

      -reuse-socket (udp:// only)
             Allows a socket to be reused by other processes as soon as it is
             closed.

      -bandwidth <Bytes> (network only)
             Specify the maximum bandwidth for network streaming (for servers
             that are able to send content in different bitrates).  Useful if
             you  want to watch live streamed media behind a slow connection.
             With Real RTSP streaming, it is also used to set the maximum de‐
             livery  bandwidth allowing faster cache filling and stream dump‐
             ing.

      -bluray-angle <angle ID> (Blu-ray only)
             Some Blu-ray discs contain scenes that can be viewed from multi‐
             ple  angles.  Here you can tell MPlayer which angles to use (de‐
             fault: 1).

      -bluray-device <path to disc> (Blu-ray only)
             Specify the Blu-ray disc location.  Must  be  a  directory  with
             Blu-ray structure.

      -cache <kBytes>
             This  option  specifies  how much memory (in kBytes) to use when
             precaching a file or URL.  Especially useful on slow media.

      -nocache
             Turns off caching.

      -cache-min <percentage>
             Playback will start when the cache has been filled up  to  <per‐
             centage> of the total.

      -cache-seek-min <percentage>
             If a seek is to be made to a position within <percentage> of the
             cache size from the current position, MPlayer will wait for  the
             cache  to  be  filled  to this position rather than performing a
             stream seek (default: 50).

      -capture (MPlayer only)
             Allows capturing the primary stream (not additional audio tracks
             or  other  kind of streams) into the file specified by -dumpfile
             or by default.  If this option is given, capturing can be start‐
             ed  and  stopped by pressing the key bound to this function (see
             section INTERACTIVE CONTROL).  Same  as  for  -dumpstream,  this
             will  likely  not  produce usable results for anything else than
             MPEG streams.  Note that, due to cache latencies, captured  data
             may begin and end somewhat delayed compared to what you see dis‐
             played.

      -cdda <option1:option2> (CDDA only)
             This option can be used to tune the CD Audio reading feature  of
             MPlayer.

             Available options are:

                speed=<value>
                     Set CD spin speed.

                paranoia=<0-2>
                     Set  paranoia  level.  Values other than 0 seem to break
                     playback of anything but the first track.
                        0: disable checking (default)
                        1: overlap checking only
                        2: full data correction and verification

                generic-dev=<value>
                     Use specified generic SCSI device.

                sector-size=<value>
                     Set atomic read size.

                overlap=<value>
                     Force minimum  overlap  search  during  verification  to
                     <value> sectors.

                toc-bias
                     Assume  that the beginning offset of track 1 as reported
                     in the TOC will be addressed as  LBA  0.   Some  Toshiba
                     drives need this for getting track boundaries correct.

                toc-offset=<value>
                     Add <value> sectors to the values reported when address‐
                     ing tracks.  May be negative.

                (no)skip
                     (Never) accept imperfect data reconstruction.

      -cdrom-device <path to device>
             Specify the CD-ROM device (default: /dev/cdrom).

      -channels <number> (also see -af channels)
             Request the number of playback channels (default:  2).   MPlayer
             asks  the  decoder  to decode the audio into as many channels as
             specified.  Then it is up to the decoder to fulfill the require‐
             ment.   This  is usually only important when playing videos with
             AC-3 audio (like DVDs).  In that case liba52 does  the  decoding
             by  default and correctly downmixes the audio into the requested
             number of channels.  To directly control the  number  of  output
             channels independently of how many channels are decoded, use the
             channels filter.
             NOTE: This option is honored  by  codecs  (AC-3  only),  filters
             (surround) and audio output drivers (OSS at least).

             Available options are:

                2    stereo
                4    surround
                6    full 5.1
                8    full 7.1

      -chapter <chapter ID>[-<endchapter ID>]
             Specify  which  chapter to start playing at.  Optionally specify
             which chapter to end playing at (default: 1).

      -cookies (network only)
             Send cookies when making HTTP requests.

      -cookies-file <filename> (network only)
             Read HTTP cookies  from  <filename>  (default:  ~/.mozilla/  and
             ~/.netscape/) and skip reading from default locations.  The file
             is assumed to be in Netscape format.

      -delay <sec>
             audio delay in seconds (positive or negative float value)
             Negative values delay the audio, and positive values  delay  the
             video.  Note that this is the exact opposite of the -audio-delay
             MEncoder option.
             NOTE: When used with MEncoder, this is not  guaranteed  to  work
             correctly with -ovc copy; use -audio-delay instead.

      -ignore-start
             Ignore the specified starting time for streams in AVI files.  In
             MPlayer, this nullifies stream delays in files encoded with  the
             -audio-delay option.  During encoding, this option prevents MEn‐
             coder from transferring original stream start times to  the  new
             file;  the  -audio-delay option is not affected.  Note that MEn‐
             coder sometimes adjusts stream starting times  automatically  to
             compensate  for  anticipated decoding delays, so do not use this
             option for encoding without testing it first.

      -demuxer <[+]name>
             Force demuxer type.  Use a '+' before the name to force it, this
             will skip some checks!  Give the demuxer name as printed by -de‐
             muxer help.  For backward compatibility it also accepts the  de‐
             muxer ID as defined in libmpdemux/demuxer.h.

      -dumpaudio (MPlayer only)
             Dumps  raw compressed audio stream to ./stream.dump (useful with
             MPEG/AC-3, in most other cases the resulting file  will  not  be
             playable).  If you give more than one of -dumpaudio, -dumpvideo,
             -dumpstream on the command line only the last one will work.

      -dumpfile <filename> (MPlayer only)
             Specify which file MPlayer should dump to.  Should be  used  to‐
             gether with -dumpaudio / -dumpvideo / -dumpstream / -capture.

      -dumpstream (MPlayer only)
             Dumps the raw stream to ./stream.dump.  Useful when ripping from
             DVD or network.  If  you  give  more  than  one  of  -dumpaudio,
             -dumpvideo,  -dumpstream  on  the command line only the last one
             will work.

      -dumpvideo (MPlayer only)
             Dump raw compressed video stream to ./stream.dump (not very  us‐
             able).   If  you  give  more than one of -dumpaudio, -dumpvideo,
             -dumpstream on the command line only the last one will work.

      -dvbin <options> (DVB only)
             Pass the following parameters to the DVB input module, in  order
             to override the default ones:

                card=<1-4>
                     Specifies using card number 1-4 (default: 1).
                file=<filename>
                     Instructs  MPlayer to read the channels list from <file‐
                     name>.         Default        is        ~/.mplayer/chan‐
                     nels.conf.{sat,ter,cbl,atsc}  (based  on your card type)
                     or ~/.mplayer/channels.conf as a last resort.
                timeout=<1-240>
                     Maximum number of seconds to wait when trying to tune  a
                     frequency before giving up (default: 30).

      -dvd-device <path to device> (DVD only)
             Specify  the  DVD  device  or .iso filename (default: /dev/dvd).
             You can also specify a directory that contains files  previously
             copied directly from a DVD (with e.g. vobcopy).

      -dvd-speed <factor or speed in KB/s> (DVD only)
             Try  to limit DVD speed (default: 0, no change).  DVD base speed
             is about 1350KB/s, so a 8x  drive  can  read  at  speeds  up  to
             10800KB/s.   Slower speeds make the drive more quiet, for watch‐
             ing DVDs 2700KB/s should be quiet and fast enough.  MPlayer  re‐
             sets the speed to the drive default value on close.  Values less
             than 100 mean multiples of 1350KB/s, i.e. -dvd-speed  8  selects
             10800KB/s.
             NOTE:  You  need  write  access  to the DVD device to change the
             speed.

      -dvdangle <angle ID> (DVD only)
             Some DVD discs contain scenes that can be viewed  from  multiple
             angles.  Here you can tell MPlayer which angles to use (default:
             1).

      -edl <filename>
             Enables edit decision list (EDL) actions during playback.  Video
             will be skipped over and audio will be muted and unmuted accord‐
             ing to the entries in the given file.   See  http://www.mplayer‐
             hq.hu/DOCS/HTML/en/edl.html for details on how to use this.

      -endpos <[[hh:]mm:]ss[.ms]|size[b|kb|mb]> (also see -ss and -sb)
             Stop at given time or byte position.
             NOTE:  Byte position may not be accurate, as it can only stop at
             a frame boundary.  When used in  conjunction  with  -ss  option,
             -endpos time will shift forward by seconds specified with -ss if
             not a byte position.  In addition it may not work well or not at
             all when used with any of the -dump options.

             EXAMPLE:
                -endpos 56
                     Stop at 56 seconds.
                -endpos 01:10:00
                     Stop at 1 hour 10 minutes.
                -ss 10 -endpos 56
                     Stop at 1 minute 6 seconds.
                mplayer -endpos 100mb
                     Stop playback after reading 100MB of the input file.
                mencoder -endpos 100mb
                     Encode only 100 MB.

      -forceidx
             Force index rebuilding.  Useful for files with broken index (A/V
             desync, etc).  This will enable seeking in files  where  seeking
             was  not  possible.  You can fix the index permanently with MEn‐
             coder (see the documentation).
             NOTE: This option only works if the  underlying  media  supports
             seeking (i.e. not with stdin, pipe, etc).

      -fps <float value>
             Override video framerate.  Useful if the original value is wrong
             or missing.

      -frames <number>
             Play/convert only first <number> frames, then quit.

      -hr-mp3-seek (MP3 only)
             Hi-res MP3 seeking.  Enabled when playing from an  external  MP3
             file,  as we need to seek to the very exact position to keep A/V
             sync.  Can be slow especially when seeking  backwards  since  it
             has to rewind to the beginning to find an exact frame position.

      -http-header-fields <field1,field2>
             Set custom HTTP fields when accessing HTTP stream.

             EXAMPLE:
                     mplayer  -http-header-fields  'Field1:  value1','Field2:
                     value2' http://localhost:1234
                     Will generate HTTP request:
                        GET / HTTP/1.0
                        Host: localhost:1234
                        User-Agent: MPlayer
                        Icy-MetaData: 1
                        Field1: value1
                        Field2: value2
                        Connection: close

      -idx (also see -forceidx)
             Rebuilds index of files if no index was found, allowing seeking.
             Useful with broken/incomplete downloads, or badly created files.
             NOTE:  This  option  only works if the underlying media supports
             seeking (i.e. not with stdin, pipe, etc).

      -noidx Skip rebuilding index file.  MEncoder skips  writing  the  index
             with this option.

      -ipv4-only-proxy (network only)
             Skip  the  proxy  for IPv6 addresses.  It will still be used for
             IPv4 connections.

      -lavfstreamopts <options>
             Specify extra options for libavformat based streams.

      -loadidx <index file>
             The file from which to  read  the  video  index  data  saved  by
             -saveidx.   This  index will be used for seeking, overriding any
             index data contained in the AVI itself.  MPlayer will  not  pre‐
             vent  you  from loading an index file generated from a different
             AVI, but this is sure to cause unfavorable results.
             NOTE: This option is obsolete now that MPlayer has OpenDML  sup‐
             port.

      -mc <seconds/frame>
             maximum A-V sync correction per frame (in seconds)
             -mc  0 should always be combined with -noskip for mencoder, oth‐
             erwise it will almost certainly cause A-V desync.

      -mf <option1:option2:...>
             Used when decoding from multiple PNG or JPEG files.

             Available options are:

                w=<value>
                     input file width (default: autodetect)
                h=<value>
                     input file height (default: autodetect)
                fps=<value>
                     output fps (default: 25)
                type=<value>
                     input file type (available: jpeg, png, tga, sgi)

      -ni    Force treating files as non-interleaved.  In  particular  forces
             usage  of non-interleaved AVI parser (fixes playback of some bad
             AVI files).  Can also help playing files that otherwise play au‐
             dio and video alternating instead of at the same time.  This can
             significantly increase memory usage, thus it would be preferable
             to fix interleaving of affected files.

      -nobps (AVI only)
             Do  not  use average byte/second value for A-V sync.  Helps with
             some AVI files with broken header.

      -noextbased
             Disables extension-based demuxer selection.   By  default,  when
             the  file  type  (demuxer) cannot be detected reliably (the file
             has no header or it is not reliable enough), the filename exten‐
             sion  is  used to select the demuxer.  Always falls back on con‐
             tent-based demuxer selection.

      -passwd <password> (also see -user) (network only)
             Specify password for HTTP authentication.

      -prefer-ipv4 (network only)
             Use IPv4 on network connections.  Falls back on  IPv6  automati‐
             cally.

      -prefer-ipv6 (IPv6 network only)
             Use  IPv6  on network connections.  Falls back on IPv4 automati‐
             cally.

      -psprobe <byte position>
             When playing an MPEG-PS or MPEG-PES streams,  this  option  lets
             you  specify  how  many  bytes in the stream you want MPlayer to
             scan in order to identify the video codec used.  This option  is
             needed to play EVO or VDR files containing H.264 streams.

      -pvr <option1:option2:...> (PVR only)
             This option tunes various encoding properties of the PVR capture
             module.  It has to be used with any hardware MPEG encoder  based
             card   supported  by  the  V4L2  driver.   The  Hauppauge  WinTV
             PVR-150/250/350/500 and all IVTV based cards are  known  as  PVR
             capture cards.  Be aware that only Linux 2.6.18 kernel and above
             is able to handle MPEG stream through V4L2 layer.  For  hardware
             capture of an MPEG stream and watching it with MPlayer/MEncoder,
             use 'pvr://' as a movie URL.

             Available options are:

                aspect=<0-3>
                     Specify input aspect ratio:
                        0: 1:1
                        1: 4:3 (default)
                        2: 16:9
                        3: 2.21:1

                arate=<32000-48000>
                     Specify encoding audio rate (default: 48000  Hz,  avail‐
                     able: 32000, 44100 and 48000 Hz).

                alayer=<1-5>
                     Specify MPEG audio layer encoding (default: 2).

                abitrate=<32-448>
                     Specify audio encoding bitrate in kbps (default: 384).

                amode=<value>
                     Specify  audio  encoding  mode.  Available preset values
                     are 'stereo', 'joint_stereo',  'dual'  and  'mono'  (de‐
                     fault: stereo).

                vbitrate=<value>
                     Specify average video bitrate encoding in Mbps (default:
                     6).

                vmode=<value>
                     Specify video encoding mode:
                        vbr: Variable BitRate (default)
                        cbr: Constant BitRate

                vpeak=<value>
                     Specify peak video bitrate encoding in Mbps (only useful
                     for VBR encoding, default: 9.6).

                fmt=<value>
                     Choose an MPEG format for encoding:
                        ps:    MPEG-2 Program Stream (default)
                        ts:    MPEG-2 Transport Stream
                        mpeg1: MPEG-1 System Stream
                        vcd:   Video CD compatible stream
                        svcd:  Super Video CD compatible stream
                        dvd:   DVD compatible stream

      -radio <option1:option2:...> (radio only)
             These  options  set various parameters of the radio capture mod‐
             ule.  For listening to radio with MPlayer use 'radio://<frequen‐
             cy>' (if channels option is not given) or 'radio://<channel_num‐
             ber>' (if channels option is given) as a movie URL.  You can see
             allowed  frequency range by running MPlayer with '-v'.  To start
             the grabbing subsystem, use 'radio://<frequency or channel>/cap‐
             ture'.   If  the  capture keyword is not given you can listen to
             radio using the line-in cable only.  Using capture to listen  is
             not  recommended  due  to  synchronization problems, which makes
             this process uncomfortable.

             Available options are:

                device=<value>
                     Radio device to use (default: /dev/radio0 for Linux  and
                     /dev/tuner0 for *BSD).

                driver=<value>
                     Radio  driver to use (default: v4l2 if available, other‐
                     wise v4l).  Currently, v4l and v4l2 drivers are support‐
                     ed.

                volume=<0..100>
                     sound volume for radio device (default 100)

                freq_min=<value> (*BSD BT848 only)
                     minimum allowed frequency (default: 87.50)

                freq_max=<value> (*BSD BT848 only)
                     maximum allowed frequency (default: 108.00)

                channels=<frequency>-<name>,<frequency>-<name>,...
                     Set  channel  list.   Use _ for spaces in names (or play
                     with quoting ;-).  The channel names will then be  writ‐
                     ten  using OSD and the slave commands radio_step_channel
                     and radio_set_channel will be usable for a  remote  con‐
                     trol  (see LIRC).  If given, number in movie URL will be
                     treated as channel position in channel list.
                     EXAMPLE: radio://1, radio://104.4, radio_set_channel 1

                adevice=<value> (radio capture only)
                     Name of device to capture sound from.   Without  such  a
                     name  capture will be disabled, even if the capture key‐
                     word appears in the URL.  For ALSA devices use it in the
                     form  hw=<card>.<device>.  If the device name contains a
                     '=', the module will use ALSA to capture, otherwise OSS.

                arate=<value> (radio capture only)
                     Rate in samples per second (default: 44100).
                     NOTE:  When  using  audio  capture  set  also  -rawaudio
                     rate=<value>  option  with  the same value as arate.  If
                     you have problems with sound speed (runs  too  quickly),
                     try   to   play   with   different   rate  values  (e.g.
                     48000,44100,32000,...).

                achannels=<value> (radio capture only)
                     Number of audio channels to capture.

      -rawaudio <option1:option2:...>
             This option lets you play raw audio files.  You have to use -de‐
             muxer  rawaudio  as well.  It may also be used to play audio CDs
             which are not 44kHz 16-bit stereo.  For playing raw AC-3 streams
             use -rawaudio format=0x2000 -demuxer rawaudio.

             Available options are:

                channels=<value>
                     number of channels
                rate=<value>
                     rate in samples per second
                samplesize=<value>
                     sample size in bytes
                bitrate=<value>
                     bitrate for rawaudio files
                format=<value>
                     fourcc in hex

      -rawvideo <option1:option2:...>
             This option lets you play raw video files.  You have to use -de‐
             muxer rawvideo as well.

             Available options are:

                fps=<value>
                     rate in frames per second (default: 25.0)
                sqcif|qcif|cif|4cif|pal|ntsc
                     set standard image size
                w=<value>
                     image width in pixels
                h=<value>
                     image height in pixels
                i420|yv12|yuy2|y8
                     set colorspace
                format=<value>
                     colorspace (fourcc) in  hex  or  string  constant.   Use
                     -rawvideo format=help for a list of possible strings.
                size=<value>
                     frame size in Bytes

             EXAMPLE:
                mplayer foreman.qcif -demuxer rawvideo -rawvideo qcif
                     Play the famous "foreman" sample video.
                mplayer   sample-720x576.yuv   -demuxer   rawvideo  -rawvideo
                w=720:h=576
                     Play a raw YUV sample.

      -referrer <string> (network only)
             Specify a referrer path or URL for HTTP requests.

      -rtsp-port
             Used with 'rtsp://' URLs to  force  the  client's  port  number.
             This option may be useful if you are behind a router and want to
             forward the RTSP stream from the server to a specific client.

      -rtsp-destination
             Used with 'rtsp://' URLs to force the destination IP address  to
             be bound.  This option may be useful with some RTSP server which
             do not send RTP packets to the right interface.  If the  connec‐
             tion  to  the  RTSP server fails, use -v to see which IP address
             MPlayer tries to bind to and try to force it to one assigned  to
             your computer instead.

      -rtsp-stream-over-tcp (LIVE555 and NEMESI only)
             Used  with 'rtsp://' URLs to specify that the resulting incoming
             RTP and RTCP packets be streamed over TCP (using  the  same  TCP
             connection  as  RTSP).   This option may be useful if you have a
             broken internet connection that does not pass incoming UDP pack‐
             ets (see http://www.live555.com/mplayer/).

      -rtsp-stream-over-http (LIVE555 only)
             Used  with 'http://' URLs to specify that the resulting incoming
             RTP and RTCP packets be streamed over HTTP.

      -saveidx <filename>
             Force index rebuilding and dump the index to  <filename>.   Cur‐
             rently this only works with AVI files.
             NOTE:  This option is obsolete now that MPlayer has OpenDML sup‐
             port.

      -sb <byte position> (also see -ss)
             Seek to byte position.  Useful for playback from  CD-ROM  images
             or VOB files with junk at the beginning.

      -speed <0.01-100>
             Slow down or speed up playback by the factor given as parameter.
             Not guaranteed to  work  correctly  with  -oac  copy.   Add  -af
             scaletempo to get past the 4x limit on playback.

      -srate <Hz>
             Select  the output sample rate to be used (of course sound cards
             have limits on this).  If the sample frequency selected is  dif‐
             ferent  from  that of the current media, the resample or lavcre‐
             sample audio filter will be inserted into the audio filter layer
             to compensate for the difference.  The type of resampling can be
             controlled by the -af-adv option.  The default  is  fast  resam‐
             pling that may cause distortion.

      -ss <time> (also see -sb)
             Seek  to given time position.  Use -ss nopts to disable seeking,
             -ss 0 has different behaviour.

             EXAMPLE:
                -ss 56
                     Seeks to 56 seconds.
                -ss 01:10:00
                     Seeks to 1 hour 10 min.

      -tskeepbroken
             Tells MPlayer not to discard TS packets reported  as  broken  in
             the stream.  Sometimes needed to play corrupted MPEG-TS files.

      -tsprobe <byte position>
             When playing an MPEG-TS stream, this option lets you specify how
             many bytes in the stream you want MPlayer to search for the  de‐
             sired audio and video IDs.

      -tsprog <1-65534>
             When playing an MPEG-TS stream, you can specify with this option
             which program (if present) you want to play.  Can be  used  with
             -vid and -aid.

      -tv <option1:option2:...> (TV/PVR only)
             This  option  tunes various properties of the TV capture module.
             For watching TV with MPlayer, use 'tv://' or 'tv://<channel_num‐
             ber>'  or  even  'tv://<channel_name>  (see  option channels for
             channel_name  below)  as  a  movie  URL.   You  can   also   use
             'tv:///<input_id>' to start watching a movie from a composite or
             S-Video input (see option input for details).

             Available options are:

                noaudio
                     no sound

                automute=<0-255> (v4l and v4l2 only)
                     If signal strength reported by device is less than  this
                     value, audio and video will be muted.  In most cases au‐
                     tomute=100 will be enough.  Default is 0 (automute  dis‐
                     abled).

                driver=<value>
                     See  -tv  driver=help for a list of compiled-in TV input
                     drivers.  available: dummy,  v4l,  v4l2,  bsdbt848  (de‐
                     fault: autodetect)

                device=<value>
                     Specify TV device (default: /dev/video0).  NOTE: For the
                     bsdbt848 driver you can provide both bktr and tuner  de‐
                     vice names separating them with a comma, tuner after bk‐
                     tr (e.g. -tv device=/dev/bktr1,/dev/tuner1).

                input=<value>
                     Specify input (default: 0 (TV), see console  output  for
                     available inputs).

                freq=<value>
                     Specify   the  frequency  to  set  the  tuner  to  (e.g.
                     511.250).  Not compatible with the channels parameter.

                outfmt=<value>
                     Specify the output format of the  tuner  with  a  preset
                     value  supported  by the V4L driver (yv12, rgb32, rgb24,
                     rgb16, rgb15, uyvy, yuy2, i420) or an  arbitrary  format
                     given  as  hex value.  Try outfmt=help for a list of all
                     available formats.

                width=<value>
                     output window width

                height=<value>
                     output window height

                fps=<value>
                     framerate at which to capture video (frames per second)

                buffersize=<value>
                     maximum size of the capture  buffer  in  megabytes  (de‐
                     fault: dynamical)

                norm=<value>
                     For  bsdbt848  and  v4l, PAL, SECAM, NTSC are available.
                     For v4l2, see the console  output  for  a  list  of  all
                     available norms, also see the normid option below.

                normid=<value> (v4l2 only)
                     Sets  the  TV norm to the given numeric ID.  The TV norm
                     depends on the capture card.  See the console output for
                     a list of available TV norms.

                channel=<value>
                     Set tuner to <value> channel.

                chanlist=<value>
                     available:  argentina,  australia,  china-bcast, europe-
                     east, europe-west, france, ireland, italy,  japan-bcast,
                     japan-cable,  newzealand, russia, southafrica, us-bcast,
                     us-cable, us-cable-hrc

                channels=<chan>-<name>[=<norm>],<chan>-<name>[=<norm>],...
                     Set names for channels.  NOTE: If <chan> is  an  integer
                     greater  than  1000, it will be treated as frequency (in
                     kHz) rather than channel name from frequency table.
                     Use _ for spaces in names (or  play  with  quoting  ;-).
                     The  channel  names  will then be written using OSD, and
                     the slave commands tv_step_channel,  tv_set_channel  and
                     tv_last_channel will be usable for a remote control (see
                     LIRC).  Not compatible with the frequency parameter.
                     NOTE: The channel number will then be  the  position  in
                     the 'channels' list, beginning with 1.
                     EXAMPLE:    tv://1,    tv://TV1,    tv_set_channel    1,
                     tv_set_channel TV1

                [brightness|contrast|hue|saturation]=<-100-100>
                     Set the image equalizer on the card.

                audiorate=<value>
                     Set input audio sample rate.

                forceaudio
                     Capture audio even if there are no audio sources report‐
                     ed by v4l.

                alsa
                     Capture from ALSA.

                amode=<0-3>
                     Choose an audio mode:
                        0: mono
                        1: stereo
                        2: language 1
                        3: language 2

                forcechan=<1-2>
                     By  default, the count of recorded audio channels is de‐
                     termined automatically by querying the audio  mode  from
                     the  TV  card.   This  option allows forcing stereo/mono
                     recording regardless of the amode option and the  values
                     returned  by  v4l.  This can be used for troubleshooting
                     when the TV card is unable to report the  current  audio
                     mode.

                adevice=<value>
                     Set an audio device.  <value> should be /dev/xxx for OSS
                     and a hardware ID for ALSA.  You must replace any ':' by
                     a '.' in the hardware ID for ALSA.

                audioid=<value>
                     Choose  an  audio  output of the capture card, if it has
                     more than one.

                [volume|bass|treble|balance]=<0-65535> (v4l1)

                [volume|bass|treble|balance]=<0-100> (v4l2)
                     These options set parameters of the mixer on  the  video
                     capture  card.   They  will have no effect, if your card
                     does not have one.  For v4l2 50 maps to the default val‐
                     ue of the control, as reported by the driver.

                gain=<0-100> (v4l2)
                     Set  gain control for video devices (usually webcams) to
                     the desired value and switch off automatic  control.   A
                     value of 0 enables automatic control.  If this option is
                     omitted, gain control will not be modified.

                immediatemode=<bool>
                     A value of 0 means capture and buffer  audio  and  video
                     together  (default for MEncoder).  A value of 1 (default
                     for MPlayer) means to do video capture only and let  the
                     audio  go  through  a loopback cable from the TV card to
                     the sound card.

                mjpeg
                     Use hardware MJPEG compression  (if  the  card  supports
                     it).  When using this option, you do not need to specify
                     the width and  height  of  the  output  window,  because
                     MPlayer will determine it automatically from the decima‐
                     tion value (see below).

                decimation=<1|2|4>
                     choose the size of the picture that will  be  compressed
                     by hardware MJPEG compression:
                        1: full size
                            704x576    PAL
                            704x480    NTSC
                        2: medium size
                            352x288    PAL
                            352x240    NTSC
                        4: small size
                            176x144    PAL
                            176x120    NTSC

                quality=<0-100>
                     Choose  the quality of the JPEG compression (< 60 recom‐
                     mended for full size).

                tdevice=<value>
                     Specify TV teletext  device  (example:  /dev/vbi0)  (de‐
                     fault: none).

                tformat=<format>
                     Specify TV teletext display format (default: 0):
                        0: opaque
                        1: transparent
                        2: opaque with inverted colors
                        3: transparent with inverted colors

                tpage=<100-899>
                     Specify initial TV teletext page number (default: 100).

                tlang=<-1-127>
                     Specify  default  teletext  language  code (default: 0),
                     which will be used as primary language until a  type  28
                     packet is received.  Useful when the teletext system us‐
                     es a non-latin character set, but language codes are not
                     transmitted  via  teletext type 28 packets for some rea‐
                     son.  To see a list of supported language codes set this
                     option to -1.

                hidden_video_renderer (dshow only)
                     Terminate  stream  with  video  renderer instead of Null
                     renderer (default: off).  Will help if video freezes but
                     audio does not.  NOTE: May not work with -vo directx and
                     -vf crop combination.

                hidden_vp_renderer (dshow only)
                     Terminate VideoPort pin stream with video  renderer  in‐
                     stead  of  removing  it  from  the graph (default: off).
                     Useful if your card has a VideoPort  pin  and  video  is
                     choppy.   NOTE:  May  not  work with -vo directx and -vf
                     crop combination.

                system_clock (dshow only)
                     Use the system clock as sync source instead of  the  de‐
                     fault  graph  clock  (usually  the clock from one of the
                     live sources in graph).

                normalize_audio_chunks (dshow only)
                     Create audio chunks with a time length  equal  to  video
                     frame time length (default: off).  Some audio cards cre‐
                     ate audio chunks about 0.5s in size, resulting in choppy
                     video when using immediatemode=0.

      -tvscan <option1:option2:...> (TV and MPlayer only)
             Tune  the TV channel scanner.  MPlayer will also print value for
             "-tv channels=" option, including existing and just found  chan‐
             nels.

             Available suboptions are:

                autostart
                     Begin  channel  scanning  immediately after startup (de‐
                     fault: disabled).

                period=<0.1-2.0>
                     Specify delay in seconds before switching to next  chan‐
                     nel  (default:  0.5).   Lower  values  will cause faster
                     scanning, but can detect inactive TV channels as active.

                threshold=<1-100>
                     Threshold value for the signal strength (in percent), as
                     reported by the device (default: 50).  A signal strength
                     higher than this value will indicate that the  currently
                     scanning channel is active.

      -user <username> (also see -passwd) (network only)
             Specify username for HTTP authentication.

      -user-agent <string>
             Use <string> as user agent for HTTP streaming.

      -vid <ID>
             Select  video channel (MPG: 0-15, ASF: 0-255, MPEG-TS: 17-8190).
             When playing an MPEG-TS stream, MPlayer/MEncoder  will  use  the
             first program (if present) with the chosen video stream.

      -vivo <suboption> (DEBUG CODE)
             Force  audio parameters for the VIVO demuxer (for debugging pur‐
             poses).  FIXME: Document this.

OSD/SUBTITLE OPTIONS
      NOTE: Also see -vf expand.

      -ass (FreeType only)
             Turn on SSA/ASS subtitle rendering.  With  this  option,  libass
             will be used for SSA/ASS external subtitles and Matroska tracks.
             You may also want to use -embeddedfonts.
             NOTE: Unlike normal OSD, libass uses fontconfig by  default.  To
             disable it, use -nofontconfig.

      -ass-border-color <value>
             Sets  the  border (outline) color for text subtitles.  The color
             format is RRGGBBAA.

      -ass-bottom-margin <value>
             Adds a black band at the bottom of the frame.  The SSA/ASS  ren‐
             derer can place subtitles there (with -ass-use-margins).

      -ass-color <value>
             Sets  the  color for text subtitles.  The color format is RRGGB‐
             BAA.

      -ass-font-scale <value>
             Set the scale coefficient to be used for fonts  in  the  SSA/ASS
             renderer.

      -ass-force-style <[Style.]Param=Value[,...]>
             Override some style or script info parameters.

             EXAMPLE:
                -ass-force-style FontName=Arial,Default.Bold=1
                -ass-force-style PlayResY=768

      -ass-hinting <type>
             Set hinting type.  <type> can be:
                0    no hinting
                1    FreeType autohinter, light mode
                2    FreeType autohinter, normal mode
                3    font native hinter
                0-3 + 4
                     The  same, but hinting will only be performed if the OSD
                     is rendered at screen resolution and will therefore  not
                     be scaled.
                The  default  value  is 7 (use native hinter for unscaled OSD
                and no hinting otherwise).

      -ass-line-spacing <value>
             Set line spacing value for SSA/ASS renderer.

      -ass-styles <filename>
             Load all SSA/ASS styles found in the specified file and use them
             for rendering text subtitles.  The syntax of the file is exactly
             like the [V4 Styles] / [V4+ Styles] section of SSA/ASS.

      -ass-top-margin <value>
             Adds a black band at the top of the frame.  The SSA/ASS renderer
             can place toptitles there (with -ass-use-margins).

      -ass-use-margins
             Enables  placing  toptitles  and subtitles in black borders when
             they are available.

      -dumpjacosub (MPlayer only)
             Convert the given subtitle (specified with the -sub  option)  to
             the  time-based  JACOsub  subtitle format.  Creates a dumpsub.js
             file in the current directory.

      -dumpmicrodvdsub (MPlayer only)
             Convert the given subtitle (specified with the -sub  option)  to
             the MicroDVD subtitle format.  Creates a dumpsub.sub file in the
             current directory.

      -dumpmpsub (MPlayer only)
             Convert the given subtitle (specified with the -sub  option)  to
             MPlayer's  subtitle format, MPsub.  Creates a dump.mpsub file in
             the current directory.

      -dumpsami (MPlayer only)
             Convert the given subtitle (specified with the -sub  option)  to
             the time-based SAMI subtitle format.  Creates a dumpsub.smi file
             in the current directory.

      -dumpsrtsub (MPlayer only)
             Convert the given subtitle (specified with the -sub  option)  to
             the time-based SubViewer (SRT) subtitle format.  Creates a dump‐
             sub.srt file in the current directory.
             NOTE: Some broken hardware players choke on SRT  subtitle  files
             with  Unix line endings.  If you are unlucky enough to have such
             a box, pass your subtitle files through unix2dos  or  a  similar
             program  to replace Unix line endings with DOS/Windows line end‐
             ings.

      -dumpsub (MPlayer only) (BETA CODE)
             Dumps the subtitle substream from VOB  streams.   Also  see  the
             -dump*sub and -vobsubout* options.

      -embeddedfonts (FreeType only)
             Enables  extraction  of  Matroska  embedded fonts (default: dis‐
             abled).  These fonts can be used for SSA/ASS subtitle  rendering
             (-ass  option).   Font files are created in the ~/.mplayer/fonts
             directory.
             NOTE: With FontConfig 2.4.2 or newer, embedded fonts are  opened
             directly from memory, and this option is enabled by default.

      -ffactor <number>
             Resample the font alphamap.  Can be:
                0    plain white fonts
                0.75 very narrow black outline (default)
                1    narrow black outline
                10   bold black outline

      -flip-hebrew (FriBiDi only)
             Turns on flipping subtitles using FriBiDi.

      -noflip-hebrew-commas
             Change  FriBiDi's  assumptions about the placements of commas in
             subtitles.  Use this if commas in subtitles  are  shown  at  the
             start of a sentence instead of at the end.

      -font  <path  to  font.desc file, path to font (FreeType), font pattern
      (Fontconfig)>
             Search for the OSD/subtitle fonts in  an  alternative  directory
             (default  for  normal  fonts: ~/.mplayer/font/font.desc, default
             for FreeType fonts: ~/.mplayer/subfont.ttf, default for Fontcon‐
             fig: "sans-serif").
             NOTE: With FreeType, this option determines the path to the font
             file.  With Fontconfig, this option  determines  the  Fontconfig
             font pattern.

             EXAMPLE:
                -font ~/.mplayer/arial-14/font.desc
                -font ~/.mplayer/arialuni.ttf
                -font 'Bitstream Vera Sans'
                -font 'Bitstream Vera Sans:style=Bold'

      -fontconfig (fontconfig only)
             Enables  the usage of fontconfig managed fonts (default: autode‐
             tect).
             NOTE: By default fontconfig is used for  libass-rendered  subti‐
             tles  and not used for OSD. With -fontconfig it is used for both
             libass and OSD, with -nofontconfig it is not used at  all,  i.e.
             only  then  -font  and  -subfont  will work with a given path to
             font.

      -forcedsubsonly
             Display only forced subtitles for the DVD  subtitle  stream  se‐
             lected by e.g. -slang.

      -fribidi-charset <charset name> (FriBiDi only)
             Specifies  the character set that will be passed to FriBiDi when
             decoding non-UTF-8 subtitles (default: ISO8859-8).

      -ifo <VOBsub IFO file>
             Indicate the file that will be used to load  palette  and  frame
             size for VOBsub subtitles.

      -noautosub
             Turns  off automatic subtitle file loading.  Note: VOBsub subti‐
             tles are not affected.

      -osd-duration <time>
             Set the duration of the OSD messages in ms (default: 1000).

      -osd-fractions <0-2>
             Set how fractions of seconds of the current timestamp are print‐
             ed on the OSD:
                0    Do not display fractions (default).
                1    Show the first two decimals.
                2    Show  approximated  frame  count  within current second.
                     This frame count is not accurate but only an  approxima‐
                     tion.   For  variable fps, the approximation is known to
                     be far off the correct frame count.

      -osdlevel <0-3> (MPlayer only)
             Specifies which mode the OSD should start in.
                0    subtitles only
                1    volume + seek (default)
                2    volume + seek + timer + percentage
                3    volume + seek + timer + percentage + total time

      -overlapsub
             Allows the next subtitle to be displayed while the  current  one
             is still visible (default is to enable the support only for spe‐
             cific formats).

      -progbar-align <0-100>
             Specify the vertical alignment of the progress bar (0: top, 100:
             bottom, default is 50, i.e. centered).

      -sid <ID> (also see -slang, -vobsubid)
             Display  the  subtitle stream specified by <ID> (0-31).  MPlayer
             prints the available subtitle IDs when run in verbose (-v) mode.
             If  you  cannot  select  one of the subtitles on a DVD, also try
             -vobsubid.

      -nosub Disables any otherwise auto-selected internal subtitles (as e.g.
             the  Matroska/mkv  demuxer supports).  Use -noautosub to disable
             the loading of external subtitle files.

      -slang <language code[,language code,...]> (also see -sid)
             Specify a priority list of subtitle languages to use.  Different
             container formats employ different language codes.  DVDs use ISO
             639-1 two letter language codes, Matroska uses ISO  639-2  three
             letter  language  codes  while  OGM uses a free-form identifier.
             MPlayer prints the available languages when run in verbose  (-v)
             mode.

             EXAMPLE:
                mplayer dvd://1 -slang hu,en
                     Chooses  the Hungarian subtitle track on a DVD and falls
                     back on English if Hungarian is not available.
                mplayer -slang jpn example.mkv
                     Plays a Matroska file with Japanese subtitles.

      -spuaa <mode>
             Antialiasing/scaling mode for DVD/VOBsub.  A value of 16 may  be
             added to <mode> in order to force scaling even when original and
             scaled frame size already match.  This can be employed  to  e.g.
             smooth subtitles with gaussian blur.  Available modes are:
                0    none (fastest, very ugly)
                1    approximate (broken?)
                2    full (slow)
                3    bilinear (default, fast and not too bad)
                4    uses swscaler gaussian blur (looks very good)

      -spualign <-1-2>
             Specify how SPU (DVD/VOBsub) subtitles should be aligned.
                -1   original position
                 0   Align at top (original behavior, default).
                 1   Align at center.
                 2   Align at bottom.

      -spugauss <0.0-3.0>
             Variance  parameter  of gaussian used by -spuaa 4.  Higher means
             more blur (default: 1.0).

      -sub <subtitlefile1,subtitlefile2,...>
             Use/display these subtitle files.  Only one  file  can  be  dis‐
             played at the same time.

      -sub-bg-alpha <0-255>
             Specify  the  alpha  channel  value  for subtitles and OSD back‐
             grounds.  Big values mean more transparency.  0 means completely
             transparent.

      -sub-bg-color <0-255>
             Specify the color value for subtitles and OSD backgrounds.  Cur‐
             rently subtitles are grayscale so this value  is  equivalent  to
             the intensity of the color.  255 means white and 0 black.

      -sub-demuxer <[+]name> (-subfile only) (BETA CODE)
             Force  subtitle demuxer type for -subfile.  Use a '+' before the
             name to force it, this will skip some checks!  Give the  demuxer
             name as printed by -sub-demuxer help.  For backward compatibili‐
             ty it also accepts the demuxer ID as defined in subreader.h.

      -sub-fuzziness <mode>
             Adjust matching fuzziness when searching for subtitles (does not
             apply to VOBsub):
                0    exact match (default)
                1    Load all subs containing movie name.
                2    Load all subs in the current and -sub-paths directories.

      -sub-no-text-pp
             Disables any kind of text post processing done after loading the
             subtitles.  Used for debug purposes.

      -subalign <0-2>
             Specify which edge of the subtitles should  be  aligned  at  the
             height given by -subpos.
                0    Align subtitle top edge (original behavior).
                1    Align subtitle center.
                2    Align subtitle bottom edge (default).

      -subcc <1-8>
             Display  DVD  Closed  Caption  (CC) subtitles from the specified
             channel.  Values 5 to 8 select a mode that can  extract  EIA-608
             compatibility  streams from EIA-708 data.  These are not the VOB
             subtitles, these are special ASCII subtitles for the hearing im‐
             paired encoded in the VOB userdata stream on most region 1 DVDs.
             CC subtitles have not been spotted on DVDs from other regions so
             far.

      -subcp <codepage> (iconv only)
             If  your  system  supports  iconv(3), you can use this option to
             specify the subtitle codepage. It takes priority over both -utf8
             and -unicode.

             EXAMPLE:
                -subcp latin2
                -subcp cp1250

      -subcp enca:<language>:<fallback codepage> (ENCA only)
             You  can  specify your language using a two letter language code
             to make ENCA detect the codepage automatically.  If unsure,  en‐
             ter  anything  and  watch  mplayer  -v output for available lan‐
             guages.  Use __ (two underscores) if your language is  not  sup‐
             ported.   Fallback  codepage specifies the codepage to use, when
             autodetection fails.

             EXAMPLE:
                -subcp enca:cs:latin2
                     Guess the encoding, assuming the  subtitles  are  Czech,
                     fall back on latin 2, if the detection fails.
                -subcp enca:pl:cp1250
                     Guess the encoding for Polish, fall back on cp1250.
                -subcp enca:__:latin1
                     Generic detection (mostly unicode) with latin1 fallback.

      -sub-paths <path1,path2,...>
             Specify extra subtitle paths to track in the media directory.

             EXAMPLE:  Assuming  that  /path/to/movie/movie.avi is played and
             -sub-paths sub,subtitles,/tmp/subs is specified, MPlayer search‐
             es for subtitle files in these directories:
                /path/to/movie/
                /path/to/movie/sub/
                /path/to/movie/subtitles/
                /tmp/subs/
                ~/.mplayer/sub/

      -subdelay <sec>
             Delays subtitles by <sec> seconds.  Can be negative.

      -subfile <filename> (BETA CODE)
             Currently useless.  Same as -audiofile, but for subtitle streams
             (OggDS?).

      -subfont <path to font (FreeType), font pattern (Fontconfig)> (FreeType
      only)
             Sets  the  subtitle  font (see -font).  If no -subfont is given,
             -font is used.

      -subfont-autoscale <0-3> (FreeType only)
             Sets the autoscale mode.
             NOTE: 0 means that text scale and OSD scale are font heights  in
             points.

             The mode can be:

                0    no autoscale
                1    proportional to movie height
                2    proportional to movie width
                3    proportional to movie diagonal (default)

      -subfont-blur <0-8> (FreeType only)
             Sets the font blur radius (default: 2).

      -subfont-encoding <value>
             Sets  the  font encoding.  When set to 'unicode', all the glyphs
             from the font file will be rendered and  unicode  will  be  used
             (default:  unicode).  (Without FreeType, setting any other value
             than  'unicode'  will  disable  unicode  glyphs  rendering   for
             font.desc  files.  With FreeType and for other values than 'uni‐
             code' your system has to support iconv(3) in order for  this  to
             work.)

      -subfont-osd-scale <0-100> (FreeType only)
             Sets the autoscale coefficient of the OSD elements (default: 6).

      -subfont-outline <0-8> (FreeType only)
             Sets the font outline thickness (default: 2).

      -subfont-text-scale <0-100> (FreeType only)
             Sets  the  subtitle  text autoscale coefficient as percentage of
             the screen size (default: 5).

      -subfps <rate>
             Specify the framerate of the subtitle file (default: movie fps).
             NOTE: <rate> > movie fps speeds the subtitles up for frame-based
             subtitle files and slows them down for time-based ones.

      -subpos <0-150> (useful with -vf expand)
             Specify  the  position of subtitles on the screen.  The value is
             the vertical position of the subtitle in % of the screen height.
             Values larger than 100 allow part of the subtitle to be cut off.

      -subwidth <10-100>
             Specify  the  maximum  width of subtitles on the screen.  Useful
             for TV-out.  The value is the width of the subtitle in % of  the
             screen width.

      -noterm-osd
             Disable the display of OSD messages on the console when no video
             output is available.

      -term-osd-esc <escape sequence>
             Specify the escape sequence to use before writing an OSD message
             on  the console.  The escape sequence should move the pointer to
             the beginning of the line used for the OSD  and  clear  it  (de‐
             fault: ^[[A\r^[[K).

      -unicode
             Tells  MPlayer  to handle the subtitle file as unicode. (It will
             only take effect if neither -subcp nor -utf8 is given.)

      -unrarexec <path to unrar executable> (not supported on MingW)
             Specify the path to the unrar executable so MPlayer can  use  it
             to  access rar-compressed VOBsub files (default: not set, so the
             feature is off).  The path must include the  executable's  file‐
             name, i.e. /usr/local/bin/unrar.

      -utf8
             Tells MPlayer to handle the subtitle file as UTF-8. (It will on‐
             ly take effect if -subcp isn't given, and it takes priority over
             -unicode.)

      -vobsub <VOBsub file without extension>
             Specify  a VOBsub file to use for subtitles.  Has to be the full
             pathname without extension, i.e. without the '.idx',  '.ifo'  or
             '.sub'.

      -vobsubid <0-31>
             Specify the VOBsub subtitle ID.

AUDIO OUTPUT OPTIONS (MPLAYER ONLY)
      -abs <value> (-ao oss only) (OBSOLETE)
             Override audio driver/card buffer size detection.

      -format <format> (also see the format audio filter)
             Select  the  sample format used for output from the audio filter
             layer to the sound card.  The values that <format> can adopt are
             listed below in the description of the format audio filter.

      -mixer <device>
             Use  a  mixer device different from the default /dev/mixer.  For
             ALSA this is the mixer name.

      -mixer-channel <mixer line>[,mixer index] (-ao oss and -ao alsa only)
             This option will tell MPlayer to use  a  different  channel  for
             controlling  volume  than  the default PCM.  Options for OSS in‐
             clude vol, pcm, line.  For a complete list of options  look  for
             SOUND_DEVICE_NAMES  in /usr/include/linux/soundcard.h.  For ALSA
             you can use the names  e.g.  alsamixer  displays,  like  Master,
             Line, PCM.
             NOTE:  ALSA  mixer  channel  names  followed by a number must be
             specified in the <name,number> format, i.e.  a  channel  labeled
             'PCM 1' in alsamixer must be converted to PCM,1.

      -softvol
             Force  the use of the software mixer, instead of using the sound
             card mixer.

      -softvol-max <10.0-10000.0>
             Set the maximum amplification level in percent  (default:  110).
             A value of 200 will allow you to adjust the volume up to a maxi‐
             mum of double the current level.  With values below 100 the ini‐
             tial  volume  (which  is  100%) will be above the maximum, which
             e.g. the OSD cannot display correctly.

      -volstep <0-100>
             Set the step size of mixer volume  changes  in  percent  of  the
             whole range (default: 3).

      -volume <-1-100> (also see -af volume)
             Set the startup volume in the mixer, either hardware or software
             (if used with -softvol).  A value of -1 (the default)  will  not
             change the volume.

AUDIO OUTPUT DRIVERS (MPLAYER ONLY)
      Audio  output  drivers are interfaces to different audio output facili‐
      ties.  The syntax is:

      -ao <driver1[:suboption1[=value]:...],driver2,...[,]>
             Specify a priority list of audio output drivers to be used.

      If the list has a trailing ',' MPlayer will fall back  on  drivers  not
      contained in the list.  Suboptions are optional and can mostly be omit‐
      ted.
      NOTE: See -ao help for a list of compiled-in audio output drivers.

      EXAMPLE:
                -ao alsa,oss,
                     Try the ALSA driver, then the OSS driver, then others.
                -ao alsa:noblock:device=hw=0.3
                     Sets noblock-mode and the  device-name  as  first  card,
                     fourth device.

      Available audio output drivers are:

      alsa
             ALSA 0.9/1.x audio output driver
                noblock
                     Sets noblock-mode.
                device=<device>
                     Sets  the device name.  Replace any ',' with '.' and any
                     ':' with '=' in the ALSA device name.  For hwac3  output
                     via  S/PDIF,  use  an "iec958" or "spdif" device, unless
                     you really know how to set it correctly.

      oss
             OSS audio output driver
                <dsp-device>
                     Sets the audio output device (default: /dev/dsp).
                <mixer-device>
                     Sets the audio mixer device (default: /dev/mixer).
                <mixer-channel>
                     Sets the audio mixer channel (default: pcm).

      sdl (SDL only)
             highly platform independent SDL (Simple Directmedia  Layer)  li‐
             brary audio output driver
                <driver>
                     Explicitly  choose the SDL audio driver to use (default:
                     let SDL choose).

      arts
             audio output through the aRts daemon

      esd
             audio output through the ESD daemon
                <server>
                     Explicitly choose the ESD server to use (default: local‐
                     host).

      jack
             audio output through JACK (Jack Audio Connection Kit)
                (no)connect
                     Automatically  create  connections  to output ports (de‐
                     fault: enabled).  When enabled, the  maximum  number  of
                     output  channels will be limited to the number of avail‐
                     able output ports.
                port=<name>
                     Connects to the ports  with  the  given  name  (default:
                     physical ports).
                name=<client
                     Client  name  that  is  passed to JACK (default: MPlayer
                     [<PID>]).  Useful if you want to  have  certain  connec‐
                     tions established automatically.
                (no)estimate
                     Estimate  the  audio  delay,  supposed to make the video
                     playback smoother (default: enabled).
                (no)autostart
                     Automatically start jackd if  necessary  (default:  dis‐
                     abled).   Note  that this seems unreliable and will spam
                     stdout with server messages.

      nas
             audio output through NAS

      coreaudio (Mac OS X only)
             native Mac OS X audio output driver
                device_id=<id>
                     ID of output device to use (0 = default device)
                help List all available output devices with their IDs.

      openal
             Experimental OpenAL audio output driver

      pulse
             PulseAudio audio output driver
                [<host>[:<output sink>[:broken_pause]]]
                     Specify the host and optionally output sink to use.   An
                     empty <host> string uses a local connection, "localhost"
                     uses network transfer (most likely not what  you  want).
                     You  can also explicitly force the workaround for broken
                     pause functionality (default:  autodetected).   To  only
                     enable that without specifying a host/sink the syntax is
                     -ao pulse:::broken_pause

      sgi (SGI only)
             native SGI audio output driver
                <output device name>
                     Explicitly choose the  output  device/interface  to  use
                     (default:  system-wide  default).   For example, 'Analog
                     Out' or 'Digital Out'.

      sun (Sun only)
             native Sun audio output driver
                <device>
                     Explicitly choose the  audio  device  to  use  (default:
                     /dev/audio).

      win32 (Windows only)
             native Windows waveout audio output driver

      dsound (Windows only)
             DirectX DirectSound audio output driver
                device=<devicenum>
                     Sets  the  device number to use.  Playing a file with -v
                     will show a list of available devices.

      kai (OS/2 only)
             OS/2 KAI audio output driver
                uniaud
                     Force UNIAUD mode.
                dart Force DART mode.
                (no)share
                     Open audio in shareable or exclusive mode.
                bufsize=<size>
                     Set buffer size to <size> in samples (default: 2048).

      dart (OS/2 only)
             OS/2 DART audio output driver
                (no)share
                     Open DART in shareable or exclusive mode.
                bufsize=<size>
                     Set buffer size to <size> in samples (default: 2048).

      dxr2 (also see -dxr2) (DXR2 only)
             Creative DXR2 specific output driver

      v4l2 (requires Linux 2.6.22+ kernel)
             Audio output driver for V4L2 cards with hardware MPEG decoder.

      mpegpes (DVB only)
             Audio output driver for DVB cards that writes the output  to  an
             MPEG-PES file if no DVB card is installed.
                card=<1-4>
                     DVB  card  to  use if more than one card is present.  If
                     not specified MPlayer will search the first usable card.
                file=<filename>
                     output filename

      null
             Produces no audio output but  maintains  video  playback  speed.
             Use -nosound for benchmarking.

      pcm
             raw PCM/wave file writer audio output
                (no)waveheader
                     Include  or do not include the wave header (default: in‐
                     cluded).  When not included, raw PCM will be generated.
                file=<filename>
                     Write the sound to <filename> instead of the default au‐
                     diodump.wav.   If nowaveheader is specified, the default
                     is audiodump.pcm.
                fast
                     Try to dump faster than realtime.  Make sure the  output
                     does  not  get  truncated  (usually with "Too many video
                     packets in buffer" message).  It is normal that you  get
                     a "Your system is too SLOW to play this!" message.

      plugin
             plugin audio output driver

VIDEO OUTPUT OPTIONS (MPLAYER ONLY)
      -adapter <value>
             Set  the graphics card that will receive the image.  You can get
             a list of available cards when you  run  this  option  with  -v.
             Currently only works with the directx video output driver.

      -bpp <depth>
             Override  the  autodetected  color depth.  Only supported by the
             fbdev, dga, svga, vesa video output drivers.

      -border
             Play movie with window border and decorations.  Since this is on
             by default, use -noborder to disable the standard window decora‐
             tions.

      -brightness <-100-100>
             Adjust the brightness of the video  signal  (default:  0).   Not
             supported by all video output drivers.

      -contrast <-100-100>
             Adjust  the contrast of the video signal (default: 0).  Not sup‐
             ported by all video output drivers.

      -display <name> (X11 only)
             Specify the hostname and display number of the X server you want
             to display on.

             EXAMPLE:
                -display xtest.localdomain:0

      -dr
             Turns on direct rendering (not supported by all codecs and video
             outputs).  This can result in significantly faster  blitting  on
             some  systems,  on most the difference will be minimal.  In some
             cases, particularly with decoders specifying  their  buffer  re‐
             quirements badly, it can be vastly slower.
             WARNING: May cause OSD/SUB corruption!

      -dxr2 <option1:option2:...>
             This option is used to control the dxr2 video output driver.

                ar-mode=<value>
                     aspect  ratio  mode  (0  = normal, 1 = pan-and-scan, 2 =
                     letterbox (default))

                iec958-encoded
                     Set iec958 output mode to encoded.

                iec958-decoded
                     Set iec958 output mode to decoded (default).

                macrovision=<value>
                     macrovision mode (0 = off (default), 1 = agc, 2 = agc  2
                     colorstripe, 3 = agc 4 colorstripe)

                mute
                     mute sound output

                unmute
                     unmute sound output

                ucode=<value>
                     path to the microcode

             TV output

                75ire
                     enable 7.5 IRE output mode

                no75ire
                     disable 7.5 IRE output mode (default)

                bw
                     b/w TV output

                color
                     color TV output (default)

                interlaced
                     interlaced TV output (default)

                nointerlaced
                     disable interlaced TV output

                norm=<value>
                     TV norm (ntsc (default), pal, pal60, palm, paln, palnc)

                square-pixel
                     set pixel mode to square

                ccir601-pixel
                     set pixel mode to ccir601

             overlay

                cr-left=<0-500>
                     Set the left cropping value (default: 50).

                cr-right=<0-500>
                     Set the right cropping value (default: 300).

                cr-top=<0-500>
                     Set the top cropping value (default: 0).

                cr-bottom=<0-500>
                     Set the bottom cropping value (default: 0).

                ck-[r|g|b]=<0-255>
                     Set  the  r(ed),  g(reen)  or b(lue) gain of the overlay
                     color-key.

                ck-[r|g|b]min=<0-255>
                     minimum value for the respective color key

                ck-[r|g|b]max=<0-255>
                     maximum value for the respective color key

                ignore-cache
                     Ignore cached overlay settings.

                update-cache
                     Update cached overlay settings.

                ol-osd
                     Enable overlay onscreen display.

                nool-osd
                     Disable overlay onscreen display (default).

                ol[h|w|x|y]-cor=<-20-20>
                     Adjust the overlay size (h,w) and position (x,y) in case
                     it does not match the window perfectly (default: 0).

                overlay
                     Activate overlay (default).

                nooverlay
                     Activate TV-out.

                overlay-ratio=<1-2500>
                     Tune the overlay (default: 1000).

      -fbmode <modename> (-vo fbdev only)
             Change  video  mode  to the one that is labeled as <modename> in
             /etc/fb.modes.
             NOTE: VESA framebuffer does not support mode changing.

      -fbmodeconfig <filename> (-vo fbdev only)
             Override framebuffer mode  configuration  file  (default:  /etc/
             fb.modes).

      -fs (also see -zoom)
             Fullscreen  playback  (centers  movie,  and  paints  black bands
             around it).  Not supported by all video output drivers.

      -fsmode-dontuse <0-31> (OBSOLETE, use the -fs option)
             Try this option if you still experience fullscreen problems.

      -fstype <type1,type2,...> (X11 only)
             Specify a priority list of fullscreen modes to be used.  You can
             negate  the modes by prefixing them with '-'.  If you experience
             problems like the fullscreen window being covered by other  win‐
             dows try using a different order.
             NOTE: See -fstype help for a full list of available modes.

             The available types are:

                above
                     Use the _NETWM_STATE_ABOVE hint if available.
                below
                     Use the _NETWM_STATE_BELOW hint if available.
                fullscreen
                     Use the _NETWM_STATE_FULLSCREEN hint if available.
                layer
                     Use the _WIN_LAYER hint with the default layer.
                layer=<0...15>
                     Use the _WIN_LAYER hint with the given layer number.
                netwm
                     Force NETWM style.
                none
                     Clear the list of modes; you can add modes to enable af‐
                     terward.
                stays_on_top
                     Use _NETWM_STATE_STAYS_ON_TOP hint if available.

             EXAMPLE:
                layer,stays_on_top,above,fullscreen
                     Default order, will be used as a fallback  if  incorrect
                     or unsupported modes are specified.
                -fullscreen
                     Fixes fullscreen switching on OpenBox 1.x.

      -fs-border-left <pixels>

      -fs-border-right <pixels>

      -fs-border-top <pixels>

      -fs-border-bottom <pixels>
             Specify extra borders in full screen mode.  The borders apply to
             all displayed elements: video, OSD and EOSD.  The number of pix‐
             els  is specified in terms of screen resolution.  Currently only
             supported with by the gl video output driver.

      -gamma <-100-100>
             Adjust the gamma of the video signal (default: 0).  Not support‐
             ed by all video output drivers.

      -geometry x[%][:y[%]] or [WxH][+-x+-y]
             Adjust where the output is on the screen initially.  The x and y
             specifications are in pixels measured from the top-left  of  the
             screen  to the top-left of the image being displayed, however if
             a percentage sign is given after the argument it turns the value
             into a percentage of the screen size in that direction.  It also
             supports the standard X11 -geometry option format, in which e.g.
             +10-50 means "place 10 pixels from the left border and 50 pixels
             from the lower border" and "--20+-10" means "place 20 pixels be‐
             yond  the right and 10 pixels beyond the top border".  If an ex‐
             ternal window is specified using the -wid option, then the x and
             y  coordinates are relative to the top-left corner of the window
             rather than the screen.  The coordinates  are  relative  to  the
             screen given with -screen for the video output drivers that ful‐
             ly support -screen (direct3d,  gl,  gl_tiled,  vdpau,  x11,  xv,
             xvmc, corevideo).
             NOTE:  This option is only supported by the x11, xmga, xv, xvmc,
             xvidix, gl, gl_tiled, direct3d, directx, fbdev, sdl,  dfxfb  and
             corevideo video output drivers.

             EXAMPLE:
                50:40
                     Places the window at x=50, y=40.
                50%:50%
                     Places the window in the middle of the screen.
                100%
                     Places the window at the middle of the right edge of the
                     screen.
                100%:100%
                     Places the window at the  bottom  right  corner  of  the
                     screen.

      -gui-wid <window ID> (also see -wid) (GUI only)
             This tells the GUI to also use an X11 window and stick itself to
             the bottom of the video, which is useful to embed a mini-GUI  in
             a browser (with the MPlayer plugin for instance).

      -hue <-100-100>
             Adjust  the hue of the video signal (default: 0).  You can get a
             colored negative of the image with this option.   Not  supported
             by all video output drivers.

      -monitor-dotclock <range[,range,...]> (-vo fbdev and vesa only)
             Specify the dotclock or pixelclock range of the monitor.

      -monitor-hfreq <range[,range,...]> (-vo fbdev and vesa only)
             Specify the horizontal frequency range of the monitor.

      -monitor-vfreq <range[,range,...]> (-vo fbdev and vesa only)
             Specify the vertical frequency range of the monitor.

      -monitoraspect <ratio> (also see -aspect)
             Set the aspect ratio of your monitor or TV screen.  A value of 0
             disables a previous setting (e.g. in the  config  file).   Over‐
             rides the -monitorpixelaspect setting if enabled.

             EXAMPLE:
                -monitoraspect 4:3  or 1.3333
                -monitoraspect 16:9 or 1.7777

      -monitorpixelaspect <ratio> (also see -aspect)
             Set  the  aspect  of a single pixel of your monitor or TV screen
             (default: 1).  A value of 1 means  square  pixels  (correct  for
             (almost?) all LCDs).

      -name (X11 only)
             Set the window class name.

      -nodouble
             Disables  double buffering, mostly for debugging purposes.  Dou‐
             ble buffering fixes flicker by storing two frames in memory, and
             displaying  one while decoding another.  It can affect OSD nega‐
             tively, but often removes OSD flickering.

      -nograbpointer
             Do not grab the mouse pointer after a video mode  change  (-vm).
             Useful for multihead setups.

      -nokeepaspect
             Do  not  keep  window  aspect ratio when resizing windows.  Only
             works with the x11, xv, xmga, xvidix, directx video output driv‐
             ers.   Furthermore  under  X11  your window manager has to honor
             window aspect hints.

      -ontop
             Makes the player window stay on top of other windows.  Supported
             by  video  output  drivers which use X11, except SDL, as well as
             directx, corevideo, quartz, ggi and gl_tiled.

      -panscan <0.0-1.0>
             Enables pan-and-scan functionality (cropping the sides of e.g. a
             16:9  movie  to  make it fit a 4:3 display without black bands).
             The range controls how much of the image is cropped.  Only works
             with the directx, xv, xmga, mga, gl, gl_tiled, quartz, corevideo
             and xvidix video output drivers.
             NOTE: Values between -1 and 0 are allowed as  well,  but  highly
             experimental and may crash or worse.  Use at your own risk!

      -panscanrange <-19.0-99.0> (experimental)
             Change the range of the pan-and-scan functionality (default: 1).
             Positive values mean multiples of the default  range.   Negative
             numbers  mean you can zoom in up to a factor of -panscanrange+1.
             E.g. -panscanrange -3 allows a zoom factor of  up  to  4.   This
             feature is experimental.

      -border-pos-x  <0.0-1.0>  (-vo  gl,xv,xvmc,vdpau,direct3d only, default
      0.5)
             When black borders are added to adjust for aspect,  this  deter‐
             mines  where  they are placed.  0.0 places borders on the right,
             1.0 on the left.  Values outside the range 0.0 -  1.0  will  add
             extra  black borders on one side and remove part of the image on
             the other side.

      -border-pos-y <0.0-1.0> (-vo  gl,xv,xvmc,vdpau,direct3d  only,  default
      0.5)
             As -border-pos-x but for top/bottom borders.  0.0 places borders
             on the bottom, 1.0 on the top.

      -monitor-orientation <0-3> (experimental)
             Rotate display by 90, 180 or 270 degrees.  Rotates also the OSD,
             not  just  the  video image itself.  Currently only supported by
             the gl video output driver.  For all  other  video  outputs  -vf
             ass,expand=osd=1,rotate=n  can be used, in the future this might
             even happen automatically.

      -refreshrate <Hz>
             Set the monitor refreshrate in Hz.  Currently only supported  by
             -vo directx combined with the -vm option.

      -rootwin
             Play  movie  in  the  root window (desktop background).  Desktop
             background images may cover  the  movie  window,  though.   Only
             works  with the x11, xv, xmga, xvidix, quartz, corevideo and di‐
             rectx video output drivers.

      -saturation <-100-100>
             Adjust the saturation of the video signal (default: 0).  You can
             get  grayscale  output  with  this option.  Not supported by all
             video output drivers.

      -screenh <pixels>
             Specify the screen height for video output drivers which do  not
             know the screen resolution like fbdev, x11 and TV-out.

      -screenw <pixels>
             Specify  the  screen width for video output drivers which do not
             know the screen resolution like fbdev, x11 and TV-out.

      -(no)stop-xscreensaver (X11 only)
             Turns off xscreensaver at startup and turns it on again on  exit
             (default:  enabled).   If  your screensaver supports neither the
             XSS nor XResetScreenSaver API please use -heartbeat-cmd instead.

      -title (also see -use-filename-title)
             Set the window title.  Supported by X11-based video output driv‐
             ers.

      -use-filename-title (also see -title)
             Set the window title using the media filename, when not set with
             -title.  Supported by X11-based video output drivers.

      -vm
             Try to change to a different video mode.  Supported by the  dga,
             x11, xv, sdl and directx video output drivers.  If used with the
             directx video output driver the  -screenw,  -screenh,  -bpp  and
             -refreshrate options can be used to set the new display mode.

      -vsync
             Enables VBI for the vesa, dfbmga and svga video output drivers.

      -wid <window ID> (also see -gui-wid) (X11, OpenGL and DirectX only)
             This  tells  MPlayer to attach to an existing window.  Useful to
             embed MPlayer in a browser (e.g. the plugger  extension).   This
             option  fills  the given window completely, thus aspect scaling,
             panscan, etc are no longer handled by MPlayer but must  be  man‐
             aged by the application that created the window.

      -screen <-2-...> (alias for -xineramascreen)
             In  Xinerama  configurations  (i.e.  a single desktop that spans
             across multiple displays) this option tells MPlayer which screen
             to  display the movie on.  A value of -2 means fullscreen across
             the whole virtual display (in this case Xinerama information  is
             completely ignored), -1 means fullscreen on the display the win‐
             dow currently is on.  The initial position set via the -geometry
             option  is  relative to the specified screen.  Will usually only
             work with "-fstype -fullscreen" or "-fstype none".  This  option
             is  not suitable to only set the startup screen (because it will
             always display on the given screen in fullscreen mode),  -geome‐
             try  is  the  best that is available for that purpose currently.
             Supported by at least the direct3d, gl, gl_tiled,  x11,  xv  and
             corevideo video output drivers.

      -zrbw (-vo zr only)
             Display  in  black and white.  For optimal performance, this can
             be combined with '-lavdopts gray'.

      -zrcrop <[width]x[height]+[x offset]+[y offset]> (-vo zr only)
             Select a part of the input image  to  display,  multiple  occur‐
             rences of this option switch on cinerama mode.  In cinerama mode
             the movie is distributed over more than one TV  (or  beamer)  to
             create a larger image.  Options appearing after the n-th -zrcrop
             apply to the n-th MJPEG card, each card should at least  have  a
             -zrdev in addition to the -zrcrop.  For examples, see the output
             of -zrhelp and the Zr section of the documentation.

      -zrdev <device> (-vo zr only)
             Specify the device special file that belongs to your MJPEG card,
             by default the zr video output driver takes the first v4l device
             it can find.

      -zrfd (-vo zr only)
             Force  decimation:  Decimation,  as  specified  by  -zrhdec  and
             -zrvdec, only happens if the hardware scaler can stretch the im‐
             age to its original size.  Use this option to force decimation.

      -zrhdec <1|2|4> (-vo zr only)
             Horizontal decimation: Ask the driver to send only every 2nd  or
             4th  line/pixel of the input image to the MJPEG card and use the
             scaler of the MJPEG card to stretch the image  to  its  original
             size.

      -zrhelp (-vo zr only)
             Display  a  list of all -zr* options, their default values and a
             cinerama mode example.

      -zrnorm <norm> (-vo zr only)
             Specify the TV norm as PAL or NTSC (default: no change).

      -zrquality <1-20> (-vo zr only)
             A number from 1 (best) to 20 (worst) representing the  JPEG  en‐
             coding quality.

      -zrvdec <1|2|4> (-vo zr only)
             Vertical  decimation:  Ask  the driver to send only every 2nd or
             4th line/pixel of the input image to the MJPEG card and use  the
             scaler  of  the  MJPEG card to stretch the image to its original
             size.

      -zrxdoff <x display offset> (-vo zr only)
             If the movie is smaller than the TV screen, this  option  speci‐
             fies  the  x  offset from the upper-left corner of the TV screen
             (default: centered).

      -zrydoff <y display offset> (-vo zr only)
             If the movie is smaller than the TV screen, this  option  speci‐
             fies  the  y  offset from the upper-left corner of the TV screen
             (default: centered).

VIDEO OUTPUT DRIVERS (MPLAYER ONLY)
      Video output drivers are interfaces to different video  output  facili‐
      ties.  The syntax is:

      -vo <driver1[:suboption1[=value]:...],driver2,...[,]>
             Specify a priority list of video output drivers to be used.

      If  the  list  has a trailing ',' MPlayer will fall back on drivers not
      contained in the list.  Suboptions are optional and can mostly be omit‐
      ted.
      NOTE: See -vo help for a list of compiled-in video output drivers.

      EXAMPLE:
                -vo xmga,xv,
                     Try the Matrox X11 driver, then the Xv driver, then oth‐
                     ers.
                -vo directx:noaccel
                     Uses  the  DirectX  driver  with  acceleration  features
                     turned off.

      Available video output drivers are:

      xv (X11 only)
             Uses  the XVideo extension of XFree86 4.x to enable hardware ac‐
             celerated playback.  If you cannot use a hardware specific driv‐
             er,  this  is  probably  the best option.  For information about
             what colorkey is used and how it is drawn run  MPlayer  with  -v
             option and look out for the lines tagged with [xv common] at the
             beginning.
                adaptor=<number>
                     Select a specific XVideo adaptor (check xvinfo results).
                port=<number>
                     Select a specific XVideo port.
                ck=<cur|use|set>
                     Select the source from which the colorkey is taken  (de‐
                     fault: cur).
                        cur  The  default takes the colorkey currently set in
                             Xv.
                        use  Use but do not set  the  colorkey  from  MPlayer
                             (use -colorkey option to change it).
                        set  Same as use but also sets the supplied colorkey.
                ck-method=<man|bg|auto>
                     Sets the colorkey drawing method (default: man).
                        man  Draw  the  colorkey manually (reduces flicker in
                             some cases).
                        bg   Set the colorkey as window background.
                        auto Let Xv draw the colorkey.

      x11 (X11 only)
             Shared memory video output driver without hardware  acceleration
             that works whenever X11 is present.

      xover (X11 only)
             Adds  X11  support  to  all  overlay based video output drivers.
             Currently only supported by tdfx_vid.
                <vo_driver>
                     Select the driver to use as source to overlay on top  of
                     X11.

      vdpau  (with  -vc ffmpeg2vdpau, ffwmv3vdpau, ffvc1vdpau, ffh264vdpau or
      ffodivxvdpau)
             Video output that uses VDPAU to decode video via hardware.  Also
             supports displaying of software-decoded video.
                sharpen=<-1-1>
                     For positive values, apply a sharpening algorithm to the
                     video, for negative values  a  blurring  algorithm  (de‐
                     fault: 0).
                denoise=<0-1>
                     Apply a noise reduction algorithm to the video (default:
                     0, no noise reduction).
                deint=<0-4>
                     Select the deinterlacer (default: 0).  All modes > 0 re‐
                     spect -field-dominance.
                        0    no deinterlacing
                        1    Show only first field, similar to -vf field.
                        2    Bob deinterlacing, similar to -vf tfields=1.
                        3    motion  adaptive temporal deinterlacing May lead
                             to A/V desync with slow  video  hardware  and/or
                             high  resolution.  This is the default if "D" is
                             used to enable deinterlacing.
                        4    motion  adaptive  temporal  deinterlacing   with
                             edge-guided  spatial  interpolation  Needs  fast
                             video hardware.
                chroma-deint
                     Makes temporal deinterlacers operate both  on  luma  and
                     chroma (default).  Use nochroma-deint to solely use luma
                     and speed up advanced deinterlacing.  Useful  with  slow
                     video memory.
                pullup
                     Try to skip deinterlacing for progressive frames, useful
                     for watching telecined content, needs fast  video  hard‐
                     ware for high resolutions.  Only works with motion adap‐
                     tive temporal deinterlacing.
                colorspace
                     Select the color space for YUV to  RGB  conversion.   In
                     general  BT.601  should  be used for standard definition
                     (SD) content and BT.709 for high  definition  (HD)  con‐
                     tent.   Using  incorrect color space results in slightly
                     under or over saturated and shifted colors.
                        0    Guess the color space based on video resolution.
                             Video  with width >= 1280 or height > 576 is as‐
                             sumed to be HD and BT.709 color  space  will  be
                             used.
                        1    Use ITU-R BT.601 color space (default).
                        2    Use ITU-R BT.709 color space.
                        3    Use SMPTE-240M color space.
                hqscaling
                        0    Use default VDPAU scaling (default).
                        1-9  Apply  high quality VDPAU scaling (needs capable
                             hardware).
                force-mixer
                     Forces the use of the VDPAU mixer, which implements  all
                     above  options  (default).   Use  noforce-mixer to allow
                     displaying BGRA colorspace.  (Disables all above options
                     and the hardware equalizer if image format BGRA is actu‐
                     ally used.)

      xvmc (X11 with FFmpeg MPEG-1/2 decoder only)
             Video output driver that uses the XvMC (X Video Motion Compensa‐
             tion) extension of XFree86 4.x to speed up MPEG-1/2 and VCR2 de‐
             coding.
                adaptor=<number>
                     Select a specific XVideo adaptor (check xvinfo results).
                port=<number>
                     Select a specific XVideo port.
                (no)benchmark
                     Disables image display.  Necessary for proper benchmark‐
                     ing  of drivers that change image buffers on monitor re‐
                     trace only (nVidia).  Default is not  to  disable  image
                     display (nobenchmark).
                (no)bobdeint
                     Very  simple  deinterlacer.   Might not look better than
                     -vf tfields=1, but it is the only deinterlacer for  xvmc
                     (default: nobobdeint).
                (no)queue
                     Queue  frames for display to allow more parallel work of
                     the video hardware.  May add a  small  (not  noticeable)
                     constant A/V desync (default: noqueue).
                (no)sleep
                     Use sleep function while waiting for rendering to finish
                     (not recommended on Linux) (default: nosleep).
                ck=cur|use|set
                     Same as -vo xv:ck (see -vo xv).
                ck-method=man|bg|auto
                     Same as -vo xv:ck-method (see -vo xv).

      dga (X11 only)
             Play video through the XFree86 Direct Graphics Access extension.
             Considered obsolete.

      sdl (SDL only, buggy/outdated)
             Highly  platform  independent SDL (Simple Directmedia Layer) li‐
             brary video output driver.  Since SDL uses its  own  X11  layer,
             MPlayer X11 options do not have any effect on SDL.  Note that it
             has several minor bugs (-vm/-novm is mostly ignored, -fs behaves
             like  -novm  should, window is in top-left corner when returning
             from fullscreen, panscan is not supported, ...).
                driver=<driver>
                     Explicitly choose the SDL driver to use.
                (no)forcexv
                     Use XVideo through the sdl video output driver (default:
                     forcexv).
                (no)hwaccel
                     Use hardware accelerated scaler (default: hwaccel).

      vidix
             VIDIX  (VIDeo  Interface  for *niX) is an interface to the video
             acceleration features of different graphics  cards.   Very  fast
             video output driver on cards that support it.
                <subdevice>
                     Explicitly  choose  the  VIDIX  subdevice driver to use.
                     Available  subdevice  drivers  are   cyberblade,   ivtv,
                     mach64,   mga_crtc2,  mga,  nvidia,  pm2,  pm3,  radeon,
                     rage128, s3, sh_veu, sis_vid and unichrome.

      xvidix (X11 only)
             X11 frontend for VIDIX
                <subdevice>
                     same as vidix

      cvidix
             Generic and platform independent VIDIX frontend, can even run in
             a text console with nVidia cards.
                <subdevice>
                     same as vidix

      winvidix (Windows only)
             Windows frontend for VIDIX
                <subdevice>
                     same as vidix

      direct3d (Windows only) (BETA CODE!)
             Video output driver that uses the Direct3D interface (useful for
             Vista).

      directx (Windows only)
             Video output driver that uses the DirectX interface.
                noaccel
                     Turns off hardware acceleration.  Try this option if you
                     have display problems.

      kva (OS/2 only)
             Video output driver that uses the libkva interface.
                snap Force SNAP mode.
                wo   Force WarpOverlay! mode.
                dive Force DIVE mode.
                (no)t23
                     Enable  or  disable  workaround for T23 laptop (default:
                     disabled).  Try to enable this option if your video card
                     supports upscaling only.

      quartz (Mac OS X only)
             Mac  OS X Quartz video output driver.  Under some circumstances,
             it might be more efficient to force a packed YUV output  format,
             with e.g. -vf format=yuy2.
                device_id=<number>
                     Choose the display device to use in fullscreen.
                fs_res=<width>:<height>
                     Specify  the  fullscreen resolution (useful on slow sys‐
                     tems).

      corevideo (Mac OS X 10.4 or 10.3.9 with QuickTime 7)
             Mac OS X CoreVideo video output driver
                device_id=<number>
                     DEPRECATED, use -screen instead.  Choose the display de‐
                     vice to use for fullscreen or set it to -1 to always use
                     the same screen the video window is on  (default:  -1  -
                     auto).
                shared_buffer
                     Write  output  to a shared memory buffer instead of dis‐
                     playing it and try to open an existing NSConnection  for
                     communication with a GUI.
                buffer_name=<name>
                     Name  of the shared buffer created with shm_open as well
                     as the name of the NSConnection MPlayer will try to open
                     (default: "mplayerosx").  Setting buffer_name implicitly
                     enables shared_buffer.

      fbdev (Linux only)
             Uses the kernel framebuffer to play video.
                <device>
                     Explicitly choose the fbdev device  name  to  use  (e.g.
                     /dev/fb0)  or the name of the VIDIX subdevice if the de‐
                     vice name starts with 'vidix' (e.g.  'vidixsis_vid'  for
                     the sis driver).

      fbdev2 (Linux only)
             Uses the kernel framebuffer to play video, alternative implemen‐
             tation.
                <device>
                     Explicitly choose the fbdev device name to use (default:
                     /dev/fb0).

      vesa
             Very  general  video  output driver that should work on any VESA
             VBE 2.0 compatible card.
                (no)dga
                     Turns DGA mode on or off (default: on).
                neotv_pal
                     Activate the NeoMagic TV out and set it to PAL norm.
                neotv_ntsc
                     Activate the NeoMagic TV out and set it to NTSC norm.
                vidix
                     Use the VIDIX driver.
                lvo:
                     Activate the Linux Video Overlay on top of VESA mode.

      svga
             Play video using the SVGA library.
                <video mode>
                     Specify video mode to use.  The mode can be given  in  a
                     <width>x<height>x<colors> format, e.g. 640x480x16M or be
                     a graphics mode number, e.g. 84.
                bbosd
                     Draw OSD into black bands below the movie (slower).
                native
                     Use only native drawing functions.  This  avoids  direct
                     rendering, OSD and hardware acceleration.
                retrace
                     Force  frame  switch  on  vertical retrace.  Usable only
                     with -double.  It has the same effect as the -vsync  op‐
                     tion.
                sq
                     Try to select a video mode with square pixels.
                vidix
                     Use svga with VIDIX.

      gl
             OpenGL  video output driver, simple version.  Video size must be
             smaller than the maximum texture size of your OpenGL implementa‐
             tion.   Intended  to work even with the most basic OpenGL imple‐
             mentations, but also makes use of newer extensions, which  allow
             support  for more colorspaces and direct rendering.  For optimal
             speed try adding the options
             -dr -noslices
             The code performs very few checks, so  if  a  feature  does  not
             work,  this  might  be  because  it  is  not  supported  by your
             card/OpenGL implementation even if you do not get any error mes‐
             sage.   Use  glxinfo  or a similar tool to display the supported
             OpenGL extensions.
                backend=<n>
                     Select the backend/OpenGL  implementation  to  use  (de‐
                     fault: -1).
                        -1: Autoselect
                        0: Win32/WGL
                        1: X11/GLX
                        2: SDL
                        3: X11/EGL (highly experimental)
                        4: OSX/Cocoa
                        5: Android (very bad hack, only for testing)
                (no)ati-hack
                     ATI  drivers  may  give  a corrupted image when PBOs are
                     used (when using -dr or force-pbo).  This  option  fixes
                     this, at the expense of using a bit more memory.
                (no)force-pbo
                     Always  uses  PBOs to transfer textures even if this in‐
                     volves an extra copy.  Currently this gives a little ex‐
                     tra  speed with NVidia drivers and a lot more speed with
                     ATI drivers.  May need -noslices and the ati-hack subop‐
                     tion to work correctly.
                (no)scaled-osd
                     Changes  the  way  the  OSD behaves when the size of the
                     window changes (default: disabled).   When  enabled  be‐
                     haves more like the other video output drivers, which is
                     better for fixed-size fonts.  Disabled looks much better
                     with  FreeType  fonts and uses the borders in fullscreen
                     mode.  Does not work correctly with ass  subtitles  (see
                     -ass),  you  can instead render them without OpenGL sup‐
                     port via -vf ass.
                osdcolor=<0xAARRGGBB>
                     Color for OSD (default: 0x00ffffff, corresponds to  non-
                     transparent white).
                rectangle=<0,1,2>
                     Select  usage  of rectangular textures which saves video
                     RAM, but often is slower (default: 0).
                        0: Use power-of-two textures (default).
                        1: Use the GL_ARB_texture_rectangle extension.
                        2: Use the GL_ARB_texture_non_power_of_two extension.
                        In  some  cases  only  supported in software and thus
                        very slow.
                swapinterval=<n>
                     Minimum interval between two buffer  swaps,  counted  in
                     displayed  frames  (default: 1).  1 is equivalent to en‐
                     abling VSYNC, 0 to disabling VSYNC.  Values below 0 will
                     leave it at the system default.  This limits the framer‐
                     ate  to  (horizontal  refresh  rate  /   n).    Requires
                     GLX_SGI_swap_control   support   to   work.   With  some
                     (most/all?)   implementations   this   only   works   in
                     fullscreen mode.
                ycbcr
                     Use  the  GL_APPLE_ycbcr_422 extension to convert YUV to
                     RGB.  Default is disabled if yuv= is specified, auto-de‐
                     tected otherwise.  Note that this will enable a few spe‐
                     cial settings to get into a special driver fast-path.
                yuv=<n>
                     Select the type of YUV to RGB conversion.   The  default
                     is auto-detection deciding between values 0 and 2.
                        0:  Use  software  conversion.   Compatible  with all
                        OpenGL versions.  Provides brightness,  contrast  and
                        saturation control.
                        1:  Use register combiners.  This uses an nVidia-spe‐
                        cific extension (GL_NV_register_combiners).  At least
                        three  texture units are needed.  Provides saturation
                        and hue control.  This method is fast but inexact.
                        2: Use a fragment program using the POW  instruction.
                        Needs  the  GL_ARB_fragment_program  extension and at
                        least three texture units.  Provides brightness, con‐
                        trast,  saturation, hue and gamma control.  Gamma can
                        also be set independently for red,  green  and  blue.
                        Method 4 is usually faster.
                        3:  Same  as  2.   They  exist as distinct values for
                        legacy reasons, MPlayer now  inserts  the  extra  in‐
                        structions for gamma control on-demand.
                        4:  Use  a  fragment  program with additional lookup.
                        Needs the GL_ARB_fragment_program  extension  and  at
                        least  four texture units.  Provides brightness, con‐
                        trast, saturation, hue and gamma control.  Gamma  can
                        also be set independently for red, green and blue.
                        5:  Use  ATI-specific method (for older cards).  This
                        uses an ATI-specific extension (GL_ATI_fragment_shad‐
                        er  -  not  GL_ARB_fragment_shader!).  At least three
                        texture units are needed.   Provides  saturation  and
                        hue control.  This method is fast but inexact.
                        6:  Use  a  3D  texture  to do conversion via lookup.
                        Needs the GL_ARB_fragment_program  extension  and  at
                        least  four  texture units.  Extremely slow (software
                        emulation) on some (all?) ATI cards since it  uses  a
                        texture  with  border  pixels.   Provides brightness,
                        contrast, saturation, hue and gamma  control.   Gamma
                        can  also  be  set  independently  for red, green and
                        blue.  Speed depends more  on  GPU  memory  bandwidth
                        than other methods.
                colorspace
                     Select the color space for YUV to RGB conversion.
                        0    Use  the  formula  used normally by MPlayer (de‐
                             fault).
                        1    Use ITU-R BT.601 color space.
                        2    Use ITU-R BT.709 color space.
                        3    Use SMPTE-240M color space.
                levelconv=<n>
                     Select the brightness level conversion to  use  for  the
                     YUV to RGB conversion
                        0    Convert TV to PC levels (default).
                        1    Convert PC to TV levels.
                        2    Do not do any conversion.
                lscale=<n>
                     Select  the  scaling function to use for luminance scal‐
                     ing.  Only valid for yuv modes 2, 3, 4 and 6.
                        0    Use simple linear filtering (default).
                        1    Use bicubic B-spline filtering (better quality).
                             Needs  one additional texture unit.  Older cards
                             will not be able to handle this  for  chroma  at
                             least in fullscreen mode.
                        2    Use  cubic  filtering in horizontal, linear fil‐
                             tering in vertical direction.  Works  on  a  few
                             more cards than method 1.
                        3    Same  as  1  but  does not use a lookup texture.
                             Might be faster on some cards.
                        4    Use experimental unsharp masking with  3x3  sup‐
                             port  and a default strength of 0.5 (see filter-
                             strength).
                        5    Use experimental unsharp masking with  5x5  sup‐
                             port  and a default strength of 0.5 (see filter-
                             strength).
                        64   Use nearest-neighbor scaling.
                cscale=<n>
                     Select the scaling function to use for chrominance scal‐
                     ing.  For details see lscale.
                filter-strength=<value>
                     Set  the  effect  strength for the lscale/cscale filters
                     that support it.
                noise-strength=<value>
                     Set how much noise to add. 0 to disable  (default),  1.0
                     for level suitable for dithering to 6 bit.
                stereo=<value>
                     Select a method for stereo display.  You may have to use
                     -aspect to fix the aspect value.  Add 32  to  swap  left
                     and  right  side.   Experimental, do not expect too much
                     from it.
                        0    normal 2D display
                        1    Convert side by side input  to  full-color  red-
                             cyan stereo.
                        2    Convert  side by side input to full-color green-
                             magenta stereo.
                        3    Convert  side  by  side  input  to  quadbuffered
                             stereo.   Only  supported  by  very  few  OpenGL
                             cards.
                        4    Mix left and right in a pixel pattern.   Pattern
                             is given by stipple option.
                stipple=<bit
                     Lowest  16  bit  give  the  4x4 pattern to use (default:
                     0x0f0f).  Examples to try:  0x0f0f,  0xf0f0:  horizontal
                     lines;  0xaaaa,  0x5555: vertical lines; 0xa5a5, 0x5a5a:
                     checkerboard pattern

             The following options are only useful if writing your own  frag‐
             ment programs.

                customprog=<filename>
                     Load  a  custom  fragment  program from <filename>.  See
                     TOOLS/edgedect.fp for an example.
                customtex=<filename>
                     Load a custom  "gamma  ramp"  texture  from  <filename>.
                     This  can  be used in combination with yuv=4 or with the
                     customprog option.
                (no)customtlin
                     If enabled (default) use GL_LINEAR interpolation, other‐
                     wise use GL_NEAREST for customtex texture.
                (no)customtrect
                     If enabled, use texture_rectangle for customtex texture.
                     Default is disabled.
                (no)mipmapgen
                     If enabled, mipmaps for the video are automatically gen‐
                     erated.  This should be useful together with the custom‐
                     prog and the TXB instruction to implement  blur  filters
                     with  a  large  radius.  For most OpenGL implementations
                     this is very slow for any non-RGB formats.   Default  is
                     disabled.

             Normally  there  is no reason to use the following options, they
             mostly exist for testing purposes.

                (no)glfinish
                     Call glFinish() before swapping buffers.  Slower but  in
                     some cases more correct output (default: disabled).
                (no)manyfmts
                     Enables  support  for  more  (RGB and BGR) color formats
                     (default: enabled).  Needs OpenGL version >= 1.2.
                slice-height=<0-...>
                     Number of lines copied to texture in one piece (default:
                     0).  0 for whole image.
                     NOTE:  If  YUV  colorspace  is used (see yuv suboption),
                     special rules apply:
                        If the decoder uses slice rendering (see  -noslices),
                        this setting has no effect, the size of the slices as
                        provided by the decoder is used.
                        If the decoder does not use slice rendering, the  de‐
                        fault is 16.
                (no)osd
                     Enable  or  disable support for OSD rendering via OpenGL
                     (default: enabled).  This option is for testing; to dis‐
                     able the OSD use -osdlevel 0 instead.
                (no)aspect
                     Enable  or  disable aspect scaling and pan-and-scan sup‐
                     port  (default:  enabled).   Disabling  might   increase
                     speed.

      gl_tiled
             Variant  of  the  OpenGL  video  output driver.  Supports videos
             larger than the maximum texture size but lacks many of  the  ad‐
             vanced  features  and  optimizations of the gl driver and is un‐
             likely to be extended further.
                (no)glfinish
                     same as gl (default: enabled)
                yuv=<n>
                     Select the type of YUV to RGB  conversion.   If  set  to
                     anything  except  0 OSD will be disabled and brightness,
                     contrast and gamma setting is  only  available  via  the
                     global  X  server  settings.  Apart from this the values
                     have the same meaning as for -vo gl.

      matrixview
             OpenGL-based renderer creating a  Matrix-like  running-text  ef‐
             fect.
                cols=<n>
                     Number  of  text columns to display.  Very low values (<
                     16) will probably fail due to scaler limitations.   Val‐
                     ues not divisible by 16 may cause issues as well.
                rows=<n>
                     Number  of text rows to display.  Very low values (< 16)
                     will probably fail due to  scaler  limitations.   Values
                     not divisible by 16 may cause issues as well.

      null
             Produces no video output.  Useful for benchmarking.

      aa
             ASCII art video output driver that works on a text console.
             NOTE: The driver does not handle -aspect correctly.
             HINT:  You  probably  have  to specify -monitorpixelaspect.  Try
             'mplayer -vo aa -monitorpixelaspect 0.5'.

      caca
             Color ASCII art video output driver that works on  a  text  con‐
             sole.

      bl
             Video playback using the Blinkenlights UDP protocol.  This driv‐
             er is highly hardware specific.
                <subdevice>
                     Explicitly choose the Blinkenlights subdevice driver  to
                     use.  It is something like arcade:host=localhost:2323 or
                     hdl:file=name1,file=name2.  You must  specify  a  subde‐
                     vice.

      ggi
             GGI graphics system video output driver
                <driver>
                     Explicitly  choose  the  GGI driver to use.  Replace any
                     ',' that would appear in the driver string by a '.'.

      directfb
             Play video using the DirectFB library.
                (no)input
                     Use the DirectFB instead of the  MPlayer  keyboard  code
                     (default: enabled).
                buffermode=single|double|triple
                     Double  and  triple  buffering  give best results if you
                     want to avoid tearing issues.  Triple buffering is  more
                     efficient  than  double  buffering  as it does not block
                     MPlayer while waiting for the vertical retrace.   Single
                     buffering should be avoided (default: single).
                fieldparity=top|bottom
                     Control the output order for interlaced frames (default:
                     disabled).  Valid values are top  =  top  fields  first,
                     bottom = bottom fields first.  This option does not have
                     any effect on progressive film material like  most  MPEG
                     movies  are.  You need to enable this option if you have
                     tearing issues or unsmooth motions  watching  interlaced
                     film material.
                layer=N
                     Will  force  layer with ID N for playback (default: -1 -
                     auto).
                dfbopts=<list>
                     Specify a parameter list for DirectFB.

      dfbmga
             Matrox G400/G450/G550 specific video output driver that uses the
             DirectFB  library to make use of special hardware features.  En‐
             ables CRTC2 (second head), displaying video independently of the
             first head.
                (no)input
                     same as directfb (default: disabled)
                buffermode=single|double|triple
                     same as directfb (default: triple)
                fieldparity=top|bottom
                     same as directfb
                (no)bes
                     Enable  the  use of the Matrox BES (backend scaler) (de‐
                     fault: disabled).  Gives very  good  results  concerning
                     speed  and  output  quality as interpolated picture pro‐
                     cessing is done in hardware.  Works only on the  primary
                     head.
                (no)spic
                     Make  use of the Matrox sub picture layer to display the
                     OSD (default: enabled).
                (no)crtc2
                     Turn on TV-out on the second  head  (default:  enabled).
                     The output quality is amazing as it is a full interlaced
                     picture with proper sync to every odd/even field.
                tvnorm=pal|ntsc|auto
                     Will set the TV norm of the Matrox card without the need
                     for   modifying   /etc/directfbrc  (default:  disabled).
                     Valid norms are pal = PAL, ntsc = NTSC.  Special norm is
                     auto  (auto-adjust  using  PAL/NTSC)  because it decides
                     which norm to use by looking at  the  framerate  of  the
                     movie.

      mga (Linux only)
             Matrox  specific  video  output driver that makes use of the YUV
             back end scaler on Gxxx cards through a kernel module.   If  you
             have a Matrox card, this is the fastest option.
                <device>
                     Explicitly  choose  the  Matrox  device name to use (de‐
                     fault: /dev/mga_vid).

      xmga (Linux, X11 only)
             The mga video output driver, running in an X11 window.
                <device>
                     Explicitly choose the Matrox device  name  to  use  (de‐
                     fault: /dev/mga_vid).

      s3fb (Linux only) (also see -dr)
             S3 Virge specific video output driver.  This driver supports the
             card's YUV conversion and scaling, double buffering  and  direct
             rendering  features.  Use -vf format=yuy2 to get hardware-accel‐
             erated YUY2 rendering, which is much faster than  YV12  on  this
             card.
                <device>
                     Explicitly choose the fbdev device name to use (default:
                     /dev/fb0).

      wii (Linux only)
             Nintendo Wii/GameCube specific video output driver.

      3dfx (Linux only)
             3dfx-specific video output driver that directly uses  the  hard‐
             ware on top of X11.  Only 16 bpp are supported.

      tdfxfb (Linux only)
             This driver employs the tdfxfb framebuffer driver to play movies
             with YUV acceleration on 3dfx cards.
                <device>
                     Explicitly choose the fbdev device name to use (default:
                     /dev/fb0).

      tdfx_vid (Linux only)
             3dfx-specific video output driver that works in combination with
             the tdfx_vid kernel module.
                <device>
                     Explicitly choose the device name to use (default: /dev/
                     tdfx_vid).

      dxr2 (also see -dxr2) (DXR2 only)
             Creative DXR2 specific video output driver.
                <vo_driver>
                     Output video subdriver to use as overlay (x11, xv).

      dxr3 (DXR3 only)
             Sigma Designs em8300 MPEG decoder chip (Creative DXR3, Sigma De‐
             signs Hollywood Plus) specific video output  driver.   Also  see
             the lavc video filter.
                overlay
                     Activates the overlay instead of TV-out.
                prebuf
                     Turns on prebuffering.
                sync
                     Will turn on the new sync-engine.
                norm=<norm>
                     Specifies the TV norm.
                        0: Does not change current norm (default).
                        1: Auto-adjust using PAL/NTSC.
                        2: Auto-adjust using PAL/PAL-60.
                        3: PAL
                        4: PAL-60
                        5: NTSC
                <0-3>
                     Specifies the device number to use if you have more than
                     one em8300 card.

      v4l2 (requires Linux 2.6.22+ kernel)
             Video output driver for V4L2 compliant cards with built-in hard‐
             ware MPEG decoder.  Also see the lavc video filter.
                <device>
                     Explicitly  choose  the  MPEG decoder device name to use
                     (default: /dev/video16).
                <output>
                     Explicitly choose the TV-out output to be used  for  the
                     video signal.

      mpegpes (DVB only)
             Video  output  driver for DVB cards that writes the output to an
             MPEG-PES file if no DVB card is installed.
                card=<1-4>
                     Specifies the device number to use if you have more than
                     one  DVB  output card (V3 API only, such as 1.x.y series
                     drivers).  If not  specified  MPlayer  will  search  the
                     first usable card.
                <filename>
                     output filename (default: ./grab.mpg)

      zr (also see -zr* and -zrhelp)
             Video  output  driver  for  a  number  of MJPEG capture/playback
             cards.

      zr2 (also see the zrmjpeg video filter)
             Video output driver  for  a  number  of  MJPEG  capture/playback
             cards, second generation.
                dev=<device>
                     Specifies the video device to use.
                norm=<PAL|NTSC|SECAM|auto>
                     Specifies the video norm to use (default: auto).
                (no)prebuf
                     (De)Activate prebuffering, not yet supported.

      md5sum
             Calculate MD5 sums of each frame and write them to a file.  Sup‐
             ports RGB24 and YV12 colorspaces.  Useful for debugging.
                outfile=<value>
                     Specify the output filename (default: ./md5sums).

      yuv4mpeg
             Transforms the video stream into a sequence of uncompressed  YUV
             4:2:0  images  and  stores it in a file (default: ./stream.yuv).
             The format is the same as the one  employed  by  mjpegtools,  so
             this  is useful if you want to process the video with the mjpeg‐
             tools suite.  It supports the YV12 format.  If your source  file
             has  a  different format and is interlaced, make sure to use -vf
             scale=::1 to ensure the conversion uses  interlaced  mode.   You
             can  combine  it  with the -fixed-vo option to concatenate files
             with the same dimensions and fps value.
                interlaced
                     Write the output as interlaced frames, top field first.
                interlaced_bf
                     Write the output  as  interlaced  frames,  bottom  field
                     first.
                file=<filename>
                     Write  the  output  to <filename> instead of the default
                     stream.yuv.

             NOTE: If you do not specify any option the output is progressive
             (i.e. not interlaced).

      gif89a
             Output each frame into a single animated GIF file in the current
             directory.  It supports only RGB format with 24 bpp and the out‐
             put is converted to 256 colors.
                <fps>
                     Float value to specify framerate (default: 5.0).
                <output>
                     Specify the output filename (default: ./out.gif).

             NOTE:  You must specify the framerate before the filename or the
             framerate will be part of the filename.

             EXAMPLE:
                mplayer video.nut -vo gif89a:fps=15:output=test.gif

      jpeg
             Output each frame into a JPEG file  in  the  current  directory.
             Each  file  takes  the frame number padded with leading zeros as
             name.
                [no]progressive
                     Specify standard or progressive  JPEG  (default:  nopro‐
                     gressive).
                [no]baseline
                     Specify use of baseline or not (default: baseline).
                optimize=<0-100>
                     optimization factor (default: 100)
                smooth=<0-100>
                     smooth factor (default: 0)
                quality=<0-100>
                     quality factor (default: 75)
                outdir=<dirname>
                     Specify  the  directory  to  save the JPEG files to (de‐
                     fault: ./).
                subdirs=<prefix>
                     Create numbered subdirectories with the specified prefix
                     to save the files in instead of the current directory.
                maxfiles=<value> (subdirs only)
                     Maximum  number  of  files to be saved per subdirectory.
                     Must be equal to or larger than 1 (default: 1000).

      pnm
             Output each frame into a PNM  file  in  the  current  directory.
             Each  file  takes  the frame number padded with leading zeros as
             name.  It supports PPM, PGM and PGMYUV files  in  both  raw  and
             ASCII mode.  Also see pnm(5), ppm(5) and pgm(5).
                ppm
                     Write PPM files (default).
                pgm
                     Write PGM files.
                pgmyuv
                     Write  PGMYUV  files.   PGMYUV  is like PGM, but it also
                     contains the U and V plane, appended at  the  bottom  of
                     the picture.
                raw
                     Write PNM files in raw mode (default).
                ascii
                     Write PNM files in ASCII mode.
                outdir=<dirname>
                     Specify the directory to save the PNM files to (default:
                     ./).
                subdirs=<prefix>
                     Create numbered subdirectories with the specified prefix
                     to save the files in instead of the current directory.
                maxfiles=<value> (subdirs only)
                     Maximum  number  of  files to be saved per subdirectory.
                     Must be equal to or larger than 1 (default: 1000).

      png
             Output each frame into a PNG  file  in  the  current  directory.
             Each  file  takes  the frame number padded with leading zeros as
             name.  24bpp RGB and BGR formats are supported.
                z=<0-9>
                     Specifies the compression level.  0 is no compression, 9
                     is maximum compression.
                outdir=<dirname>
                     Specify the directory to save the PNG files to (default:
                     ./).
                prefix=<prefix>
                     Specify the prefix to be used for the PNG filenames (de‐
                     fault: no prefix).
                alpha
                     Create  PNG  files  with  an  alpha  channel.  Note that
                     MPlayer in general does not support alpha, so this  will
                     only be useful in some rare cases.

      mng
             Output  video  into an animated MNG file using 24 bpp RGB images
             with lossless compression.
                output=<filename>
                     Specify the output filename (default: out.mng).

             EXAMPLE:
                mplayer video.mkv -vo mng:output=test.mng

      tga
             Output each frame into a Targa file in  the  current  directory.
             Each  file  takes  the frame number padded with leading zeros as
             name.  The purpose of this video output driver is to have a sim‐
             ple  lossless  image writer to use without any external library.
             It supports the BGR[A] color format, with 15,  24  and  32  bpp.
             You can force a particular format with the format video filter.

             EXAMPLE:
                mplayer video.nut -vf format=bgr15 -vo tga

DECODING/FILTERING OPTIONS
      -ac <[-|+]codec1,[-|+]codec2,...[,]>
             Specify a priority list of audio codecs to be used, according to
             their codec name in codecs.conf.  Use a  '-'  before  the  codec
             name  to  omit it.  Use a '+' before the codec name to force it,
             this will likely crash!  If the list has a trailing ','  MPlayer
             will fall back on codecs not contained in the list.
             NOTE: See -ac help for a full list of available codecs.

             EXAMPLE:
                -ac mp3acm
                     Force the l3codeca.acm MP3 codec.
                -ac mad,
                     Try libmad first, then fall back on others.
                -ac hwac3,a52,
                     Try  hardware AC-3 passthrough, software AC-3, then oth‐
                     ers.
                -ac hwdts,
                     Try hardware DTS passthrough, then fall back on others.
                -ac -ffmp3,
                     Skip FFmpeg's MP3 decoder.

      -af-adv <force=(0-7):list=(filters)> (also see -af)
             Specify advanced audio filter options:

                force=<0-7>
                     Forces the insertion of audio filters to one of the fol‐
                     lowing:
                        0:  Use  completely  automatic filter insertion (cur‐
                        rently identical to 1).
                        1: Optimize for accuracy (default).
                        2: Optimize for speed.  Warning: Some features in the
                        audio filters may silently fail, and the sound quali‐
                        ty may drop.
                        3: Use no automatic insertion of filters and no opti‐
                        mization.   Warning:  It  may  be  possible  to crash
                        MPlayer using this setting.
                        4: Use automatic insertion of filters according to  0
                        above,  but use floating point processing when possi‐
                        ble.
                        5: Use automatic insertion of filters according to  1
                        above,  but use floating point processing when possi‐
                        ble.
                        6: Use automatic insertion of filters according to  2
                        above,  but use floating point processing when possi‐
                        ble.
                        7: Use no automatic insertion of filters according to
                        3  above, and use floating point processing when pos‐
                        sible.

                list=<filters>
                     Same as -af.

      -afm <driver1,driver2,...>
             Specify a priority list of audio codec families to be used,  ac‐
             cording  to  their codec name in codecs.conf.  Falls back on the
             default codecs if none of the given codec families work.
             NOTE: See -afm help for a full list of available codec families.

             EXAMPLE:
                -afm ffmpeg
                     Try FFmpeg's libavcodec codecs first.
                -afm acm,dshow
                     Try Win32 codecs first.

      -aspect <ratio> (also see -zoom)
             Override movie aspect ratio, in case aspect information  is  in‐
             correct or missing in the file being played.

             EXAMPLE:
                -aspect 4:3  or -aspect 1.3333
                -aspect 16:9 or -aspect 1.7777

      -noaspect
             Disable automatic movie aspect ratio compensation.

      -field-dominance <-1-1>
             Set  first field for interlaced content.  Useful for deinterlac‐
             ers that double the framerate: -vf tfields=1, -vf  yadif=1,  -vo
             vdpau:deint and -vo xvmc:bobdeint.
                -1   auto  (default):  If the decoder does not export the ap‐
                     propriate information, it falls back  to  0  (top  field
                     first).
                0    top field first
                1    bottom field first

      -flip
             Flip image upside-down.

      -lavdopts <option1:option2:...> (DEBUG CODE)
             Specify  libavcodec  decoding parameters.  Separate multiple op‐
             tions with a colon.

             EXAMPLE:
                -lavdopts gray:skiploopfilter=all:skipframe=nonref

             Available options are:

                bitexact
                     Only use bit-exact algorithms in all decoding steps (for
                     codec testing).

                bug=<value>
                     Manually work around encoder bugs.
                        0: nothing
                        1: autodetect bugs (default)
                        2  (msmpeg4v3):  some  old  lavc  generated msmpeg4v3
                        files (no autodetection)
                        4 (mpeg4):  Xvid  interlacing  bug  (autodetected  if
                        fourcc==XVIX)
                        8 (mpeg4): UMP4 (autodetected if fourcc==UMP4)
                        16 (mpeg4): padding bug (autodetected)
                        32 (mpeg4): illegal vlc bug (autodetected per fourcc)
                        64  (mpeg4): Xvid and DivX qpel bug (autodetected per
                        fourcc/version)
                        128 (mpeg4):  old  standard  qpel  (autodetected  per
                        fourcc/version)
                        256 (mpeg4): another qpel bug (autodetected per four‐
                        cc/version)
                        512 (mpeg4): direct-qpel-blocksize bug  (autodetected
                        per fourcc/version)
                        1024  (mpeg4):  edge  padding  bug  (autodetected per
                        fourcc/version)

                debug=<value>
                     Display debugging information.
                        0: disabled
                        1: picture info
                        2: rate control
                        4: bitstream
                        8: macroblock (MB) type
                        16: per-block quantization parameter (QP)
                        32: motion vector
                        0x0040: motion vector visualization (use -noslices)
                        0x0080: macroblock (MB) skip
                        0x0100: startcode
                        0x0200: PTS
                        0x0400: error resilience
                        0x0800: memory management control operations (H.264)
                        0x1000: bugs
                        0x2000: Visualize quantization parameter (QP),  lower
                        QP are tinted greener.
                        0x4000: Visualize block types.

                ec=<value>
                     Set error concealment strategy.
                        1: Use strong deblock filter for damaged MBs.
                        2: iterative motion vector (MV) search (slow)
                        3: all (default)

                er=<value>
                     Set error resilience strategy.
                        0: disabled
                        1: careful (Should work with broken encoders.)
                        2: normal (default) (Works with compliant encoders.)
                        3:  aggressive (More checks, but might cause problems
                        even for valid bitstreams.)
                        4: very aggressive

                fast (MPEG-2, MPEG-4, and H.264 only)
                     Enable optimizations which do not comply to the specifi‐
                     cation  and  might potentially cause problems, like sim‐
                     pler dequantization, simpler motion compensation, assum‐
                     ing use of the default quantization matrix, assuming YUV
                     4:2:0 and skipping a few checks to detect  damaged  bit‐
                     streams.

                gray
                     grayscale only decoding (a bit faster than with color)

                idct=<0-99> (see -lavcopts)
                     For  best  decoding  quality use the same IDCT algorithm
                     for decoding and encoding.  This may come at a price  in
                     accuracy, though.

                lowres=<number>[,<w>]
                     Decode at lower resolutions.  Low resolution decoding is
                     not supported by all codecs, and it will often result in
                     ugly artifacts.  This is not a bug, but a side effect of
                     not decoding at full resolution.
                        0: disabled
                        1: 1/2 resolution
                        2: 1/4 resolution
                        3: 1/8 resolution
                     If <w> is specified lowres decoding will be used only if
                     the width of the video is major than or equal to <w>.
                o=<key>=<value>[,<key>=<value>[,...]]    Pass   AVOptions  to
                libavcodec decoder.  Note, a patch to make  the  o=  unneeded
                and  pass  all unknown options through the AVOption system is
                welcome.  A full list of AVOptions can be found in the FFmpeg
                manual.   Note  that  some options may conflict with MEncoder
                options.

                     EXAMPLE:
                          o=debug=pict

                sb=<number> (MPEG-2 only)
                     Skip the given number of macroblock rows at the bottom.

                st=<number> (MPEG-2 only)
                     Skip the given number of macroblock rows at the top.

                skiploopfilter=<skipvalue> (H.264 only)
                     Skips the loop filter (AKA deblocking) during H.264  de‐
                     coding.  Since the filtered frame is supposed to be used
                     as reference for decoding dependent frames  this  has  a
                     worse  effect  on  quality  than not doing deblocking on
                     e.g. MPEG-2 video.  But at least for high  bitrate  HDTV
                     this  provides  a  big  speedup  with no visible quality
                     loss.

                     <skipvalue> can be either one of the following:
                        none: Never skip.
                        default: Skip useless processing steps (e.g.  0  size
                        packets in AVI).
                        nonref: Skip frames that are not referenced (i.e. not
                        used for decoding  other  frames,  the  error  cannot
                        "build up").
                        bidir: Skip B-Frames.
                        nonkey: Skip all frames except keyframes.
                        all: Skip all frames.

                skipidct=<skipvalue> (MPEG-1/2 only)
                     Skips  the IDCT step.  This degrades quality a lot of in
                     almost all cases (see skiploopfilter for available  skip
                     values).

                skipframe=<skipvalue>
                     Skips  decoding  of frames completely.  Big speedup, but
                     jerky motion and sometimes bad artifacts (see  skiploop‐
                     filter for available skip values).

                threads=<1-32>
                     Maximum  number  of  threads used for decoding.  Not all
                     decoders support threading  and  decoders  with  support
                     might  stay  below  the  provided limit at their choice.
                     (default: 1)


                vismv=<value>
                     Visualize motion vectors.
                        0: disabled
                        1: Visualize forward predicted MVs of P-frames.
                        2: Visualize forward predicted MVs of B-frames.
                        4: Visualize backward predicted MVs of B-frames.

                vstats
                     Prints  some  statistics  and  stores  them   in   ./vs‐
                     tats_*.log.

                wait_keyframe
                     Wait  for a keyframe before displaying anything.  Avoids
                     broken frames at startup or after seeking with some for‐
                     mats.

      -noslices
             Disable  drawing  video by 16-pixel height slices/bands, instead
             draws the whole frame in a single run.  May be faster or slower,
             depending on video card and available cache.  It has effect only
             with libmpeg2 and libavcodec codecs.

      -nosound
             Do not play/encode sound.  Useful for benchmarking.

      -novideo
             Do not play/encode video.  In many cases this will not work, use
             -vc null -vo null instead.

      -pp <quality> (also see -vf pp)
             Set  the DLL postprocess level.  This option is no longer usable
             with -vf pp.  It only works with Win32 DirectShow DLLs with  in‐
             ternal  postprocessing  routines.  The valid range of -pp values
             varies by codec, it is mostly 0-6, where  0=disable,  6=slowest/
             best.

      -pphelp (also see -vf pp)
             Show a summary about the available postprocess filters and their
             usage.

      -ssf <mode>
             Specifies software scaler parameters.

             EXAMPLE:
                -vf scale -ssf lgb=3.0
                lgb=<0-100>
                     gaussian blur filter (luma)
                cgb=<0-100>
                     gaussian blur filter (chroma)
                ls=<-100-100>
                     sharpen filter (luma)
                cs=<-100-100>
                     sharpen filter (chroma)
                chs=<h>
                     chroma horizontal shifting
                cvs=<v>
                     chroma vertical shifting

      -stereo <mode>
             Select type of MP2/MP3 stereo output.
                0    stereo
                1    left channel
                2    right channel

      -sws <software scaler type> (also see -vf scale and -zoom)
             Specify the software scaler algorithm to be used with the  -zoom
             option.   This  affects video output drivers which lack hardware
             acceleration, e.g. x11.

             Available types are:

                0    fast bilinear
                1    bilinear
                2    bicubic (good quality) (default)
                3    experimental
                4    nearest neighbor (bad quality)
                5    area
                6    luma bicubic / chroma bilinear
                7    gauss
                8    sincR
                9    lanczos
                10   natural bicubic spline

             NOTE: Some -sws options are tunable.   The  description  of  the
             scale video filter has further information.

      -swsopts <software scaler options>
             Specify advanced software scaler options provided by libswscale.
             Set to "help" to get a list of supported options.

      -vc <[-|+]codec1,[-|+]codec2,...[,]>
             Specify a priority list of video codecs to be used, according to
             their  codec  name  in  codecs.conf.  Use a '-' before the codec
             name to omit it.  Use a '+' before the codec name to  force  it,
             this  will likely crash!  If the list has a trailing ',' MPlayer
             will fall back on codecs not contained in the list.
             NOTE: See -vc help for a full list of available codecs.

             EXAMPLE:
                -vc divx
                     Force Win32/VfW DivX codec, no fallback.
                -vc -divxds,-divx,
                     Skip Win32 DivX codecs.
                -vc ffmpeg12,mpeg12,
                     Try libavcodec's MPEG-1/2  codec,  then  libmpeg2,  then
                     others.

      -vfm <driver1,driver2,...>
             Specify  a priority list of video codec families to be used, ac‐
             cording to their names in codecs.conf.  Falls back  on  the  de‐
             fault codecs if none of the given codec families work.
             NOTE: See -vfm help for a full list of available codec families.

             EXAMPLE:
                -vfm ffmpeg,dshow,vfw
                     Try the libavcodec, then Directshow, then VfW codecs and
                     fall back on others, if they do not work.
                -vfm xanim
                     Try XAnim codecs first.

      -x <x> (also see -zoom) (MPlayer only)
             Scale image to width <x> (if software/hardware scaling is avail‐
             able).  Disables aspect calculations.

      -xvidopts <option1:option2:...>
             Specify additional parameters when decoding with Xvid.
             NOTE: Since libavcodec is faster than Xvid you might want to use
             the libavcodec postprocessing filter (-vf pp) and decoder  (-vfm
             ffmpeg) instead.

             Xvid's internal postprocessing filters:
                deblock-chroma (also see -vf pp)
                     chroma deblock filter
                deblock-luma (also see -vf pp)
                     luma deblock filter
                dering-luma (also see -vf pp)
                     luma deringing filter
                dering-chroma (also see -vf pp)
                     chroma deringing filter
                filmeffect (also see -vf noise)
                     Adds  artificial  film grain to the video.  May increase
                     perceived quality, while lowering true quality.

             rendering methods:
                dr2
                     Activate direct rendering method 2.
                nodr2
                     Deactivate direct rendering method 2.

      -xy <value> (also see -zoom)
                value<=8
                     Scale image by factor <value>.
                value>8
                     Set width to value and calculate height to keep  correct
                     aspect ratio.

      -y <y> (also see -zoom) (MPlayer only)
             Scale  image  to  height  <y>  (if  software/hardware scaling is
             available).  Disables aspect calculations.

      -zoom
             Allow software scaling, where available.  This will allow  scal‐
             ing  with  output  drivers (like x11, fbdev) that do not support
             hardware scaling where MPlayer disables scaling by  default  for
             performance reasons.

AUDIO FILTERS
      Audio  filters allow you to modify the audio stream and its properties.
      The syntax is:

      -af <filter1[=parameter1:parameter2:...],filter2,...>
             Setup a chain of audio filters.

      NOTE: To get a full list of available audio filters, see -af help.

      Audio filters are managed in lists.  There are a few commands to manage
      the filter list.

      -af-add <filter1[,filter2,...]>
             Appends the filters given as arguments to the filter list.

      -af-pre <filter1[,filter2,...]>
             Prepends the filters given as arguments to the filter list.

      -af-del <index1[,index2,...]>
             Deletes  the  filters at the given indexes.  Index numbers start
             at 0, negative numbers address the end of the list  (-1  is  the
             last).

      -af-clr
             Completely empties the filter list.

      Available filters are:

      resample[=srate[:sloppy[:type]]]
             Changes the sample rate of the audio stream.  Can be used if you
             have a fixed frequency sound card or if you are  stuck  with  an
             old sound card that is only capable of max 44.1kHz.  This filter
             is automatically enabled if necessary.  It only supports  16-bit
             integer and float in native-endian format as input.
             NOTE: With MEncoder, you need to also use -srate <srate>.
                <srate>
                     output sample frequency in Hz.  The valid range for this
                     parameter is 8000 to 192000.  If the  input  and  output
                     sample  frequency  are  the same or if this parameter is
                     omitted the filter is automatically  unloaded.   A  high
                     sample  frequency  normally  improves the audio quality,
                     especially when used in combination with other filters.
                <sloppy>
                     Allow (1) or disallow (0) the output frequency to differ
                     slightly  from  the frequency given by <srate> (default:
                     1).  Can be used if the startup of the playback  is  ex‐
                     tremely slow.
                <type>
                     Select which resampling method to use.
                        0: linear interpolation (fast, poor quality especial‐
                        ly when upsampling)
                        1: polyphase filterbank and integer processing
                        2: polyphase filterbank and floating point processing
                        (slow, best quality)

             EXAMPLE:
                mplayer -af resample=44100:0:0
                     would set the output frequency of the resample filter to
                     44100Hz using exact output frequency scaling and  linear
                     interpolation.

      lavcresample[=srate[:length[:linear[:count[:cutoff]]]]]
             Changes  the  sample  rate  of  the  audio  stream to an integer
             <srate> in Hz.  It only supports the 16-bit  native-endian  for‐
             mat.
             NOTE: With MEncoder, you need to also use -srate <srate>.
                <srate>
                     the output sample rate
                <length>
                     length  of the filter with respect to the lower sampling
                     rate (default: 16)
                <linear>
                     if 1 then filters will be linearly interpolated  between
                     polyphase entries
                <count>
                     log2  of the number of polyphase entries (..., 10->1024,
                     11->2048, 12->4096, ...)  (default: 10->1024)
                <cutoff>
                     cutoff frequency (0.0-1.0), default set  depending  upon
                     filter length

      lavcac3enc[=tospdif[:bitrate[:minchn]]]
             Encode  multi-channel audio to AC-3 at runtime using libavcodec.
             Supports 16-bit native-endian input format, maximum 6  channels.
             The  output is big-endian when outputting a raw AC-3 stream, na‐
             tive-endian when outputting to S/PDIF.  The output  sample  rate
             of  this  filter is same with the input sample rate.  When input
             sample rate is 48kHz, 44.1kHz, or 32kHz,  this  filter  directly
             use  it.   Otherwise a resampling filter is auto-inserted before
             this filter to make the input and output sample rate  be  48kHz.
             You need to specify '-channels N' to make the decoder decode au‐
             dio into N-channel, then the filter can encode the N-channel in‐
             put to AC-3.
                <tospdif>
                     Output  raw  AC-3  stream  if zero or not set, output to
                     S/PDIF for passthrough when <tospdif> is set non-zero.
                <bitrate>
                     The bitrate to encode the AC-3 stream.  Set it to either
                     384  or  384000  to get 384kbits.  Valid values: 32, 40,
                     48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256,
                                   320, 384, 448, 512, 576, 640  Default  bi‐
                     trate  is  based  on  the input channel number: 1ch: 96,
                     2ch: 192,  3ch: 224,  4ch: 384,  5ch: 448,  6ch: 448
                <minchn>
                     If the input channel number is less than  <minchn>,  the
                     filter will detach itself (default: 5).

      sweep[=speed]
             Produces a sine sweep.
                <0.0-1.0>
                     Sine  function  delta,  use  very low values to hear the
                     sweep.

      sinesuppress[=freq:decay]
             Remove a sine at the specified frequency.  Useful to get rid  of
             the  50/60Hz  noise on low quality audio equipment.  It probably
             only works on mono input.
                <freq>
                     The frequency of the sine which should  be  removed  (in
                     Hz) (default: 50)
                <decay>
                     Controls  the  adaptivity  (a larger value will make the
                     filter adapt to amplitude and phase changes  quicker,  a
                     smaller value will make the adaptation slower) (default:
                     0.0001).  Reasonable values are around 0.001.

      bs2b[=option1:option2:...]
             Bauer stereophonic to  binaural  transformation  using  libbs2b.
             Improves  the headphone listening experience by making the sound
             similar to that from loudspeakers, allowing  each  ear  to  hear
             both  channels  and  taking into account the distance difference
             and the head shadowing effect.  It is applicable only to 2 chan‐
             nel audio.
                fcut=<300-1000>
                     Set cut frequency in Hz.
                feed=<10-150>
                     Set feed level for low frequencies in 0.1*dB.
                profile=<value>
                     Several profiles are available for convenience:
                          default
                               will  be  used  if  nothing else was specified
                               (fcut=700, feed=45)
                          cmoy
                               Chu  Moy  circuit  implementation   (fcut=700,
                               feed=60)
                          jmeier
                               Jan  Meier  circuit  implementation (fcut=650,
                               feed=95)

             If fcut or feed options are specified together with  a  profile,
             they will be applied on top of the selected profile.

      hrtf[=flag]
             Head-related transfer function: Converts multichannel audio to 2
             channel output for headphones, preserving the spatiality of  the
             sound.

             Flag  Meaning
             m     matrix decoding of the rear channel
             s     2-channel matrix decoding
             0     no matrix decoding (default)

      equalizer=[g1:g2:g3:...:g10]
             10  octave band graphic equalizer, implemented using 10 IIR band
             pass filters.  This means that it works regardless of what  type
             of  audio  is being played back.  The center frequencies for the
             10 bands are:

             No. frequency
             0    31.25 Hz
             1    62.50 Hz
             2   125.00 Hz
             3   250.00 Hz
             4   500.00 Hz
             5    1.00 kHz
             6    2.00 kHz
             7    4.00 kHz
             8    8.00 kHz
             9   16.00 kHz

             If the sample rate of the sound being played is lower  than  the
             center  frequency  for  a frequency band, then that band will be
             disabled.  A known bug with this filter is that the characteris‐
             tics  for the uppermost band are not completely symmetric if the
             sample rate is close to the center frequency of that band.  This
             problem  can  be worked around by upsampling the sound using the
             resample filter before it reaches this filter.
                <g1>:<g2>:<g3>:...:<g10>
                     floating point numbers representing the gain in  dB  for
                     each frequency band (-12-12)

             EXAMPLE:
                mplayer -af equalizer=11:11:10:5:0:-12:0:5:12:12 media.avi
                     Would amplify the sound in the upper and lower frequency
                     region while canceling it almost completely around 1kHz.

      channels=nch[:nr:from1:to1:from2:to2:from3:to3:...]
             Can be used for adding,  removing,  routing  and  copying  audio
             channels.   If  only <nch> is given the default routing is used,
             it works as follows: If the number of output channels is  bigger
             than  the  number  of input channels empty channels are inserted
             (except mixing from mono to stereo, then the mono channel is re‐
             peated in both of the output channels).  If the number of output
             channels is smaller than the number of input  channels  the  ex‐
             ceeding channels are truncated.
                <nch>
                     number of output channels (1-8)
                <nr>
                     number of routes (1-8)
                <from1:to1:from2:to2:from3:to3:...>
                     Pairs  of  numbers  between 0 and 7 that define where to
                     route each channel.

             EXAMPLE:
                mplayer -af channels=4:4:0:1:1:0:2:2:3:3 media.avi
                     Would change the number of channels to 4 and  set  up  4
                     routes that swap channel 0 and channel 1 and leave chan‐
                     nel 2 and 3 intact.  Observe that  if  media  containing
                     two  channels  was  played  back, channels 2 and 3 would
                     contain silence but 0 and 1 would still be swapped.
                mplayer -af channels=6:4:0:0:0:1:0:2:0:3 media.avi
                     Would change the number of channels to 6 and  set  up  4
                     routes  that copy channel 0 to channels 0 to 3.  Channel
                     4 and 5 will contain silence.

      format[=format] (also see -format)
             Convert between different sample formats.  Automatically enabled
             when needed by the sound card or another filter.
                <format>
                     Sets  the  desired  format.   The general form is 'sbe',
                     where 's' denotes the sign (either 's' for signed or 'u'
                     for unsigned), 'b' denotes the number of bits per sample
                     (16, 24 or 32) and  'e'  denotes  the  endianness  ('le'
                     means  little-endian, 'be' big-endian and 'ne' the endi‐
                     anness of the computer MPlayer is  running  on).   Valid
                     values   (amongst  others)  are:  's16le',  'u32be'  and
                     'u24ne'.  Exceptions to this rule that  are  also  valid
                     format  specifiers:  u8,  s8, floatle, floatbe, floatne,
                     mulaw, alaw, mpeg2, ac3 and imaadpcm.

      volume[=v[:sc]]
             Implements software volume control.  Use this filter  with  cau‐
             tion since it can reduce the signal to noise ratio of the sound.
             In most cases it is best to set the level for the PCM  sound  to
             max,  leave this filter out and control the output level to your
             speakers with the master volume control of the mixer.   In  case
             your  sound  card  has  a digital PCM mixer instead of an analog
             one, and you hear distortion, use the MASTER mixer instead.   If
             there  is  an external amplifier connected to the computer (this
             is almost always the case), the noise level can be minimized  by
             adjusting  the master level and the volume knob on the amplifier
             until the hissing noise in the background is gone.
             This filter has a second feature: It measures the overall  maxi‐
             mum  sound  level  and prints out that level when MPlayer exits.
             This volume estimate can be used for setting the sound level  in
             MEncoder  such that the maximum dynamic range is utilized.  This
             feature currently only works with floating-point data, use  e.g.
             -af-adv force=5, or use -af stats.
             NOTE: This filter is not reentrant and can therefore only be en‐
             abled once for every audio stream.
                <v>
                     Sets the desired gain in dB  for  all  channels  in  the
                     stream  from  -200dB  to  +60dB,  where -200dB mutes the
                     sound completely and +60dB equals a gain  of  1000  (de‐
                     fault: 0).
                <sc>
                     Turns  soft  clipping  on (1) or off (0).  Soft-clipping
                     can make the sound more smooth if very high volume  lev‐
                     els  are  used.  Enable this option if the dynamic range
                     of the loudspeakers is very low.
                     WARNING: This feature creates distortion and  should  be
                     considered a last resort.

             EXAMPLE:
                mplayer -af volume=10.1:0 media.avi
                     Would  amplify  the sound by 10.1dB and hard-clip if the
                     sound level is too high.

      pan=n[:L00:L01:L02:...L10:L11:L12:...Ln0:Ln1:Ln2:...]
             Mixes channels arbitrarily.  Basically a combination of the vol‐
             ume  and  the  channels filter that can be used to down-mix many
             channels to only a few, e.g. stereo to mono or vary the  "width"
             of  the  center speaker in a surround sound system.  This filter
             is hard to use, and will require some tinkering before  the  de‐
             sired result is obtained.  The number of options for this filter
             depends on the number of output channels.   An  example  how  to
             downmix  a six-channel file to two channels with this filter can
             be found in the examples section near the end.
                <n>
                     number of output channels (1-8)
                <Lij>
                     How much of input channel i is mixed into output channel
                     j  (0-1).  So in principle you first have n numbers say‐
                     ing what to do with the first input channel, then n num‐
                     bers  that  act on the second input channel etc.  If you
                     do not specify any numbers for some input channels, 0 is
                     assumed.

             EXAMPLE:
                mplayer -af pan=1:0.5:0.5 media.avi
                     Would down-mix from stereo to mono.
                mplayer -af pan=3:1:0:0.5:0:1:0.5 media.avi
                     Would give 3 channel output leaving channels 0 and 1 in‐
                     tact, and mix channels 0 and 1  into  output  channel  2
                     (which could be sent to a subwoofer for example).

      sub[=fc:ch]
             Adds  a  subwoofer  channel to the audio stream.  The audio data
             used for creating the subwoofer channel is  an  average  of  the
             sound  in  channel 0 and channel 1.  The resulting sound is then
             low-pass filtered by a 4th order Butterworth filter with  a  de‐
             fault  cutoff  frequency of 60Hz and added to a separate channel
             in the audio stream.
             Warning: Disable this filter when you are playing DVDs with Dol‐
             by  Digital  5.1  sound,  otherwise this filter will disrupt the
             sound to the subwoofer.
                <fc>
                     cutoff frequency in Hz for the low-pass filter (20Hz  to
                     300Hz)  (default:  60Hz) For the best result try setting
                     the cutoff frequency as low as possible.  This will  im‐
                     prove the stereo or surround sound experience.
                <ch>
                     Determines  the  channel  number  in which to insert the
                     sub-channel audio.  Channel number can be between 0  and
                     7  (default:  5).   Observe  that the number of channels
                     will automatically be increased to <ch> if necessary.

             EXAMPLE:
                mplayer -af sub=100:4 -channels 5 media.avi
                     Would add a sub-woofer channel with a  cutoff  frequency
                     of 100Hz to output channel 4.

      center
             Creates a center channel from the front channels.  May currently
             be low quality as it does not implement a high-pass  filter  for
             proper  extraction yet, but averages and halves the channels in‐
             stead.
                <ch>
                     Determines the channel number in  which  to  insert  the
                     center  channel.   Channel number can be between 0 and 7
                     (default: 5).  Observe that the number of channels  will
                     automatically be increased to <ch> if necessary.

      surround[=delay]
             Decoder  for  matrix encoded surround sound like Dolby Surround.
             Many files with 2 channel audio actually contain  matrixed  sur‐
             round  sound.  Requires a sound card supporting at least 4 chan‐
             nels.
                <delay>
                     delay time in ms for the rear speakers (0 to 1000)  (de‐
                     fault: 20) This delay should be set as follows: If d1 is
                     the distance from the listening position  to  the  front
                     speakers and d2 is the distance from the listening posi‐
                     tion to the rear speakers, then the delay should be  set
                     to 15ms if d1 <= d2 and to 15 + 5*(d1-d2) if d1 > d2.

             EXAMPLE:
                mplayer -af surround=15 -channels 4 media.avi
                     Would  add  surround  sound decoding with 15ms delay for
                     the sound to the rear speakers.

      delay[=ch1:ch2:...]
             Delays the sound to the loudspeakers such that  the  sound  from
             the  different channels arrives at the listening position simul‐
             taneously.  It is only useful if you have more than 2 loudspeak‐
             ers.
                ch1,ch2,...
                     The  delay  in ms that should be imposed on each channel
                     (floating point number between 0 and 1000).

             To calculate the required delay for the different channels do as
             follows:

             1. Measure  the  distance to the loudspeakers in meters in rela‐
                tion to your listening position, giving you the distances  s1
                to  s5 (for a 5.1 system).  There is no point in compensating
                for the subwoofer (you will not hear the difference anyway).

             2. Subtract the distances s1 to s5 from  the  maximum  distance,
                i.e. s[i] = max(s) - s[i]; i = 1...5.

             3. Calculate  the required delays in ms as d[i] = 1000*s[i]/342;
                i = 1...5.

             EXAMPLE:
                mplayer -af delay=10.5:10.5:0:0:7:0 media.avi
                     Would delay front left and right by 10.5ms, the two rear
                     channels  and  the  sub by 0ms and the center channel by
                     7ms.

      export[=mmapped_file[:nsamples]]
             Exports the incoming signal to other processes using memory map‐
             ping (mmap()).  Memory mapped areas contain a header:

             int nch                      /*number of channels*/
             int size                     /*buffer size*/
             unsigned long long counter   /*Used to keep sync, updated every
                                            time new data is exported.*/

             The rest is payload (non-interleaved) 16 bit data.
                <mmapped_file>
                     file  to map data to (default: ~/.mplayer/mplayer-af_ex‐
                     port)
                <nsamples>
                     number of samples per channel (default: 512)

             EXAMPLE:
                mplayer -af export=/tmp/mplayer-af_export:1024 media.avi
                     Would export 1024 samples per channel to  '/tmp/mplayer-
                     af_export'.

      extrastereo[=mul]
             (Linearly) increases the difference between left and right chan‐
             nels which adds some sort of "live" effect to playback.
                <mul>
                     Sets the difference  coefficient  (default:  2.5).   0.0
                     means  mono  sound  (average of both channels), with 1.0
                     sound will be unchanged, with -1.0 left and right  chan‐
                     nels will be swapped.

      volnorm[=method:target]
             Maximizes the volume without distorting the sound.
                <method>
                     Sets the used method.
                        1:  Use  a single sample to smooth the variations via
                        the standard weighted mean  over  past  samples  (de‐
                        fault).
                        2:  Use  several samples to smooth the variations via
                        the standard weighted mean over past samples.

                <target>
                     Sets the target amplitude as a fraction of  the  maximum
                     for the sample type (default: 0.25).

      ladspa=file:label[:controls...]
             Load  a LADSPA (Linux Audio Developer's Simple Plugin API) plug‐
             in.  This filter is reentrant, so multiple LADSPA plugins can be
             used at once.
                <file>
                     Specifies   the   LADSPA   plugin   library   file.   If
                     LADSPA_PATH is set, it searches for the specified  file.
                     If  it  is  not  set,  you must supply a fully specified
                     pathname.
                <label>
                     Specifies the filter within the library.  Some libraries
                     contain  only  one  filter,  but  others contain many of
                     them.  Entering 'help' here,  will  list  all  available
                     filters  within  the specified library, which eliminates
                     the use of 'listplugins' from the LADSPA SDK.
                <controls>
                     Controls are zero or more floating point values that de‐
                     termine  the  behavior of the loaded plugin (for example
                     delay, threshold or gain).  In verbose mode (add  -v  to
                     the  MPlayer  command  line), all available controls and
                     their valid ranges are printed.  This eliminates the use
                     of 'analyseplugin' from the LADSPA SDK.

      comp
             Compressor/expander  filter  usable  for microphone input.  Pre‐
             vents artifacts on very loud sound and raises the volume on very
             low sound.  This filter is untested, maybe even unusable.

      gate
             Noise gate filter similar to the comp audio filter.  This filter
             is untested, maybe even unusable.

      karaoke
             Simple voice removal filter exploiting the fact  that  voice  is
             usually  recorded  with  mono gear and later 'center' mixed onto
             the final audio stream.  Beware that this filter will turn  your
             signal into mono.  Works well for 2 channel tracks; do not both‐
             er trying it on anything but 2 channel stereo.

      scaletempo[=option1:option2:...]
             Scales audio tempo without altering pitch, optionally synced  to
             playback speed (default).
             This  works by playing ´stride´ ms of audio at normal speed then
             consuming ´stride*scale´ ms  of  input  audio.   It  pieces  the
             strides  together  by  blending  ´overlap´% of stride with audio
             following the previous stride.  It optionally performs  a  short
             statistical  analysis on the next ´search´ ms of audio to deter‐
             mine the best overlap position.
                scale=<amount>
                     Nominal amount to scale tempo.  Scales  this  amount  in
                     addition to speed.  (default: 1.0)
                stride=<amount>
                     Length  in milliseconds to output each stride.  Too high
                     of value will  cause  noticeable  skips  at  high  scale
                     amounts and an echo at low scale amounts.  Very low val‐
                     ues will alter pitch.  Increasing improves  performance.
                     (default: 60)
                overlap=<percent>
                     Percentage  of  stride  to overlap.  Decreasing improves
                     performance.  (default: .20)
                search=<amount>
                     Length in milliseconds to search for best overlap  posi‐
                     tion.  Decreasing improves performance greatly.  On slow
                     systems, you will probably want to set  this  very  low.
                     (default: 14)
                speed=<tempo|pitch|both|none>
                     Set response to speed change.
                        tempo
                             Scale tempo in sync with speed (default).
                        pitch
                             Reverses effect of filter.  Scales pitch without
                             altering    tempo.     Add     ´[     speed_mult
                             0.9438743126816935´     and     ´]    speed_mult
                             1.059463094352953´ to your input.conf to step by
                             musical  semi-tones.   WARNING:  Loses sync with
                             video.
                        both Scale both tempo and pitch.
                        none Ignore speed changes.

             EXAMPLE:
                mplayer -af scaletempo -speed 1.2 media.ogg
                     Would playback media at 1.2x normal speed, with audio at
                     normal pitch.  Changing playback speed, would change au‐
                     dio tempo to match.
                mplayer -af scaletempo=scale=1.2:speed=none  -speed  1.2  me‐
                dia.ogg
                     Would playback media at 1.2x normal speed, with audio at
                     normal pitch, but changing playback speed has no  effect
                     on audio tempo.
                mplayer  -af  scaletempo=stride=30:overlap=.50:search=10  me‐
                dia.ogg
                     Would tweak the quality and performace parameters.
                mplayer -af format=floatne,scaletempo media.ogg
                     Would make scaletempo use float code.  Maybe  faster  on
                     some platforms.
                mplayer -af scaletempo=scale=1.2:speed=pitch audio.ogg
                     Would playback audio file at 1.2x normal speed, with au‐
                     dio at normal pitch.   Changing  playback  speed,  would
                     change pitch, leaving audio tempo at 1.2x.

      stats
             Collects and prints statistics about the audio stream, especial‐
             ly the volume.  These statistics are especially intended to help
             adjusting  the  volume while avoiding clipping.  The volumes are
             printed in dB and compatible with the volume audio filter,  they
             are always rounded towards -0dB.

             The 'n_samples' field is the total number of samples seen by the
             filter.  The 'mean_volume' field is the root mean  square.   The
             'max_volume' field is exactly what it says.  The 'histogram_Xdb'
             fields count how many samples were at -XdB,  for  X  just  below
             max_volume.

             For  example,  if  max_volume  is  -7dB and histogram_7dB is 19,
             'volume=7' will not cause clipping  and  'volume=8'  will  cause
             clipping on exactly 19 samples.

VIDEO FILTERS
      Video  filters allow you to modify the video stream and its properties.
      The syntax is:

      -vf <filter1[=parameter1:parameter2:...],filter2,...>
             Setup a chain of video filters.

      Many parameters are optional and set to default values if omitted.   To
      explicitly use a default value set a parameter to '-1'.  Parameters w:h
      means width x height in pixels, x:y means x;y position counted from the
      upper left corner of the bigger image.
      NOTE: To get a full list of available video filters, see -vf help.

      Video filters are managed in lists.  There are a few commands to manage
      the filter list.

      -vf-add <filter1[,filter2,...]>
             Appends the filters given as arguments to the filter list.

      -vf-pre <filter1[,filter2,...]>
             Prepends the filters given as arguments to the filter list.

      -vf-del <index1[,index2,...]>
             Deletes the filters at the given indexes.  Index  numbers  start
             at  0,  negative  numbers address the end of the list (-1 is the
             last).

      -vf-clr
             Completely empties the filter list.

      With filters that support it, you can access parameters by their name.

      -vf <filter>=help
             Prints the parameter names and parameter value ranges for a par‐
             ticular filter.

      -vf <filter=named_parameter1=value1[:named_parameter2=value2:...]>
             Sets  a  named  parameter to the given value.  Use on and off or
             yes and no to set flag parameters.

      Available filters are:

      crop[=w:h:x:y]
             Crops the given part of the image and discards the rest.  Useful
             to remove black bands from widescreen movies.
                <w>,<h>
                     Cropped width and height, defaults to original width and
                     height.
                <x>,<y>
                     Position of the cropped picture, defaults to center.

      cropdetect[=limit:round[:reset]]
             Calculates necessary cropping parameters and prints  the  recom‐
             mended parameters to stdout.
                <limit>
                     Threshold,  which can be optionally specified from noth‐
                     ing (0) to everything (255) (default: 24).
                <round>
                     Value which the width/height should be divisible by (de‐
                     fault:  16).   The  offset  is automatically adjusted to
                     center the video.  Use 2 to  get  only  even  dimensions
                     (needed  for  4:2:2 video).  16 is best when encoding to
                     most video codecs.
                <reset>
                     Counter that determines after how many frames cropdetect
                     will  reset  the  previously detected largest video area
                     and start over to detect the current optimal  crop  area
                     (default:  0).   This  can  be useful when channel logos
                     distort the video area.  0 indicates never reset and re‐
                     turn the largest area encountered during playback.

      rectangle[=w:h:x:y]
             Draws a rectangle of the requested width and height at the spec‐
             ified coordinates over the image and  prints  current  rectangle
             parameters  to  the  console.   This can be used to find optimal
             cropping parameters.   If  you  bind  the  input.conf  directive
             'change_rectangle'  to  keystrokes,  you can move and resize the
             rectangle on the fly.
                <w>,<h>
                     width and height (default: -1,  maximum  possible  width
                     where boundaries are still visible.)
                <x>,<y>
                     top  left  corner position (default: -1, uppermost left‐
                     most)

      expand[=w:h:x:y:o:a:r]
             Expands (not scales) movie resolution to  the  given  value  and
             places  the  unscaled original at coordinates x, y.  Can be used
             for placing subtitles/OSD in the resulting black bands.

                <w>,<h>
                     Expanded width,height (default: original  width,height).
                     Negative  values  for  w and h are treated as offsets to
                     the original size.

                     EXAMPLE:
                          expand=0:-50:0:0
                                 Adds a 50 pixel border to the bottom of  the
                                 picture.

                <x>,<y>
                     position  of  original  image on the expanded image (de‐
                     fault: center)

                <o>
                     OSD/subtitle rendering
                        0: disable (default)
                        1: enable

                <a>
                     Expands to fit an aspect instead of  a  resolution  (de‐
                     fault: 0).

                     EXAMPLE:
                          expand=800:::::4/3
                                 Expands  to  800x600,  unless  the source is
                                 higher resolution, in which case it  expands
                                 to fill a 4/3 aspect.

                <r>
                     Rounds up to make both width and height divisible by <r>
                     (default: 1).

      flip (also see -flip)
             Flips the image upside down.

      mirror
             Mirrors the image on the Y axis.

      rotate[=<0-7>]
             Rotates the image by 90 degrees and optionally  flips  it.   For
             values  between  4-7 rotation is only done if the movie geometry
             is portrait and not landscape.

                0    Rotate by 90 degrees clockwise and flip (default).

                1    Rotate by 90 degrees clockwise.

                2    Rotate by 90 degrees counterclockwise.

                3    Rotate by 90 degrees counterclockwise and flip.

      scale[=w:h[:interlaced[:chr_drop[:par[:par2[:pre‐
      size[:noup[:arnd]]]]]]]]
             Scales  the image with the software scaler (slow) and performs a
             YUV<->RGB colorspace conversion (also see -sws).

                <w>,<h>
                     scaled width/height (default: original width/height)
                     NOTE: If -zoom is used, and underlying filters  (includ‐
                     ing  libvo)  are  incapable  of  scaling, it defaults to
                     d_width/d_height!
                         0:   scaled d_width/d_height
                        -1:   original width/height
                        -2:   Calculate w/h using the other dimension and the
                        prescaled aspect ratio.
                        -3:   Calculate w/h using the other dimension and the
                        original aspect ratio.
                        -(n+8): Like -n above, but rounding the dimension  to
                        the closest multiple of 16.

                <interlaced>
                     Toggle interlaced scaling.
                        0: off (default)
                        1: on

                <chr_drop>
                     chroma skipping
                        0: Use all available input lines for chroma.
                        1: Use only every 2. input line for chroma.
                        2: Use only every 4. input line for chroma.
                        3: Use only every 8. input line for chroma.

                <par>[:<par2>] (also see -sws)
                     Set  some  scaling  parameters  depending on the type of
                     scaler selected with -sws.
                        -sws 2 (bicubic):  B (blurring) and C (ringing)
                        0.00:0.60 default
                        0.00:0.75 VirtualDub's "precise bicubic"
                        0.00:0.50 Catmull-Rom spline
                        0.33:0.33 Mitchell-Netravali spline
                        1.00:0.00 cubic B-spline
                        -sws 7 (gaussian): sharpness (0 (soft) - 100 (sharp))
                        -sws 9 (lanczos):  filter length (1-10)

                <presize>
                     Scale to preset sizes.
                        qntsc:   352x240 (NTSC quarter screen)
                        qpal:    352x288 (PAL quarter screen)
                        ntsc:    720x480 (standard NTSC)
                        pal:     720x576 (standard PAL)
                        sntsc:   640x480 (square pixel NTSC)
                        spal:    768x576 (square pixel PAL)

                <noup>
                     Disallow upscaling past the original dimensions.
                        0: Allow upscaling (default).
                        1: Disallow upscaling if one  dimension  exceeds  its
                        original value.
                        2: Disallow upscaling if both dimensions exceed their
                        original values.

                <arnd>
                     Accurate rounding for the vertical scaler, which may  be
                     faster or slower than the default rounding.
                        0: Disable accurate rounding (default).
                        1: Enable accurate rounding.

      dsize[=aspect|w:h:aspect-method:r]
             Changes  the  intended display size/aspect at an arbitrary point
             in the filter chain.  Aspect can be given as a fraction (4/3) or
             floating  point  number  (1.33).  Alternatively, you may specify
             the exact display width and height desired.  Note that this fil‐
             ter  does  not do any scaling itself; it just affects what later
             scalers (software or hardware) will do when auto-scaling to cor‐
             rect aspect.

                <w>,<h>
                     New display width and height.  Can also be these special
                     values:
                         0:   original display width and height
                        -1:   original video width and height (default)
                        -2:   Calculate w/h using the other dimension and the
                        original display aspect ratio.
                        -3:   Calculate w/h using the other dimension and the
                        original video aspect ratio.

                EXAMPLE:
                          dsize=800:-2
                                 Specifies a display  resolution  of  800x600
                                 for  a  4/3  aspect  video, or 800x450 for a
                                 16/9 aspect video.
                <aspect-method>
                     Modifies width and height according to  original  aspect
                     ratios.
                        -1: Ignore original aspect ratio (default).
                         0: Keep display aspect ratio by using <w> and <h> as
                        maximum resolution.
                         1: Keep display aspect ratio by using <w> and <h> as
                        minimum resolution.
                         2:  Keep  video aspect ratio by using <w> and <h> as
                        maximum resolution.
                         3: Keep video aspect ratio by using <w> and  <h>  as
                        minimum resolution.

                EXAMPLE:
                          dsize=800:600:0
                                 Specifies  a  display  resolution of at most
                                 800x600, or smaller, in order  to  keep  as‐
                                 pect.

                <r>
                     Rounds up to make both width and height divisible by <r>
                     (default: 1).

      yvu9
             Forces software YVU9 to YV12 colorspace conversion.   Deprecated
             in favor of the software scaler.

      yuvcsp
             Clamps YUV color values to the CCIR 601 range without doing real
             conversion.

      palette
             RGB/BGR 8 -> 15/16/24/32bpp colorspace conversion using palette.

      format[=fourcc[:outfourcc]]
             Restricts the colorspace for the next filter without  doing  any
             conversion.   Use together with the scale filter for a real con‐
             version.
             NOTE: For a list of available formats see format=fmt=help.
                <fourcc>
                     format name like rgb15, bgr24, yv12, etc (default: yuy2)
                <outfourcc>
                     Format name that should be substituted for  the  output.
                     If  this  is not 100% compatible with the <fourcc> value
                     it will crash.
                     Valid examples:
                     format=rgb24:bgr24 format=yuyv:yuy2
                     Invalid examples (will crash):
                     format=rgb24:yv12

      noformat[=fourcc]
             Restricts the colorspace for the next filter without  doing  any
             conversion.   Unlike the format filter, this will allow any col‐
             orspace except the one you specify.
             NOTE: For a list of available formats see noformat=fmt=help.
                <fourcc>
                     format name like rgb15, bgr24, yv12, etc (default: yv12)

      pp[=filter1[:option1[:option2...]]/[-]filter2...] (also see -pphelp)
             Enables the specified chain of postprocessing subfilters.   Sub‐
             filters must be separated by '/' and can be disabled by prepend‐
             ing a '-'.  Each subfilter and some options have a short  and  a
             long  name  that can be used interchangeably, i.e. dr/dering are
             the same.  All subfilters  share  common  options  to  determine
             their scope:
                a/autoq
                     Automatically switch the subfilter off if the CPU is too
                     slow.
                c/chrom
                     Do chrominance filtering, too (default).
                y/nochrom
                     Do luminance filtering only (no chrominance).
                n/noluma
                     Do chrominance filtering only (no luminance).

             NOTE: -pphelp shows a list of available subfilters.

             Available subfilters are

                hb/hdeblock[:difference[:flatness]]
                     horizontal deblocking filter
                        <difference>: Difference factor where  higher  values
                        mean more deblocking (default: 32).
                        <flatness>:  Flatness  threshold  where  lower values
                        mean more deblocking (default: 39).

                vb/vdeblock[:difference[:flatness]]
                     vertical deblocking filter
                        <difference>: Difference factor where  higher  values
                        mean more deblocking (default: 32).
                        <flatness>:  Flatness  threshold  where  lower values
                        mean more deblocking (default: 39).

                ha/hadeblock[:difference[:flatness]]
                     accurate horizontal deblocking filter
                        <difference>: Difference factor where  higher  values
                        mean more deblocking (default: 32).
                        <flatness>:  Flatness  threshold  where  lower values
                        mean more deblocking (default: 39).

                va/vadeblock[:difference[:flatness]]
                     accurate vertical deblocking filter
                        <difference>: Difference factor where  higher  values
                        mean more deblocking (default: 32).
                        <flatness>:  Flatness  threshold  where  lower values
                        mean more deblocking (default: 39).

                The horizontal and vertical deblocking filters share the dif‐
                ference and flatness values so you cannot set different hori‐
                zontal and vertical thresholds.


                h1/x1hdeblock
                     experimental horizontal deblocking filter

                v1/x1vdeblock
                     experimental vertical deblocking filter

                dr/dering
                     deringing filter

                tn/tmpnoise[:threshold1[:threshold2[:threshold3]]]
                     temporal noise reducer
                        <threshold1>: larger -> stronger filtering
                        <threshold2>: larger -> stronger filtering
                        <threshold3>: larger -> stronger filtering

                al/autolevels[:f/fullyrange]
                     automatic brightness / contrast correction
                        f/fullyrange: Stretch luminance to (0-255).

                lb/linblenddeint
                     Linear blend deinterlacing filter that deinterlaces  the
                     given  block  by filtering all lines with a (1 2 1) fil‐
                     ter.

                li/linipoldeint
                     Linear interpolating deinterlacing filter that  deinter‐
                     laces  the  given  block by linearly interpolating every
                     second line.

                ci/cubicipoldeint
                     Cubic interpolating  deinterlacing  filter  deinterlaces
                     the  given block by cubically interpolating every second
                     line.

                md/mediandeint
                     Median deinterlacing filter that deinterlaces the  given
                     block by applying a median filter to every second line.

                fd/ffmpegdeint
                     FFmpeg  deinterlacing filter that deinterlaces the given
                     block by filtering every second line with a (-1  4  2  4
                     -1) filter.

                l5/lowpass5
                     Vertically applied FIR lowpass deinterlacing filter that
                     deinterlaces the given block by filtering all lines with
                     a (-1 2 6 2 -1) filter.

                fq/forceQuant[:quantizer]
                     Overrides  the  quantizer  table from the input with the
                     constant quantizer you specify.
                        <quantizer>: quantizer to use

                de/default
                     default pp filter combination (hb:a,vb:a,dr:a)

                fa/fast
                     fast pp filter combination (h1:a,v1:a,dr:a)

                ac
                     high      quality      pp       filter       combination
                     (ha:a:128:7,va:a,dr:a)

             EXAMPLE:
                -vf pp=hb/vb/dr/al
                     horizontal  and vertical deblocking, deringing and auto‐
                     matic brightness/contrast
                -vf pp=de/-al
                     default filters without brightness/contrast correction
                -vf pp=default/tmpnoise:1:2:3
                     Enable default filters & temporal denoiser.
                -vf pp=hb:y/vb:a
                     Horizontal deblocking on luminance only, and switch ver‐
                     tical  deblocking  on  or off automatically depending on
                     available CPU time.

      spp[=quality[:qp[:mode]]]
             Simple postprocessing filter that  compresses  and  decompresses
             the image at several (or - in the case of quality level 6 - all)
             shifts and averages the results.

                <quality>
                     0-6 (default: 3)

                <qp>
                     Force quantization parameter (default: 0,  use  QP  from
                     video).

                <mode>
                     0: hard thresholding (default)
                     1: soft thresholding (better deringing, but blurrier)
                     4: like 0, but also use B-frames' QP (may cause flicker)
                     5: like 1, but also use B-frames' QP (may cause flicker)

      uspp[=quality[:qp]]
             Ultra  simple  &  slow postprocessing filter that compresses and
             decompresses the image at several (or - in the case  of  quality
             level  8  -  all) shifts and averages the results.  The way this
             differs from the behavior of spp is that uspp actually encodes &
             decodes  each case with libavcodec Snow, whereas spp uses a sim‐
             plified intra only 8x8 DCT similar to MJPEG.

                <quality>
                     0-8 (default: 3)

                <qp>
                     Force quantization parameter (default: 0,  use  QP  from
                     video).

      fspp[=quality[:qp[:strength[:bframes]]]]
             faster version of the simple postprocessing filter

                <quality>
                     4-5 (equivalent to spp; default: 4)

                <qp>
                     Force  quantization  parameter  (default: 0, use QP from
                     video).

                <-15-32>
                     Filter strength, lower values mean more details but also
                     more  artifacts,  while  higher  values  make  the image
                     smoother but also blurrier (default: 0 - PSNR optimal).

                <bframes>
                     0: do not use QP from B-frames (default)
                     1: use QP from B-frames too (may cause flicker)

      pp7[=qp[:mode]]
             Variant of the spp filter, similar to spp=6  with  7  point  DCT
             where only the center sample is used after IDCT.

                <qp>
                     Force  quantization  parameter  (default: 0, use QP from
                     video).

                <mode>
                     0: hard thresholding
                     1: soft thresholding (better deringing, but blurrier)
                     2: medium thresholding (default, good results)

      qp=equation
             quantization parameter (QP) change filter

                <equation>
                     some equation like "2+2*sin(PI*qp)"

      geq=equation
             generic equation change filter

                <equation>
                     Some equation, e.g.  'p(W-X\,Y)' to flip the image hori‐
                     zontally.   You  can use whitespace to make the equation
                     more readable.  There are a couple of constants that can
                     be used in the equation:
                        PI: the number pi
                        E: the number e
                        X / Y: the coordinates of the current sample
                        W / H: width and height of the image
                        SW / SH: width/height scale depending on the current‐
                        ly filtered plane,  e.g.  1,1  and  0.5,0.5  for  YUV
                        4:2:0.
                        p(x,y):  returns  the  value of the pixel at location
                        x/y of the current plane.

      test
             Generate various test patterns.

      rgbtest[=width:height]
             Generate an RGB test pattern useful for detecting RGB vs BGR is‐
             sues.   You  should see a red, green and blue stripe from top to
             bottom.

                <width>
                     Desired width of generated image (default: 0).  0  means
                     width of input image.

                <height>
                     Desired height of generated image (default: 0).  0 means
                     height of input image.

      lavc[=quality:fps]
             Fast software YV12 to MPEG-1 conversion with libavcodec for  use
             with DVB/DXR3/IVTV/V4L2.

                <quality>
                     1-31: fixed qscale
                     32-:  fixed bitrate in kbits

                <fps>
                     force  output  fps (float value) (default: 0, autodetect
                     based on height)

      dvbscale[=aspect]
             Set up optimal scaling for DVB cards,  scaling  the  x  axis  in
             hardware  and calculating the y axis scaling in software to keep
             aspect.  Only useful together with expand and scale.

                <aspect>
                     Control aspect ratio, calculate as  DVB_HEIGHT*ASPECTRA‐
                     TIO  (default:  576*4/3=768),  set it to 576*(16/9)=1024
                     for a 16:9 TV.

             EXAMPLE:
                -vf dvbscale,scale=-1:0,expand=-1:576:-1:-1:1,lavc
                     FIXME: Explain what this does.

      noise[=luma[u][t|a][h][p]:chroma[u][t|a][h][p]]
             Adds noise.
                <0-100>
                     luma noise
                <0-100>
                     chroma noise
                u    uniform noise (gaussian otherwise)
                t    temporal noise (noise pattern changes between frames)
                a    averaged temporal noise (smoother, but a lot slower)
                h    high quality (slightly better looking, slightly slower)
                p    mix random noise with a (semi)regular pattern

      denoise3d[=luma_spatial:chroma_spatial:luma_tmp:chroma_tmp]
             This filter aims to reduce image noise producing  smooth  images
             and  making  still images really still (This should enhance com‐
             pressibility.).
                <luma_spatial>
                     spatial luma strength (default: 4)
                <chroma_spatial>
                     spatial chroma strength (default: 3)
                <luma_tmp>
                     luma temporal strength (default: 6)
                <chroma_tmp>
                     chroma temporal strength (default:  luma_tmp*chroma_spa‐
                     tial/luma_spatial)

      hqdn3d[=luma_spatial:chroma_spatial:luma_tmp:chroma_tmp]
             High precision/quality version of the denoise3d filter.  Parame‐
             ters and usage are the same.

      ow[=depth[:luma_strength[:chroma_strength]]]
             Overcomplete Wavelet denoiser.
                <depth>
                     Larger depth values will denoise lower frequency  compo‐
                     nents more, but slow down filtering (default: 8).
                <luma_strength>
                     luma strength (default: 1.0)
                <chroma_strength>
                     chroma strength (default: 1.0)

      eq[=brightness:contrast] (OBSOLETE)
             Software equalizer with interactive controls just like the hard‐
             ware equalizer, for cards/drivers that do not support brightness
             and  contrast  controls  in hardware.  Might also be useful with
             MEncoder, either for  fixing  poorly  captured  movies,  or  for
             slightly  reducing  contrast  to  mask artifacts and get by with
             lower bitrates.
                <-100-100>
                     initial brightness
                <-100-100>
                     initial contrast

      eq2[=gamma:contrast:brightness:saturation:rg:gg:bg:weight]
             Alternative software equalizer that  uses  lookup  tables  (very
             slow),  allowing  gamma correction in addition to simple bright‐
             ness and contrast adjustment.  Note that it uses  the  same  MMX
             optimized  code  as -vf eq if all gamma values are 1.0.  The pa‐
             rameters are given as floating point values.
                <0.1-10>
                     initial gamma value (default: 1.0)
                <-2-2>
                     initial contrast, where negative values result in a neg‐
                     ative image (default: 1.0)
                <-1-1>
                     initial brightness (default: 0.0)
                <0-3>
                     initial saturation (default: 1.0)
                <0.1-10>
                     gamma value for the red component (default: 1.0)
                <0.1-10>
                     gamma value for the green component (default: 1.0)
                <0.1-10>
                     gamma value for the blue component (default: 1.0)
                <0-1>
                     The weight parameter can be used to reduce the effect of
                     a high gamma value on bright image areas, e.g. keep them
                     from getting overamplified and just plain white.  A val‐
                     ue of 0.0 turns the gamma correction all  the  way  down
                     while 1.0 leaves it at its full strength (default: 1.0).

      hue[=hue:saturation]
             Software equalizer with interactive controls just like the hard‐
             ware equalizer, for cards/drivers that do not  support  hue  and
             saturation controls in hardware.
                <-180-180>
                     initial hue (default: 0.0)
                <-100-100>
                     initial  saturation,  where  negative values result in a
                     negative chroma (default: 1.0)

      halfpack[=f]
             Convert planar YUV 4:2:0 to half-height packed  4:2:2,  downsam‐
             pling luma but keeping all chroma samples.  Useful for output to
             low-resolution display devices when hardware downscaling is poor
             quality  or  is  not available.  Can also be used as a primitive
             luma-only deinterlacer with very low CPU usage.
                <f>
                     By default, halfpack averages pairs of lines when  down‐
                     sampling.  Any value different from 0 or 1 gives the de‐
                     fault (averaging) behavior.
                        0: Only use even lines when downsampling.
                        1: Only use odd lines when downsampling.

      ilpack[=mode]
             When interlaced video is stored in YUV 4:2:0 formats, chroma in‐
             terlacing does not line up properly due to vertical downsampling
             of the chroma channels.  This filter packs the planar 4:2:0 data
             into  YUY2  (4:2:2) format with the chroma lines in their proper
             locations, so that in any given scanline, the  luma  and  chroma
             data both come from the same field.
                <mode>
                     Select the sampling mode.
                        0: nearest-neighbor sampling, fast but incorrect
                        1: linear interpolation (default)

      harddup
             Only useful with MEncoder.  If harddup is used when encoding, it
             will force duplicate frames to be encoded in the  output.   This
             uses  slightly  more  space, but is necessary for output to MPEG
             files or if you plan to demux and remux the video  stream  after
             encoding.   Should  be  placed  at or near the end of the filter
             chain unless you have a good reason to do otherwise.

      softskip
             Only useful with MEncoder.  Softskip moves  the  frame  skipping
             (dropping) step of encoding from before the filter chain to some
             point during the filter chain.  This allows filters  which  need
             to  see  all frames (inverse telecine, temporal denoising, etc.)
             to function properly.  Should be placed after the filters  which
             need  to  see  all frames and before any subsequent filters that
             are CPU-intensive.

      decimate[=max:hi:lo:frac]
             Drops frames that do not differ greatly from the previous  frame
             in  order  to  reduce framerate.  The main use of this filter is
             for very-low-bitrate encoding (e.g. streaming  over  dialup  mo‐
             dem), but it could in theory be used for fixing movies that were
             inverse-telecined incorrectly.
                <max>
                     Sets the maximum number of consecutive frames which  can
                     be  dropped  (if  positive), or the minimum interval be‐
                     tween dropped frames (if negative).
                <hi>,<lo>,<frac>
                     A frame is a candidate for dropping  if  no  8x8  region
                     differs  by  more  than  a threshold of <hi>, and if not
                     more than <frac> portion (1  meaning  the  whole  image)
                     differs  by  more  than  a threshold of <lo>.  Values of
                     <hi> and <lo> are for 8x8 pixel blocks and represent ac‐
                     tual  pixel value differences, so a threshold of 64 cor‐
                     responds to 1 unit of difference for each pixel, or  the
                     same spread out differently over the block.

      dint[=sense:level]
             The  drop-deinterlace  (dint) filter detects and drops the first
             from a set of interlaced video frames.
                <0.0-1.0>
                     relative difference between neighboring pixels (default:
                     0.1)
                <0.0-1.0>
                     What  part of the image has to be detected as interlaced
                     to drop the frame (default: 0.15).

      lavcdeint (OBSOLETE)
             FFmpeg deinterlacing filter, same as -vf pp=fd

      lavfi=filtergraph
             FFmpeg libavfilter wrapper.  filtergraph defines a whole  libav‐
             filter graph with one input and one output.  See http://www.ffm‐
             peg.org/libavfilter.html#SEC4 for details.

             As a special case, if filtergraph is $word then the value of the
             word  environment  variable is used; this is necessary if commas
             are present in the graph description, as mplayer uses them as  a
             delimiter between filters.

             NOTE:  This  filter  is considered experimental, it may interact
             strangely with other filters.

             EXAMPLE:
             overlay="movie=$small_video, scale=160:120 [ca]; [in] [ca] over‐
             lay=16:8" mplayer -vf lavfi='$overlay' $big_video

      kerndeint[=thresh[:map[:order[:sharp[:twoway]]]]]
             Donald Graft's adaptive kernel deinterlacer.  Deinterlaces parts
             of a video if a configurable threshold is exceeded.
                <0-255>
                     threshold (default: 10)
                <map>
                        0: Ignore pixels exceeding the threshold (default).
                        1: Paint pixels exceeding the threshold white.

                <order>
                        0: Leave fields alone (default).
                        1: Swap fields.

                <sharp>
                        0: Disable additional sharpening (default).
                        1: Enable additional sharpening.

                <twoway>
                        0: Disable twoway sharpening (default).
                        1: Enable twoway sharpening.

      unsharp[=l|cWxH:amount[:l|cWxH:amount]]
             unsharp mask / gaussian blur

                l
                     Apply effect on luma component.

                c
                     Apply effect on chroma components.

                <width>x<height>
                     width and height of the matrix, odd sized in both direc‐
                     tions  (min  =  3x3, max = 13x11 or 11x13, usually some‐
                     thing between 3x3 and 7x7)

                amount
                     Relative amount of sharpness/blur to add to the image (a
                     sane range should be -1.5-1.5).
                        <0: blur
                        >0: sharpen

      swapuv
             Swap U & V plane.

      il[=d|i][s][:[d|i][s]]
             (De)interleaves  lines.   The  goal of this filter is to add the
             ability to process interlaced images pre-field without  deinter‐
             lacing  them.  You can filter your interlaced DVD and play it on
             a TV without  breaking  the  interlacing.   While  deinterlacing
             (with the postprocessing filter) removes interlacing permanently
             (by smoothing, averaging, etc) deinterleaving splits  the  frame
             into  2  fields  (so  called  half pictures), so you can process
             (filter) them independently and then re-interleave them.
                d    deinterleave (placing one above the other)
                i    interleave
                s    swap fields (exchange even & odd lines)

      fil[=i|d]
             (De)interleaves lines.  This filter is very similar  to  the  il
             filter  but  much  faster, the main disadvantage is that it does
             not always work.  Especially if combined with other  filters  it
             may  produce  randomly messed up images, so be happy if it works
             but do not complain if it does not for your combination of  fil‐
             ters.
                d    Deinterleave fields, placing them side by side.
                i    Interleave fields again (reversing the effect of fil=d).

      field[=n]
             Extracts  a  single  field from an interlaced image using stride
             arithmetic to avoid wasting CPU time.  The optional  argument  n
             specifies  whether to extract the even or the odd field (depend‐
             ing on whether n is even or odd).

      detc[=var1=value1:var2=value2:...]
             Attempts to reverse the 'telecine' process to recover  a  clean,
             non-interlaced stream at film framerate.  This was the first and
             most primitive inverse telecine filter to be added  to  MPlayer/
             MEncoder.   It  works  by latching onto the telecine 3:2 pattern
             and following it as long as possible.  This  makes  it  suitable
             for perfectly-telecined material, even in the presence of a fair
             degree of noise, but it will fail in  the  presence  of  complex
             post-telecine  edits.   Development  on this filter is no longer
             taking place, as ivtc, pullup, and filmdint are better for  most
             applications.  The following arguments (see syntax above) may be
             used to control detc's behavior:

                <dr>
                     Set the frame dropping mode.
                        0: Do not drop frames to maintain fixed output  fram‐
                        erate (default).
                        1:  Always drop a frame when there have been no drops
                        or telecine merges in the past 5 frames.
                        2: Always maintain exact 5:4 input  to  output  frame
                        ratio.
                        NOTE: Use mode 1 or 2 with MEncoder.

                <am>
                     Analysis mode.
                        0:  Fixed pattern with initial frame number specified
                        by <fr>.
                        1: aggressive search for telecine pattern (default)

                <fr>
                     Set initial frame number in sequence.  0-2 are the three
                     clean progressive frames; 3 and 4 are the two interlaced
                     frames.  The default, -1, means  'not  in  telecine  se‐
                     quence'.   The number specified here is the type for the
                     imaginary previous frame before the movie starts.

                <t0>, <t1>, <t2>, <t3>
                     Threshold values to be used in certain modes.

      ivtc[=1]
             Experimental 'stateless' inverse telecine filter.   Rather  than
             trying  to  lock on to a pattern like the detc filter does, ivtc
             makes its decisions independently for  each  frame.   This  will
             give  much  better results for material that has undergone heavy
             editing after telecine was applied, but as a result it is not as
             forgiving  of noisy input, for example TV capture.  The optional
             parameter (ivtc=1) corresponds to the dr=1 option for  the  detc
             filter,  and  should be used with MEncoder but not with MPlayer.
             As with detc, you must  specify  the  correct  output  framerate
             (-ofps  24000/1001) when using MEncoder.  Further development on
             ivtc has stopped, as the pullup and filmdint filters  appear  to
             be much more accurate.

      pullup[=jl:jr:jt:jb:sb:mp]
             Third-generation  pulldown  reversal  (inverse telecine) filter,
             capable of handling mixed hard-telecine, 24000/1001 fps progres‐
             sive, and 30000/1001 fps progressive content.  The pullup filter
             is designed to be much more robust than detc or ivtc, by  taking
             advantage of future context in making its decisions.  Like ivtc,
             pullup is stateless in the sense that it does not  lock  onto  a
             pattern to follow, but it instead looks forward to the following
             fields in order to  identify  matches  and  rebuild  progressive
             frames.  It is still under development, but believed to be quite
             accurate.

                jl, jr, jt, and jb
                     These options set the amount of "junk" to ignore at  the
                     left, right, top, and bottom of the image, respectively.
                     Left/right are in units of 8  pixels,  while  top/bottom
                     are  in  units  of  2 lines.  The default is 8 pixels on
                     each side.

                sb (strict breaks)
                     Setting this option to 1  will  reduce  the  chances  of
                     pullup generating an occasional mismatched frame, but it
                     may also cause an  excessive  number  of  frames  to  be
                     dropped  during high motion sequences.  Conversely, set‐
                     ting it to -1 will make pullup match fields more easily.
                     This  may help processing of video where there is slight
                     blurring between the fields, but may also cause there to
                     be interlaced frames in the output.

                mp (metric plane)
                     This  option  may be set to 1 or 2 to use a chroma plane
                     instead of the luma plane for  doing  pullup's  computa‐
                     tions.   This  may improve accuracy on very clean source
                     material, but more likely will decrease accuracy,  espe‐
                     cially  if there is chroma noise (rainbow effect) or any
                     grayscale video.  The main purpose of setting  mp  to  a
                     chroma  plane  is to reduce CPU load and make pullup us‐
                     able in realtime on slow machines.

             NOTE: Always follow pullup with the softskip filter when  encod‐
             ing to ensure that pullup is able to see each frame.  Failure to
             do so will lead to incorrect output and will usually crash,  due
             to design limitations in the codec/filter layer.

      filmdint[=options]
             Inverse telecine filter, similar to the pullup filter above.  It
             is designed to handle any pulldown pattern, including mixed soft
             and hard telecine and limited support for movies that are slowed
             down or sped up from their original framerate for TV.  Only  the
             luma  plane is used to find the frame breaks.  If a field has no
             match, it is deinterlaced with simple linear approximation.   If
             the source is MPEG-2, this must be the first filter to allow ac‐
             cess to the field-flags set by the MPEG-2 decoder.  Depending on
             the  source  MPEG, you may be fine ignoring this advice, as long
             as you do not see lots of "Bottom-first field"  warnings.   With
             no  options  it does normal inverse telecine, and should be used
             together with mencoder -fps 30000/1001 -ofps  24000/1001.   When
             this  filter  is  used with MPlayer, it will result in an uneven
             framerate during playback, but it is still generally better than
             using pp=lb or no deinterlacing at all.  Multiple options can be
             specified separated by /.

                crop=<w>:<h>:<x>:<y>
                     Just like the crop filter,  but  faster,  and  works  on
                     mixed  hard and soft telecined content as well as when y
                     is not a multiple of 4.  If x or y would  require  crop‐
                     ping  fractional pixels from the chroma planes, the crop
                     area is extended.  This usually means that x and y  must
                     be even.

                io=<ifps>:<ofps>
                     For  each  ifps input frames the filter will output ofps
                     frames.   The  ratio  of  ifps/ofps  should  match   the
                     -fps/-ofps  ratio.   This could be used to filter movies
                     that are broadcast on TV at a frame rate different  from
                     their original framerate.

                luma_only=<n>
                     If  n  is nonzero, the chroma plane is copied unchanged.
                     This is useful for YV12 sampled TV, which  discards  one
                     of the chroma fields.

                mmx2=<n>
                     On  x86,  if  n=1, use MMX2 optimized functions, if n=2,
                     use 3DNow!  optimized functions, otherwise, use plain C.
                     If this option is not specified, MMX2 and 3DNow! are au‐
                     to-detected, use this option to override auto-detection.

                fast=<n>
                     The larger n will speed up the filter at the expense  of
                     accuracy.   The  default  value  is n=3.  If n is odd, a
                     frame immediately following a frame marked with the  RE‐
                     PEAT_FIRST_FIELD MPEG flag is assumed to be progressive,
                     thus filter will not spend any  time  on  soft-telecined
                     MPEG-2 content.  This is the only effect of this flag if
                     MMX2 or 3DNow! is available.  Without MMX2 and 3DNow, if
                     n=0 or 1, the same calculations will be used as with n=2
                     or 3.  If n=2 or 3, the number of luma  levels  used  to
                     find  the frame breaks is reduced from 256 to 128, which
                     results in a faster filter without losing much accuracy.
                     If  n=4  or  5,  a faster, but much less accurate metric
                     will be used to find the frame  breaks,  which  is  more
                     likely  to  misdetect high vertical detail as interlaced
                     content.

                verbose=<n>
                     If n is nonzero, print the  detailed  metrics  for  each
                     frame.  Useful for debugging.

                dint_thres=<n>
                     Deinterlace  threshold.   Used  during de-interlacing of
                     unmatched frames.  Larger value means  less  deinterlac‐
                     ing,  use  n=256  to  completely turn off deinterlacing.
                     Default is n=8.

                comb_thres=<n>
                     Threshold for comparing a top and  bottom  fields.   De‐
                     faults to 128.

                diff_thres=<n>
                     Threshold to detect temporal change of a field.  Default
                     is 128.

                sad_thres=<n>
                     Sum of Absolute Difference threshold, default is 64.

      softpulldown
             This filter works only correct with MEncoder  and  acts  on  the
             MPEG-2 flags used for soft 3:2 pulldown (soft telecine).  If you
             want to use the ivtc or detc filter on movies  that  are  partly
             soft  telecined,  inserting  this filter before them should make
             them more reliable.

      divtc[=options]
             Inverse  telecine  for  deinterlaced  video.   If   3:2-pulldown
             telecined  video  has  lost one of the fields or is deinterlaced
             using a method that keeps one field and interpolates the  other,
             the  result is a juddering video that has every fourth frame du‐
             plicated.  This filter is intended to find and drop those dupli‐
             cates  and restore the original film framerate.  When using this
             filter, you must specify -ofps that is 4/5 of the fps of the in‐
             put  file  and  place  the softskip later in the filter chain to
             make sure that divtc sees all the frames.  Two  different  modes
             are  available: One pass mode is the default and is straightfor‐
             ward to use, but has the disadvantage that any  changes  in  the
             telecine phase (lost frames or bad edits) cause momentary judder
             until the filter can resync again.  Two pass mode avoids this by
             analyzing  the  whole  video  beforehand so it will have forward
             knowledge about the phase changes and can resync  at  the  exact
             spot.  These passes do not correspond to pass one and two of the
             encoding process.  You must run an extra pass using  divtc  pass
             one  before  the  actual  encoding  throwing the resulting video
             away.  Use -nosound -ovc raw -o /dev/null to avoid  wasting  CPU
             power  for  this  pass.  You may add something like crop=2:2:0:0
             after divtc to speed things up even more.  Then use  divtc  pass
             two  for the actual encoding.  If you use multiple encoder pass‐
             es, use divtc pass two for all of them.  The options are:

                pass=1|2
                     Use two pass mode.

                file=<filename>
                     Set  the  two  pass  log  filename  (default:   "framed‐
                     iff.log").

                threshold=<value>
                     Set  the minimum strength the telecine pattern must have
                     for the filter to believe in it (default: 0.5).  This is
                     used  to  avoid recognizing false pattern from the parts
                     of the video that are very dark or very still.

                window=<numframes>
                     Set the number of past frames to look at when  searching
                     for  pattern  (default: 30).  Longer window improves the
                     reliability of the pattern search,  but  shorter  window
                     improves  the  reaction  time  to  the  changes  in  the
                     telecine phase.  This only affects the  one  pass  mode.
                     The  two  pass mode currently uses fixed window that ex‐
                     tends to both future and past.

                phase=0|1|2|3|4
                     Sets the initial telecine phase for one pass  mode  (de‐
                     fault:  0).  The two pass mode can see the future, so it
                     is able to use the correct phase from the beginning, but
                     one  pass  mode  can only guess.  It catches the correct
                     phase when it finds it, but this option can be  used  to
                     fix  the possible juddering at the beginning.  The first
                     pass of the two pass mode also uses this, so if you save
                     the  output  from the first pass, you get constant phase
                     result.

                deghost=<value>
                     Set the deghosting threshold (0-255 for one  pass  mode,
                     -255-255  for  two  pass  mode, default 0).  If nonzero,
                     deghosting mode is used.  This is  for  video  that  has
                     been  deinterlaced  by  blending the fields together in‐
                     stead of dropping one of the fields.  Deghosting  ampli‐
                     fies any compression artifacts in the blended frames, so
                     the parameter value is used as a  threshold  to  exclude
                     those pixels from deghosting that differ from the previ‐
                     ous frame less than specified value.  If two  pass  mode
                     is  used,  then  negative  value can be used to make the
                     filter analyze the  whole  video  in  the  beginning  of
                     pass-2  to  determine whether it needs deghosting or not
                     and then select either zero or the absolute value of the
                     parameter.   Specify this option for pass-2, it makes no
                     difference on pass-1.

      phase[=t|b|p|a|u|T|B|A|U][:v]
             Delay interlaced video by one field time so that the field order
             changes.   The  intended use is to fix PAL movies that have been
             captured with the opposite  field  order  to  the  film-to-video
             transfer.  The options are:

                t    Capture  field  order  top-first, transfer bottom-first.
                     Filter will delay the bottom field.

                b    Capture bottom-first, transfer top-first.   Filter  will
                     delay the top field.

                p    Capture  and  transfer  with the same field order.  This
                     mode only exists for the documentation of the other  op‐
                     tions  to  refer  to, but if you actually select it, the
                     filter will faithfully do nothing ;-)

                a    Capture field order determined  automatically  by  field
                     flags,  transfer opposite.  Filter selects among t and b
                     modes on a frame by frame basis using field  flags.   If
                     no  field information is available, then this works just
                     like u.

                u    Capture unknown or varying, transfer  opposite.   Filter
                     selects  among t and b on a frame by frame basis by ana‐
                     lyzing the images and  selecting  the  alternative  that
                     produces best match between the fields.

                T    Capture  top-first, transfer unknown or varying.  Filter
                     selects among t and p using image analysis.

                B    Capture bottom-first, transfer unknown or varying.  Fil‐
                     ter selects among b and p using image analysis.

                A    Capture  determined  by field flags, transfer unknown or
                     varying.  Filter selects among t, b and  p  using  field
                     flags  and  image  analysis.  If no field information is
                     available, then this works just like U.  This is the de‐
                     fault mode.

                U    Both  capture  and  transfer unknown or varying.  Filter
                     selects among t, b and p using image analysis only.

                v    Verbose operation.  Prints the selected  mode  for  each
                     frame  and the average squared difference between fields
                     for t, b, and p alternatives.

      telecine[=start]
             Apply 3:2 'telecine' process to increase framerate by 20%.  This
             most  likely will not work correctly with MPlayer, but it can be
             used  with  'mencoder  -fps  30000/1001  -ofps  30000/1001   -vf
             telecine'.   Both  fps  options  are  essential!  (A/V sync will
             break if they are wrong.)  The optional  start  parameter  tells
             the filter where in the telecine pattern to start (0-3).

      tinterlace[=mode]
             Temporal  field  interlacing - merge pairs of frames into an in‐
             terlaced frame, halving the framerate.  Even  frames  are  moved
             into  the  upper field, odd frames to the lower field.  This can
             be used to fully reverse the effect of the  tfields  filter  (in
             mode 0).  Available modes are:
                0    Move odd frames into the upper field, even into the low‐
                     er field, generating a full-height frame at half framer‐
                     ate.
                1    Only  output odd frames, even frames are dropped; height
                     unchanged.
                2    Only output even frames, odd frames are dropped;  height
                     unchanged.
                3    Expand  each  frame  to  full  height, but pad alternate
                     lines with black; framerate unchanged.
                4    Interleave even lines from even frames  with  odd  lines
                     from odd frames.  Height unchanged at half framerate.

      tfields[=mode[:field_dominance]]
             Temporal  field  separation - split fields into frames, doubling
             the output framerate.  Like the telecine filter,  tfields  might
             not  work  completely  right  unless used with MEncoder and both
             -fps and -ofps set to the desired (double) framerate!
                <mode>
                     0: Leave fields unchanged (will jump/flicker).
                     1: Interpolate missing lines. (The algorithm used  might
                     not be so good.)
                     2:  Translate fields by 1/4 pixel with linear interpola‐
                     tion (no jump).
                     4: Translate fields by 1/4 pixel with 4tap filter (high‐
                     er quality) (default).
                <field_dominance> (DEPRECATED)
                     -1: auto (default) Only works if the decoder exports the
                     appropriate information and no other filters which  dis‐
                     card  that information come before tfields in the filter
                     chain, otherwise it falls back to 0 (top field first).
                     0: top field first
                     1: bottom field first
                     NOTE: This option will possibly be removed in  a  future
                     version.  Use -field-dominance instead.

      yadif=[mode[:field_dominance]]
             Yet another deinterlacing filter
                <mode>
                     0: Output 1 frame for each frame.
                     1: Output 1 frame for each field.
                     2: Like 0 but skips spatial interlacing check.
                     3: Like 1 but skips spatial interlacing check.
                <field_dominance> (DEPRECATED)
                     Operates like tfields.
                     NOTE:  This  option will possibly be removed in a future
                     version.  Use -field-dominance instead.

      mcdeint=[mode[:parity[:qp]]]
             Motion compensating deinterlacer.  It needs one field per  frame
             as  input  and  must  thus  be  used  together with tfields=1 or
             yadif=1/3 or equivalent.
                <mode>
                     0: fast
                     1: medium
                     2: slow, iterative motion estimation
                     3: extra slow, like 2 plus multiple reference frames
                <parity>
                     0 or 1 selects which field to use (note:  no  autodetec‐
                     tion yet!).
                <qp>
                     Higher  values should result in a smoother motion vector
                     field but less optimal individual vectors.

      boxblur=radius:power[:radius:power]
             box blur
                <radius>
                     blur filter strength
                <power>
                     number of filter applications

      sab=radius:pf:colorDiff[:radius:pf:colorDiff]
             shape adaptive blur
                <radius>
                     blur filter strength (~0.1-4.0) (slower if larger)
                <pf>
                     prefilter strength (~0.1-2.0)
                <colorDiff>
                     maximum difference between pixels to still be considered
                     (~0.1-100.0)

      smartblur=radius:strength:threshold[:radius:strength:threshold]
             smart blur
                <radius>
                     blur filter strength (~0.1-5.0) (slower if larger)
                <strength>
                     blur (0.0-1.0) or sharpen (-1.0-0.0)
                <threshold>
                     filter all (0), filter flat areas (0-30) or filter edges
                     (-30-0)

      perspective=x0:y0:x1:y1:x2:y2:x3:y3:t
             Correct the perspective of movies not  filmed  perpendicular  to
             the screen.
                <x0>,<y0>,...
                     coordinates  of  the  top  left, top right, bottom left,
                     bottom right corners
                <t>
                     linear (0) or cubic resampling (1)

      2xsai
             Scale and smooth the image with the 2x scale and interpolate al‐
             gorithm.

      1bpp
             1bpp bitmap to YUV/BGR 8/15/16/32 conversion

      down3dright[=lines]
             Reposition and resize stereoscopic images.  Extracts both stereo
             fields and places them side by side, resizing them  to  maintain
             the original movie aspect.
                <lines>
                     number  of  lines to select from the middle of the image
                     (default: 12)

      bmovl=hidden:opaque:fifo
             The bitmap overlay filter reads bitmaps from a FIFO and displays
             them  on  top of the movie, allowing some transformations on the
             image.  Also see TOOLS/bmovl-test.c for a small bmovl test  pro‐
             gram.
                <hidden>
                     Set  the  default value of the 'hidden' flag (0=visible,
                     1=hidden).
                <opaque>
                     Set the default value of the 'opaque' flag  (0=transpar‐
                     ent, 1=opaque).
                <fifo>
                     path/filename   for  the  FIFO  (named  pipe  connecting
                     'mplayer -vf bmovl' to the controlling application)

             FIFO commands are:
                RGBA32 width height xpos ypos alpha clear
                     followed by width*height*4 Bytes of raw RGBA32 data.
                ABGR32 width height xpos ypos alpha clear
                     followed by width*height*4 Bytes of raw ABGR32 data.
                RGB24 width height xpos ypos alpha clear
                     followed by width*height*3 Bytes of raw RGB24 data.
                BGR24 width height xpos ypos alpha clear
                     followed by width*height*3 Bytes of raw BGR24 data.
                ALPHA width height xpos ypos alpha
                     Change alpha transparency of the specified area.
                CLEAR width height xpos ypos
                     Clear area.
                OPAQUE
                     Disable all alpha transparency.  Send "ALPHA 0 0 0 0  0"
                     to enable it again.
                HIDE
                     Hide bitmap.
                SHOW
                     Show bitmap.

             Arguments are:
                <width>, <height>
                     image/area size
                <xpos>, <ypos>
                     Start blitting at position x/y.
                <alpha>
                     Set  alpha  difference.  If you set this to -255 you can
                     then send a sequence of ALPHA-commands to set  the  area
                     to -225, -200, -175 etc for a nice fade-in-effect! ;)
                        0:    same as original
                        255:  Make everything opaque.
                        -255: Make everything transparent.

                <clear>
                     Clear the framebuffer before blitting.
                        0:  The  image will just be blitted on top of the old
                        one, so you do not need to send 1.8MB of RGBA32  data
                        every time a small part of the screen is updated.
                        1: clear

      framestep=I|[i]step
             Renders only every nth frame or every intra frame (keyframe).

             If you call the filter with I (uppercase) as the parameter, then
             only keyframes are rendered.  For DVDs it generally means one in
             every  15/12  frames  (IBBPBBPBBPBBPBB),  for AVI it means every
             scene change or every keyint value (see -lavcopts keyint=  value
             if you use MEncoder to encode the video).

             When  a  keyframe is found, an 'I!' string followed by a newline
             character is printed, leaving the current line  of  MPlayer/MEn‐
             coder  output  on  the  screen, because it contains the time (in
             seconds) and frame number of the keyframe (You can use this  in‐
             formation to split the AVI.).

             If you call the filter with a numeric parameter 'step' then only
             one in every 'step' frames is rendered.

             If you put an 'i' (lowercase) before the number then an 'I!'  is
             printed (like the I parameter).

             If  you give only the i then nothing is done to the frames, only
             I! is printed.

      tile=xtiles:ytiles:output:start:delta
             Tile a series of images into a single,  bigger  image.   If  you
             omit  a  parameter  or use a value less than 0, then the default
             value is used.  You can also stop when you  are  satisfied  (...
             -vf tile=10:5 ...).  It is probably a good idea to put the scale
             filter before the tile :-)

             The parameters are:

                <xtiles>
                     number of tiles on the x axis (default: 5)
                <ytiles>
                     number of tiles on the y axis (default: 5)
                <output>
                     Render the tile  when  'output'  number  of  frames  are
                     reached,  where  'output'  should  be a number less than
                     xtile * ytile.   Missing  tiles  are  left  blank.   You
                     could,  for example, write an 8 * 7 tile every 50 frames
                     to have one image every 2 seconds @ 25 fps.
                <start>
                     outer border thickness in pixels (default: 2)
                <delta>
                     inner border thickness in pixels (default: 4)

      delogo[=x:y:w:h:t]
             Suppresses a TV station logo by a simple  interpolation  of  the
             surrounding  pixels.  Just set a rectangle covering the logo and
             watch it disappear (and sometimes something even uglier appear -
             your mileage may vary).
                <x>,<y>
                     top left corner of the logo
                <w>,<h>
                     width and height of the cleared rectangle
                <t>  Thickness of the fuzzy edge of the rectangle (added to w
                     and h).  When set to -1, a green rectangle is  drawn  on
                     the screen to simplify finding the right x,y,w,h parame‐
                     ters.
                file=<file>
                     You can specify a text  file  to  load  the  coordinates
                     from.   Each line must have a timestamp (in seconds, and
                     in ascending order) and the "x:y:w:h:t"  coordinates  (t
                     can be omitted).

      remove-logo=/path/to/logo_bitmap_file_name.pgm
             Suppresses  a  TV station logo, using a PGM or PPM image file to
             determine which pixels comprise the logo.  The width and  height
             of  the  image  file  must match those of the video stream being
             processed.  Uses the filter image and a circular blur  algorithm
             to remove the logo.

                /path/to/logo_bitmap_file_name.pgm
                     [path] + filename of the filter image.

      zrmjpeg[=options]
             Software YV12 to MJPEG encoder for use with the zr2 video output
             device.

                maxheight=<h>|maxwidth=<w>
                     These options set the maximum width and  height  the  zr
                     card can handle (the MPlayer filter layer currently can‐
                     not query those).

                {dc10+,dc10,buz,lml33}-{PAL|NTSC}
                     Use these options to set maxwidth and maxheight automat‐
                     ically to the values known for card/mode combo.  For ex‐
                     ample, valid options are:  dc10-PAL  and  buz-NTSC  (de‐
                     fault: dc10+PAL)

                color|bw
                     Select  color  or  black  and white encoding.  Black and
                     white encoding is faster.  Color is the default.

                hdec={1,2,4}
                     Horizontal decimation 1, 2 or 4.

                vdec={1,2,4}
                     Vertical decimation 1, 2 or 4.

                quality=1-20
                     Set JPEG compression quality [BEST] 1 - 20 [VERY BAD].

                fd|nofd
                     By default, decimation is only performed  if  the  Zoran
                     hardware  can  upscale the resulting MJPEG images to the
                     original size.  The option fd instructs  the  filter  to
                     always perform the requested decimation (ugly).

      screenshot=prefix
             Allows  acquiring screenshots of the movie using slave mode com‐
             mands that can be bound to keypresses.  See the slave mode docu‐
             mentation  and  the INTERACTIVE CONTROL section for details.  By
             default files named 'shotNNNN.png' will be saved in the  working
             directory,  using  the first available number - no files will be
             overwritten.  Specify a prefix to change the name  or  location,
             e.g.  -vf screenshot=shots/now will save the files in the direc‐
             tory shots with nowNNNN.png as name.  The filter has no overhead
             when not used and accepts an arbitrary colorspace, so it is safe
             to add it to the configuration file.  Make sure that the screen‐
             shot  filter  is  added after all other filters whose effect you
             want to record on the saved image.  E.g. it should be  the  last
             filter  if  you want to have an exact screenshot of what you see
             on the monitor.

      ass
             Moves SSA/ASS subtitle rendering to an arbitrary  point  in  the
             filter chain.  Only useful with the -ass option.

             EXAMPLE:
                -vf ass,screenshot
                     Moves  SSA/ASS  rendering  before the screenshot filter.
                     Screenshots taken this way will contain subtitles.

      blackframe[=amount:threshold]
             Detect frames that are (almost) completely black.  Can be useful
             to detect chapter transitions or commercials.  Output lines con‐
             sist of the frame number of the detected frame,  the  percentage
             of  blackness,  the  frame type and the frame number of the last
             encountered keyframe.

                <amount>
                     Percentage of the pixels  that  have  to  be  below  the
                     threshold (default: 98).

                <threshold>
                     Threshold  below which a pixel value is considered black
                     (default: 32).

      stereo3d[=in:out]
             Stereo3d converts between different stereoscopic image formats.

                <in> Stereoscopic image format of input. Possible values:
                     sbsl or side_by_side_left_first
                            side by side parallel (left eye left,  right  eye
                            right)
                     sbsr or side_by_side_right_first
                            side  by  side crosseye (right eye left, left eye
                            right)
                     sbs2l or side_by_side_half_width_left_first
                            side by side with half width resolution (left eye
                            left, right eye right)
                     sbs2r or side_by_side_half_width_right_first
                            side  by  side  with half width resolution (right
                            eye left, left eye right)
                     abl or above_below_left_first
                            above-below (left eye above, right eye below)
                     abl or above_below_right_first
                            above-below (right eye above, left eye below)
                     ab2l or above_below_half_height_left_first
                            above-below with half height resolution (left eye
                            above, right eye below)
                     ab2r or above_below_half_height_right_first
                            above-below  with  half  height resolution (right
                            eye above, left eye below)

                <out>
                     Stereoscopic image format of output. Possible values are
                     all the input formats as well as:
                     arcg or anaglyph_red_cyan_gray
                            anaglyph  red/cyan  gray (red filter on left eye,
                            cyan filter on right eye)
                     arch or anaglyph_red_cyan_half_color
                            anaglyph red/cyan half  colored  (red  filter  on
                            left eye, cyan filter on right eye)
                     arcc or anaglyph_red_cyan_color
                            anaglyph  red/cyan color (red filter on left eye,
                            cyan filter on right eye)
                     arcd or anaglyph_red_cyan_dubois
                            anaglyph red/cyan color optimized with the  least
                            squares  projection of dubois (red filter on left
                            eye, cyan filter on right eye)
                     agmg or anaglyph_green_magenta_gray
                            anaglyph green/magenta gray (green filter on left
                            eye, magenta filter on right eye)
                     agmh or anaglyph_green_magenta_half_color
                            anaglyph green/magenta half colored (green filter
                            on left eye, magenta filter on right eye)
                     agmc or anaglyph_green_magenta_color
                            anaglyph green/magenta colored (green  filter  on
                            left eye, magenta filter on right eye)
                     aybg or anaglyph_yellow_blue_gray
                            anaglyph  yellow/blue gray (yellow filter on left
                            eye, blue filter on right eye)
                     aybh or anaglyph_yellow_blue_half_color
                            anaglyph yellow/blue half colored (yellow  filter
                            on left eye, blue filter on right eye)
                     aybc or anaglyph_yellow_blue_color
                            anaglyph  yellow/blue  colored  (yellow filter on
                            left eye, blue filter on right eye)
                     irl or interleave_rows_left_first
                            Interleaved rows (left eye has top row, right eye
                            starts on next row)
                     irr or interleave_rows_right_first
                            Interleaved rows (right eye has top row, left eye
                            starts on next row)
                     ml or mono_left
                            mono output (left eye only)
                     mr or mono_right
                            mono output (right eye only)
                NOTE: To use either of the interleaved-rows output formats to
                display full-screen on a row-interleaved 3D display, you will
                need to scale the video to the correct height first using the
                "scale" filter, if it is not already the right height.  Typi‐
                cally,   that   is   1080   rows   (so   use    e.g.     "-vf
                scale=1440:1080,stereo3d=sbsl:irl"  for  a  720p side-by-side
                encoded movie).

      gradfun[=strength[:radius]]
             Fix the banding artifacts that  are  sometimes  introduced  into
             nearly  flat regions by truncation to 8bit colordepth.  Interpo‐
             lates the gradients that should go  where  the  bands  are,  and
             dithers them.

             This  filter is designed for playback only.  Do not use it prior
             to lossy compression, because  compression  tends  to  lose  the
             dither and bring back the bands.

                <strength>
                     Maximum  amount  by which the filter will change any one
                     pixel.  Also the threshold for detecting nearly flat re‐
                     gions (default: 1.2).

                <radius>
                     Neighborhood  to  fit  the  gradient  to.  Larger radius
                     makes for smoother gradients, but also prevents the fil‐
                     ter  from  modifying  pixels  near detailed regions (de‐
                     fault: 16).

      fixpts[=options]
             Fixes the presentation timestamps (PTS) of the frames.   By  de‐
             fault,  the  PTS  passed  to the next filter is dropped, but the
             following options can change that:

                print
                     Print the incoming PTS.

                fps=<fps>
                     Specify a frame per second value.

                start=<pts>
                     Specify an initial value for the PTS.

                autostart=<n>
                     Uses the nth incoming PTS as the initial PTS.  All  pre‐
                     vious  PTS are kept, so setting a huge value or -1 keeps
                     the PTS intact.

                autofps=<n>
                     Uses the nth incoming PTS after the end of autostart  to
                     determine the framerate.

             EXAMPLE:
                -vf fixpts=fps=24000/1001,ass,fixpts
                     Generates  a new sequence of PTS, uses it for ASS subti‐
                     tles, then drops it.  Generating a new sequence is  use‐
                     ful  when  the  timestamps are reset during the program;
                     this is frequent on DVDs.  Dropping it may be  necessary
                     to avoid confusing encoders.

             NOTE:  Using  this filter together with any sort of seeking (in‐
             cluding -ss and EDLs) may make demons fly out of your nose.

GENERAL ENCODING OPTIONS (MENCODER ONLY)
      -audio-delay <any floating-point number>
             Delays either audio or video by setting a  delay  field  in  the
             header  (default: 0.0).  This does not delay either stream while
             encoding, but the player will see the delay field and compensate
             accordingly.  Positive values delay the audio, and negative val‐
             ues delay the video.  Note that this is the  exact  opposite  of
             the -delay option.  For example, if a video plays correctly with
             -delay 0.2, you can fix the video with MEncoder  by  using  -au‐
             dio-delay -0.2.

             Currently,  this  option  only works with the default muxer (-of
             avi).  If you are using a different muxer,  then  you  must  use
             -delay instead.

      -audio-density <1-50>
             Number  of  audio  chunks per second (default is 2 for 0.5s long
             audio chunks).
             NOTE: CBR only, VBR ignores this as it puts each packet in a new
             chunk.

      -audio-preload <0.0-2.0>
             Sets up the audio buffering time interval (default: 0.5s).

      -fafmttag <format>
             Can be used to override the audio format tag of the output file.

             EXAMPLE:
                -fafmttag 0x55
                     Will  have  the  output file contain 0x55 (mp3) as audio
                     format tag.

      -ffourcc <fourcc>
             Can be used to override the video fourcc of the output file.

             EXAMPLE:
                -ffourcc div3
                     Will have the output file contain 'div3' as video  four‐
                     cc.

      -force-avi-aspect <0.2-3.0>
             Override the aspect stored in the AVI OpenDML vprp header.  This
             can be used to change the aspect ratio with '-ovc copy'.

      -frameno-file <filename> (DEPRECATED)
             Specify the name of the audio  file  with  framenumber  mappings
             created  in  the first (audio only) pass of a special three pass
             encoding mode.
             NOTE: Using this mode will most likely give you A-V desync.   Do
             not  use  it.   It  is kept for backwards compatibility only and
             will possibly be removed in a future version.

      -hr-edl-seek
             Use a more precise, but much slower method for  skipping  areas.
             Areas  marked  for  skipping  are  not  seeked over, instead all
             frames are decoded, but only the necessary frames  are  encoded.
             This allows starting at non-keyframe boundaries.
             NOTE: Not guaranteed to work right with '-ovc copy'.

      -info <option1:option2:...> (AVI only)
             Specify the info header of the resulting AVI file.

             Available options are:

                help
                     Show this description.

                name=<value>
                     title of the work

                artist=<value>
                     artist or author of the work

                genre=<value>
                     original work category

                subject=<value>
                     contents of the work

                copyright=<value>
                     copyright information

                srcform=<value>
                     original format of the digitized material

                comment=<value>
                     general comments about the work

      -noautoexpand
             Do  not automatically insert the expand filter into the MEncoder
             filter chain.  Useful to control at which point  of  the  filter
             chain  subtitles  are  rendered when hardcoding subtitles onto a
             movie.

      -noencodedups
             Do not attempt to encode duplicate frames in  duplicate;  always
             output  zero-byte  frames  to  indicate  duplicates.   Zero-byte
             frames will be written anyway unless a filter or encoder capable
             of  doing duplicate encoding is loaded.  Currently the only such
             filter is harddup.

      -noodml (-of avi only)
             Do not write OpenDML index for AVI files >1GB.

      -noskip
             Do not skip frames.

      -o <filename>
             Outputs to the given filename.
             If you want a default output filename, you can put  this  option
             in the MEncoder config file.

      -oac <codec name>
             Encode with the given audio codec (no default set).
             NOTE: Use -oac help to get a list of available audio codecs.

             EXAMPLE:
                -oac copy
                     no encoding, just streamcopy
                -oac pcm
                     Encode to uncompressed PCM.
                -oac mp3lame
                     Encode to MP3 (using LAME).
                -oac lavc
                     Encode with a libavcodec codec.

      -of <format> (BETA CODE!)
             Encode to the specified container format (default: AVI).
             NOTE: Use -of help to get a list of available container formats.

             EXAMPLE:
                -of avi
                     Encode to AVI.
                -of mpeg
                     Encode to MPEG (also see -mpegopts).
                -of lavf
                     Encode with libavformat muxers (also see -lavfopts).
                -of rawvideo
                     raw video stream (no muxing - one video stream only)
                -of rawaudio
                     raw audio stream (no muxing - one audio stream only)

      -ofps <fps>
             Specify  a  frames  per  second (fps) value for the output file,
             which can be different from that of the source  material.   Must
             be  set  for  variable  fps  (ASF,  some  MOV)  and  progressive
             (30000/1001 fps telecined MPEG) files.

      -ovc <codec name>
             Encode with the given video codec (no default set).
             NOTE: Use -ovc help to get a list of available video codecs.

             EXAMPLE:
                -ovc copy
                     no encoding, just streamcopy
                -ovc raw
                     Encode to an arbitrary  uncompressed  format  (use  '-vf
                     format' to select).
                -ovc lavc
                     Encode with a libavcodec codec.

      -passlogfile <filename>
             Dump first pass information to <filename> instead of the default
             divx2pass.log in two pass encoding mode.

      -skiplimit <value>
             Specify the maximum number of frames that may be  skipped  after
             encoding one frame (-noskiplimit for unlimited).

      -vobsubout <basename>
             Specify  the  basename for the output .idx and .sub files.  This
             turns off subtitle rendering in the encoded movie and diverts it
             to VOBsub subtitle files.

      -vobsuboutid <langid>
             Specify  the  language  two letter code for the subtitles.  This
             overrides what is read from the DVD or the .ifo file.

      -vobsuboutindex <index>
             Specify the index of the subtitles in the output files (default:
             0).

      -force-key-frames <time>,<time>,...
             Force  key frames at the specified timestamps, more precisely at
             the first frame after each specified time.

             This option can be used to ensure that a seek point  is  present
             at  a  chapter  mark or any other designated place in the output
             file.

             The timestamps must be specified in ascending order.

             Since MEncoder does not send timestamps along the filter  chain,
             you  probably  need  to use the fixpts filter for this option to
             work.

             Not all codecs support forced key frames.  Currently, support is
             only implemented for the following encoders: lavc, x264, xvid.

CODEC SPECIFIC ENCODING OPTIONS (MENCODER ONLY)
      You  can specify codec specific encoding parameters using the following
      syntax:

      -<codec>opts <option1[=value1]:option2[=value2]:...>

      Where <codec> may be: lavc, xvidenc, mp3lame,  toolame,  twolame,  nuv,
      xvfw, faac, x264enc, mpeg, lavf.

  mp3lame (-lameopts)
      help
             get help

      vbr=<0-4>
             variable bitrate method
                0    cbr
                1    mt
                2    rh (default)
                3    abr
                4    mtrh

      abr
             average bitrate

      cbr
             constant bitrate Also forces CBR mode encoding on subsequent ABR
             presets modes.

      br=<0-1024>
             bitrate in kbps (CBR and ABR only)

      q=<0-9>
             quality (0 - highest, 9 - lowest) (VBR only)

      aq=<0-9>
             algorithmic quality (0 - best/slowest, 9 - worst/fastest)

      ratio=<1-100>
             compression ratio

      vol=<0-10>
             audio input gain

      mode=<0-3>
             (default: auto)
                0    stereo
                1    joint-stereo
                2    dualchannel
                3    mono

      padding=<0-2>
                0    none
                1    all
                2    adjust

      fast
             Switch on faster encoding on subsequent VBR presets modes.  This
             results in slightly lower quality and higher bitrates.

      highpassfreq=<freq>
             Set a highpass filtering frequency in Hz.  Frequencies below the
             specified one will be cut off.  A value of -1 will disable  fil‐
             tering, a value of 0 will let LAME choose values automatically.

      lowpassfreq=<freq>
             Set  a lowpass filtering frequency in Hz.  Frequencies above the
             specified one will be cut off.  A value of -1 will disable  fil‐
             tering, a value of 0 will let LAME choose values automatically.

      preset=<value>
             preset values

                help
                     Print  additional  options and information about presets
                     settings.

                medium
                     VBR encoding, good quality, 150-180 kbps bitrate range

                standard
                     VBR encoding, high quality, 170-210 kbps bitrate range

                extreme
                     VBR encoding, very high quality,  200-240  kbps  bitrate
                     range

                insane
                     CBR encoding, highest preset quality, 320 kbps bitrate

                <8-320>
                     ABR encoding at average given kbps bitrate

             EXAMPLES:
                fast:preset=standard
                     suitable  for  most  people and most music types and al‐
                     ready quite high quality
                cbr:preset=192
                     Encode with ABR presets at a 192  kbps  forced  constant
                     bitrate.
                preset=172
                     Encode with ABR presets at a 172 kbps average bitrate.
                preset=extreme
                     for  people  with  extremely  good  hearing  and similar
                     equipment

  toolame and twolame (-toolameopts and -twolameopts respectively)
      br=<32-384>
             In CBR mode this parameter indicates the bitrate in  kbps,  when
             in  VBR  mode  it is the minimum bitrate allowed per frame.  VBR
             mode will not work with a value below 112.

      vbr=<-50-50> (VBR only)
             variability range; if negative the encoder  shifts  the  average
             bitrate towards the lower limit, if positive towards the higher.
             When set to 0 CBR is used (default).

      maxvbr=<32-384> (VBR only)
             maximum bitrate allowed per frame, in kbps

      mode=<stereo | jstereo | mono | dual>
             (default: mono for 1-channel audio, stereo otherwise)

      psy=<-1-4>
             psychoacoustic model (default: 2)

      errprot=<0 | 1>
             Include error protection.

      debug=<0-10>
             debug level

  faac (-faacopts)
      br=<bitrate>
             average bitrate in kbps (mutually exclusive with quality)

      quality=<1-1000>
             quality mode, the higher the better (mutually exclusive with br)

      object=<1-4>
             object type complexity
                1    MAIN (default)
                2    LOW
                3    SSR
                4    LTP (extremely slow)

      mpeg=<2|4>
             MPEG version (default: 4)

      tns
             Enables temporal noise shaping.

      cutoff=<0-sampling_rate/2>
             cutoff frequency (default: sampling_rate/2)

      raw
             Stores the bitstream as raw payload with extradata in  the  con‐
             tainer  header  (default:  0,  corresponds to ADTS).  Do not set
             this flag if not explicitly required or you will not be able  to
             remux the audio stream later on.

  lavc (-lavcopts)
      Many  libavcodec (lavc for short) options are tersely documented.  Read
      the source for full details.

      EXAMPLE:
                vcodec=msmpeg4:vbitrate=1800:vhq:keyint=250

      o=<key>=<value>[,<key>=<value>[,...]]
             Pass AVOptions to libavcodec encoder.  Note, a patch to make the
             o=  unneeded  and  pass all unknown options through the AVOption
             system is welcome.  A full list of AVOptions can be found in the
             FFmpeg  manual.  Note that some AVOptions may conflict with MEn‐
             coder options.

             EXAMPLE:
                o=bt=100k

      acodec=<value>
             audio codec (default: mp2)
                ac3
                     Dolby Digital (AC-3)
                adpcm_*
                     Adaptive PCM formats - see the  HTML  documentation  for
                     details.
                flac
                     Free Lossless Audio Codec (FLAC)
                g726
                     G.726 ADPCM
                libfaac
                     Advanced Audio Coding (AAC) - using FAAC
                libmp3lame
                     MPEG-1 audio layer 3 (MP3) - using LAME
                mp2
                     MPEG-1 audio layer 2 (MP2)
                pcm_*
                     PCM formats - see the HTML documentation for details.
                roq_dpcm
                     Id Software RoQ DPCM
                sonic
                     experimental simple lossy codec
                sonicls
                     experimental simple lossless codec
                vorbis
                     Vorbis
                wmav1
                     Windows Media Audio v1
                wmav2
                     Windows Media Audio v2

      abitrate=<value>
             audio bitrate in kbps (default: 224)

      atag=<value>
             Use the specified Windows audio format tag (e.g. atag=0x55).

      bit_exact
             Use  only  bit  exact  algorithms (except (I)DCT).  Additionally
             bit_exact disables several optimizations and thus should only be
             used  for  regression  tests,  which need binary identical files
             even if the encoder version changes.  This also  suppresses  the
             user_data  header in MPEG-4 streams.  Do not use this option un‐
             less you know exactly what you are doing.

      threads=<1-16>
             Maximum number of threads to  use  (default:  1).   May  have  a
             slight negative effect on motion estimation.

      vcodec=<value>
             Employ the specified codec (default: mpeg4).
                asv1
                     ASUS Video v1
                asv2
                     ASUS Video v2
                dvvideo
                     Sony Digital Video
                ffv1
                     FFmpeg's lossless video codec
                ffvhuff
                     nonstandard 20% smaller HuffYUV using YV12
                flv
                     Sorenson H.263 used in Flash Video
                h261
                     H.261
                h263
                     H.263
                h263p
                     H.263+
                huffyuv
                     HuffYUV
                libtheora
                     Theora
                libx264
                     x264 H.264/AVC MPEG-4 Part 10
                libxvid
                     Xvid MPEG-4 Part 2 (ASP)
                ljpeg
                     Lossless JPEG
                mjpeg
                     Motion JPEG
                mpeg1video
                     MPEG-1 video
                mpeg2video
                     MPEG-2 video
                mpeg4
                     MPEG-4 (DivX 4/5)
                msmpeg4
                     DivX 3
                msmpeg4v2
                     MS MPEG4v2
                roqvideo
                     ID Software RoQ Video
                rv10
                     an old RealVideo codec
                snow (also see: vstrict)
                     FFmpeg's experimental wavelet-based codec
                svq1
                     Apple Sorenson Video 1
                wmv1
                     Windows Media Video, version 1 (AKA WMV7)
                wmv2
                     Windows Media Video, version 2 (AKA WMV8)

      vqmin=<1-31>
             minimum quantizer

                1    Not  recommended  (much larger file, little quality dif‐
                     ference and weird side effects: msmpeg4,  h263  will  be
                     very low quality, ratecontrol will be confused resulting
                     in lower quality and some decoders will not be  able  to
                     decode it).

                2    Recommended  for  normal  mpeg4/mpeg1video encoding (de‐
                     fault).

                3    Recommended for h263(p)/msmpeg4.  The reason for prefer‐
                     ring  3 over 2 is that 2 could lead to overflows.  (This
                     will be fixed for h263(p) by changing the quantizer  per
                     MB in the future, msmpeg4 cannot be fixed as it does not
                     support that.)

      lmin=<0.01-255.0>
             Minimum frame-level Lagrange  multiplier  for  ratecontrol  (de‐
             fault: 2.0).  Lavc will rarely use quantizers below the value of
             lmin.  Lowering lmin will make lavc more likely to choose  lower
             quantizers  for  some  frames,  but  not lower than the value of
             vqmin.  Likewise, raising lmin will make  lavc  less  likely  to
             choose  low  quantizers,  even if vqmin would have allowed them.
             You probably want to set  lmin  approximately  equal  to  vqmin.
             When  adaptive  quantization  is  in use, changing lmin/lmax may
             have less of an effect; see mblmin/mblmax.

      lmax=<0.01-255.0>
             maximum Lagrange multiplier for ratecontrol (default: 31.0)

      mblmin=<0.01-255.0>
             Minimum macroblock-level  Lagrange  multiplier  for  ratecontrol
             (default:2.0).  This parameter affects adaptive quantization op‐
             tions like qprd, lumi_mask, etc..

      mblmax=<0.01-255.0>
             Maximum macroblock-level  Lagrange  multiplier  for  ratecontrol
             (default: 31.0).

      vqscale=<0-31>
             Constant  quantizer  / constant  quality encoding (selects fixed
             quantizer mode).  A lower value means better quality but  larger
             files (default: -1).  In case of snow codec, value 0 means loss‐
             less encoding.  Since the other codecs do not support this, vqs‐
             cale=0 will have an undefined effect.  1 is not recommended (see
             vqmin for details).

      vqmax=<1-31>
             Maximum quantizer, 10-31 should be a sane range (default: 31).

      vqdiff=<1-31>
             maximum quantizer difference between consecutive I- or  P-frames
             (default: 3)

      vmax_b_frames=<0-4>
             maximum number of B-frames between non-B-frames:
                0    no B-frames (default)
                0-2  sane range for MPEG-4

      vme=<0-5>
             motion estimation method.  Available methods are:
                0    none (very low quality)
                1    full (slow, currently unmaintained and disabled)
                2    log (low quality, currently unmaintained and disabled)
                3    phods (low quality, currently unmaintained and disabled)
                4    EPZS: size=1 diamond, size can be adjusted with the *dia
                     options (default)
                5    X1 (experimental, currently aliased to EPZS)
                8    iter (iterative overlapped block, only used in snow)

             NOTE: 0-3 currently ignores the amount of bits spent, so quality
             may be low.

      me_range=<0-9999>
             motion estimation search range (default: 0 (unlimited))

      mbd=<0-2> (also see *cmp, qpel)
             Macroblock  decision  algorithm (high quality mode), encode each
             macro block in all modes and choose the best.  This is slow  but
             results  in  better quality and file size.  When mbd is set to 1
             or 2, the value of mbcmp is ignored when  comparing  macroblocks
             (the  mbcmp  value is still used in other places though, in par‐
             ticular the motion search algorithms).  If any  comparison  set‐
             ting  (precmp,  subcmp,  cmp,  or  mbcmp) is nonzero, however, a
             slower but better half-pel motion search will be  used,  regard‐
             less  of what mbd is set to.  If qpel is set, quarter-pel motion
             search will be used regardless.
                0    Use comparison function given by mbcmp (default).
                1    Select the MB mode which needs the fewest bits (=vhq).
                2    Select the MB mode which has the best rate distortion.

      vhq
             Same as mbd=1, kept for compatibility reasons.

      v4mv
             Allow 4 motion vectors per macroblock (slightly better quality).
             Works better if used with mbd>0.

      obmc
             overlapped block motion compensation (H.263+)

      loop
             loop filter (H.263+) note, this is broken

      keyint=<0-300>
             maximum  interval  between  keyframes in frames (default: 250 or
             one keyframe every ten seconds in a 25fps movie.   This  is  the
             recommended  default  for  MPEG-4).  Most codecs require regular
             keyframes in order to limit the accumulation of mismatch  error.
             Keyframes are also needed for seeking, as seeking is only possi‐
             ble to a keyframe - but keyframes need  more  space  than  other
             frames,  so  larger numbers here mean slightly smaller files but
             less precise seeking.  0 is equivalent to 1, which  makes  every
             frame a keyframe.  Values >300 are not recommended as the quali‐
             ty might be bad depending upon decoder, encoder and luck.  It is
             common for MPEG-1/2 to use values <=30.

      sc_threshold=<-1000000000-1000000000>
             Threshold for scene change detection.  A keyframe is inserted by
             libavcodec when it detects a scene change.  You can specify  the
             sensitivity  of  the  detection  with  this option.  -1000000000
             means  there  is  a  scene  change  detected  at  every   frame,
             1000000000 means no scene changes are detected (default: 0).

      sc_factor=<any positive integer>
             Causes  frames with higher quantizers to be more likely to trig‐
             ger a scene change detection and make libavcodec use an  I-frame
             (default: 1).  1-16 is a sane range.  Values between 2 and 6 may
             yield increasing PSNR (up to approximately 0.04 dB)  and  better
             placement of I-frames in high-motion scenes.  Higher values than
             6 may give very slightly better PSNR (approximately 0.01 dB more
             than sc_factor=6), but noticably worse visual quality.

      vb_strategy=<0-2> (pass one only)
             strategy to choose between I/P/B-frames:
                0    Always use the maximum number of B-frames (default).
                1    Avoid  B-frames in high motion scenes.  See the b_sensi‐
                     tivity option to tune this strategy.
                2    Places B-frames more or less optimally to yield  maximum
                     quality  (slower).  You may want to reduce the speed im‐
                     pact of this option by tuning the option brd_scale.

      b_sensitivity=<any integer greater than 0>
             Adjusts how sensitively vb_strategy=1 detects motion and  avoids
             using  B-frames  (default: 40).  Lower sensitivities will result
             in more B-frames.  Using more B-frames  usually  improves  PSNR,
             but  too  many  B-frames can hurt quality in high-motion scenes.
             Unless there is an extremely high amount of motion, b_sensitivi‐
             ty  can  safely be lowered below the default; 10 is a reasonable
             value in most cases.

      brd_scale=<0-10>
             Downscales frames for dynamic  B-frame  decision  (default:  0).
             Each  time  brd_scale  is increased by one, the frame dimensions
             are divided by two, which improves speed by a  factor  of  four.
             Both  dimensions of the fully downscaled frame must be even num‐
             bers, so brd_scale=1 requires the original dimensions to be mul‐
             tiples  of  four,  brd_scale=2 requires multiples of eight, etc.
             In other words, the dimensions of the original frame  must  both
             be divisible by 2^(brd_scale+1) with no remainder.

      bidir_refine=<0-4>
             Refine the two motion vectors used in bidirectional macroblocks,
             rather than re-using  vectors  from  the  forward  and  backward
             searches.  This option has no effect without B-frames.
                0    Disabled (default).
                1-4  Use a wider search (larger values are slower).

      vpass=<1-3>
             Activates  internal two (or more) pass mode, only specify if you
             wish to use two (or more) pass encoding.
                1    first pass (also see turbo)
                2    second pass
                3    Nth pass (second and subsequent passes of N-pass  encod‐
                     ing)
             Here is how it works, and how to use it:
             The  first pass (vpass=1) writes the statistics file.  You might
             want to deactivate some CPU-hungry options,  like  "turbo"  mode
             does.
             In two pass mode, the second pass (vpass=2) reads the statistics
             file and bases ratecontrol decisions on it.
             In N-pass mode, the second pass (vpass=3, that is  not  a  typo)
             does  both: It first reads the statistics, then overwrites them.
             You might want to backup  divx2pass.log  before  doing  this  if
             there  is any possibility that you will have to cancel MEncoder.
             You can use all encoding options, except very CPU-hungry options
             like "qns".
             You  can  run this same pass over and over to refine the encode.
             Each subsequent pass will use the statistics from  the  previous
             pass  to improve.  The final pass can include any CPU-hungry en‐
             coding options.
             If you want a  2  pass  encode,  use  first  vpass=1,  and  then
             vpass=2.
             If  you  want a 3 or more pass encode, use vpass=1 for the first
             pass and then vpass=3 and then vpass=3 again and again until you
             are satisfied with the encode.

             huffyuv:
                pass 1
                     Saves statistics.
                pass 2
                     Encodes with an optimal Huffman table based upon statis‐
                     tics from the first pass.

      turbo (two pass only)
             Dramatically speeds up pass one using faster algorithms and dis‐
             abling  CPU-intensive options.  This will probably reduce global
             PSNR a little bit (around 0.01dB) and  change  individual  frame
             type and PSNR a little bit more (up to 0.03dB).

      aspect=<x/y>
             Store  movie aspect internally, just like with MPEG files.  Much
             nicer than rescaling, because quality is  not  decreased.   Only
             MPlayer will play these files correctly, other players will dis‐
             play them with wrong aspect.  The aspect parameter can be  given
             as a ratio or a floating point number.

             EXAMPLE:
                aspect=16/9 or aspect=1.78

      autoaspect
             Same  as  the  aspect option, but automatically computes aspect,
             taking into account all the adjustments (crop/expand/scale/etc.)
             made in the filter chain.  Does not incur a performance penalty,
             so you can safely leave it always on.

      vbitrate=<value>
             Specify bitrate (default: 800).
             WARNING: 1kbit = 1000 bits
                4-16000
                     (in kbit)
                16001-24000000
                     (in bit)

      vratetol=<value>
             approximated file size tolerance in kbit.  1000-100000 is a sane
             range.  (warning: 1kbit = 1000 bits) (default: 8000)
             NOTE: vratetol should not be too large during the second pass or
             there might be problems if vrc_(min|max)rate is used.

      vrc_maxrate=<value>
             maximum bitrate in kbit/sec (default: 0, unlimited)

      vrc_minrate=<value>
             minimum bitrate in kbit/sec (default: 0, unlimited)

      vrc_buf_size=<value>
             buffer size in kbit For MPEG-1/2 this also sets the  vbv  buffer
             size, use 327 for VCD, 917 for SVCD and 1835 for DVD.

      vrc_buf_aggressivity
             currently useless

      vrc_strategy
             Ratecontrol method.  Note that some of the ratecontrol-affecting
             options will have no effect if vrc_strategy is not set to 0.
                0    Use internal lavc ratecontrol (default).
                1    Use Xvid ratecontrol (experimental; requires MEncoder to
                     be compiled with support for Xvid 1.1 or higher).

      vb_qfactor=<-31.0-31.0>
             quantizer factor between B- and non-B-frames (default: 1.25)

      vi_qfactor=<-31.0-31.0>
             quantizer factor between I- and non-I-frames (default: 0.8)

      vb_qoffset=<-31.0-31.0>
             quantizer offset between B- and non-B-frames (default: 1.25)

      vi_qoffset=<-31.0-31.0>
             (default: 0.0)
             if v{b|i}_qfactor > 0
             I/B-frame  quantizer  =  P-frame  quantizer  *  v{b|i}_qfactor +
             v{b|i}_qoffset
             else
             do normal ratecontrol (do not lock to  next  P-frame  quantizer)
             and set q= -q * v{b|i}_qfactor + v{b|i}_qoffset
             HINT:  To do constant quantizer encoding with different quantiz‐
             ers for I/P- and B-frames you can  use:  lmin=  <ip_quant>:lmax=
             <ip_quant>:vb_qfactor= <b_quant/ip_quant>.

      vqblur=<0.0-1.0> (pass one)
             Quantizer  blur  (default:  0.5), larger values will average the
             quantizer more over time (slower change).
                0.0  Quantizer blur disabled.
                1.0  Average the quantizer over all previous frames.

      vqblur=<0.0-99.0> (pass two)
             Quantizer gaussian blur (default: 0.5), larger values will aver‐
             age the quantizer more over time (slower change).

      vqcomp=<0.0-1.0>
             Quantizer  compression, vrc_eq depends upon this (default: 0.5).
             NOTE: Perceptual quality will be optimal  somewhere  in  between
             the range's extremes.

      vrc_eq=<equation>
             main ratecontrol equation

                1+(tex/avgTex-1)*qComp
                     approximately the equation of the old ratecontrol code

                tex^qComp
                     with qcomp 0.5 or something like that (default)

             infix operators:

                +,-,*,/,^

             variables:

                tex
                     texture complexity

                iTex,pTex
                     intra, non-intra texture complexity

                avgTex
                     average texture complexity

                avgIITex
                     average intra texture complexity in I-frames

                avgPITex
                     average intra texture complexity in P-frames

                avgPPTex
                     average non-intra texture complexity in P-frames

                avgBPTex
                     average non-intra texture complexity in B-frames

                mv
                     bits used for motion vectors

                fCode
                     maximum length of motion vector in log2 scale

                iCount
                     number of intra macroblocks / number of macroblocks

                var
                     spatial complexity

                mcVar
                     temporal complexity

                qComp
                     qcomp from the command line

                isI, isP, isB
                     Is 1 if picture type is I/P/B else 0.

                Pi,E
                     See your favorite math book.

             functions:

                max(a,b),min(a,b)
                     maximum / minimum

                gt(a,b)
                     is 1 if a>b, 0 otherwise

                lt(a,b)
                     is 1 if a<b, 0 otherwise

                eq(a,b)
                     is 1 if a==b, 0 otherwise

                sin, cos, tan, sinh, cosh, tanh, exp, log, abs

      vrc_override=<options>
             User  specified  quality  for  specific  parts (ending, credits,
             ...).   The  options  are  <start-frame>,  <end-frame>,  <quali‐
             ty>[/<start-frame>, <end-frame>, <quality>[/...]]:
                quality (2-31)
                     quantizer
                quality (-500-0)
                     quality correction in %

      vrc_init_cplx=<0-1000>
             initial complexity (pass 1)

      vrc_init_occupancy=<0.0-1.0>
             initial  buffer  occupancy,  as  a fraction of vrc_buf_size (de‐
             fault: 0.9)

      vqsquish=<0|1>
             Specify how to keep the quantizer between qmin and qmax.
                0    Use clipping.
                1    Use a nice differentiable function (default).

      vlelim=<-1000-1000>
             Sets single coefficient  elimination  threshold  for  luminance.
             Negative values will also consider the DC coefficient (should be
             at least -4 or lower for encoding at quant=1):
                0    disabled (default)
                -4   JVT recommendation

      vcelim=<-1000-1000>
             Sets single coefficient elimination threshold  for  chrominance.
             Negative values will also consider the DC coefficient (should be
             at least -4 or lower for encoding at quant=1):
                0    disabled (default)
                7    JVT recommendation

      vstrict=<-2|-1|0|1>
             strict standard compliance
                0    disabled
                1    Only recommended if you want to feed the output into the
                     MPEG-4 reference decoder.
                -1   Allow libavcodec specific extensions (default).
                -2   Enables  experimental  codecs and features which may not
                     be playable with future MPlayer versions (snow).

      vdpart
             Data partitioning.  Adds 2 Bytes per video packet, improves  er‐
             ror-resistance  when transferring over unreliable channels (e.g.
             streaming over the internet).  Each video packet will be encoded
             in 3 separate partitions:
                1. MVs
                     movement
                2. DC coefficients
                     low res picture
                3. AC coefficients
                     details
             MV  &  DC  are  most important, losing them looks far worse than
             losing the AC and the 1. & 2. partition.   (MV  &  DC)  are  far
             smaller  than the 3. partition (AC) meaning that errors will hit
             the AC partition much more often than the MV  &  DC  partitions.
             Thus,  the picture will look better with partitioning than with‐
             out, as without partitioning an error will trash AC/DC/MV equal‐
             ly.

      vpsize=<0-10000> (also see vdpart)
             Video packet size, improves error-resistance.
                0
                     disabled (default)
                100-1000
                     good choice

      ss
             slice structured mode for H.263+

      gray
             grayscale only encoding (faster)

      vfdct=<0-10>
             DCT algorithm
                0    Automatically select a good one (default).
                1    fast integer
                2    accurate integer
                3    MMX
                4    mlib
                5    AltiVec
                6    floating point AAN

      idct=<0-99>
             IDCT algorithm
             NOTE:  To  the best of our knowledge all these IDCTs do pass the
             IEEE1180 tests.
                0    Automatically select a good one (default).
                1    JPEG reference integer
                2    simple
                3    simplemmx
                4    libmpeg2mmx (inaccurate, do not use  for  encoding  with
                     keyint >100)
                5    ps2
                6    mlib
                7    arm
                8    AltiVec
                9    sh4
                10   simplearm
                11   H.264
                12   VP3
                13   IPP
                14   xvidmmx
                15   CAVS
                16   simplearmv5te
                17   simplearmv6

      lumi_mask=<0.0-1.0>
             Luminance  masking is a 'psychosensory' setting that is supposed
             to make use of the fact that the human eye tends to notice fewer
             details  in very bright parts of the picture.  Luminance masking
             compresses bright areas stronger than medium ones,  so  it  will
             save bits that can be spent again on other frames, raising over‐
             all subjective quality, while possibly reducing PSNR.
             WARNING: Be careful, overly large values  can  cause  disastrous
             things.
             WARNING:  Large  values might look good on some monitors but may
             look horrible on other monitors.
                0.0
                     disabled (default)
                0.0-0.3
                     sane range

      dark_mask=<0.0-1.0>
             Darkness masking is a 'psychosensory' setting that  is  supposed
             to make use of the fact that the human eye tends to notice fewer
             details in very dark parts of  the  picture.   Darkness  masking
             compresses dark areas stronger than medium ones, so it will save
             bits that can be spent again on other  frames,  raising  overall
             subjective quality, while possibly reducing PSNR.
             WARNING:  Be  careful,  overly large values can cause disastrous
             things.
             WARNING: Large values might look good on some monitors  but  may
             look horrible on other monitors / TV / TFT.
                0.0
                     disabled (default)
                0.0-0.3
                     sane range

      tcplx_mask=<0.0-1.0>
             Temporal  complexity masking (default: 0.0 (disabled)).  Imagine
             a scene with a bird flying across the  whole  scene;  tcplx_mask
             will  raise  the  quantizers of the bird's macroblocks (thus de‐
             creasing their quality), as the human eye usually does not  have
             time  to  see  all  the  bird's  details.  Be warned that if the
             masked object stops (e.g. the bird lands) it is likely  to  look
             horrible  for  a short period of time, until the encoder figures
             out that the object is not moving and needs refined blocks.  The
             saved  bits will be spent on other parts of the video, which may
             increase subjective quality, provided that tcplx_mask  is  care‐
             fully chosen.

      scplx_mask=<0.0-1.0>
             Spatial  complexity masking.  Larger values help against blocki‐
             ness, if no deblocking filter is used  for  decoding,  which  is
             maybe not a good idea.
             Imagine a scene with grass (which usually has great spatial com‐
             plexity), a blue sky and a  house;  scplx_mask  will  raise  the
             quantizers of the grass' macroblocks, thus decreasing its quali‐
             ty, in order to spend more bits on the sky and the house.
             HINT: Crop any black borders completely as they will reduce  the
             quality of the macroblocks (also applies without scplx_mask).
                0.0
                     disabled (default)
                0.0-0.5
                     sane range

             NOTE: This setting does not have the same effect as using a cus‐
             tom matrix that would compress high frequencies harder,  as  sc‐
             plx_mask  will reduce the quality of P blocks even if only DC is
             changing.  The result of scplx_mask will probably  not  look  as
             good.

      p_mask=<0.0-1.0> (also see vi_qfactor)
             Reduces  the quality of inter blocks.  This is equivalent to in‐
             creasing the quality of intra blocks, because the  same  average
             bitrate  will be distributed by the rate controller to the whole
             video sequence (default: 0.0  (disabled)).   p_mask=1.0  doubles
             the bits allocated to each intra block.

      border_mask=<0.0-1.0>
             border-processing  for  MPEG-style  encoders.  Border processing
             increases the quantizer for  macroblocks  which  are  less  than
             1/5th  of  the  frame  width/height  away from the frame border,
             since they are often visually less important.

      naq
             Normalize  adaptive  quantization  (experimental).   When  using
             adaptive quantization (*_mask), the average per-MB quantizer may
             no longer match the requested frame-level quantizer.   Naq  will
             attempt  to  adjust the per-MB quantizers to maintain the proper
             average.

      ildct
             Use interlaced DCT.

      ilme
             Use interlaced motion estimation (mutually exclusive with qpel).

      alt
             Use alternative scantable.

      top=<-1-1>
                -1   automatic
                0    bottom field first
                1    top field first

      format=<value>
                YV12
                     default
                444P
                     for ffv1
                422P
                     for HuffYUV, lossless JPEG, dv and ffv1
                411P
                     for lossless JPEG, dv and ffv1
                YVU9
                     for lossless JPEG, ffv1 and svq1
                BGR32
                     for lossless JPEG and ffv1

      pred
             (for HuffYUV)
                0    left prediction
                1    plane/gradient prediction
                2    median prediction

      pred
             (for lossless JPEG)
                0    left prediction
                1    top prediction
                2    topleft prediction
                3    plane/gradient prediction
                6    mean prediction

      coder
             (for ffv1)
                0    vlc coding (Golomb-Rice)
                1    arithmetic coding (CABAC)

      context
             (for ffv1)
                0    small context model
                1    large context model

             (for ffvhuff)
                0    predetermined Huffman tables (builtin or two pass)
                1    adaptive Huffman tables

      qpel
             Use quarter pel motion  compensation  (mutually  exclusive  with
             ilme).
             HINT: This seems only useful for high bitrate encodings.

      mbcmp=<0-2000>
             Sets  the  comparison  function for the macroblock decision, has
             only an effect if mbd=0.  This is  also  used  for  some  motion
             search  functions,  in which case it has an effect regardless of
             mbd setting.
                0 (SAD)
                     sum of absolute differences, fast (default)
                1 (SSE)
                     sum of squared errors
                2 (SATD)
                     sum of absolute Hadamard transformed differences
                3 (DCT)
                     sum of absolute DCT transformed differences
                4 (PSNR)
                     sum of squared quantization errors (avoid, low quality)
                5 (BIT)
                     number of bits needed for the block
                6 (RD)
                     rate distortion optimal, slow
                7 (ZERO)
                     0
                8 (VSAD)
                     sum of absolute vertical differences
                9 (VSSE)
                     sum of squared vertical differences
                10 (NSSE)
                     noise preserving sum of squared differences
                11 (W53)
                     5/3 wavelet, only used in snow
                12 (W97)
                     9/7 wavelet, only used in snow
                +256
                     Also use chroma, currently  does  not  work  (correctly)
                     with B-frames.

      ildctcmp=<0-2000>
             Sets  the  comparison  function for interlaced DCT decision (see
             mbcmp for available comparison functions).

      precmp=<0-2000>
             Sets the comparison function for motion estimation pre pass (see
             mbcmp for available comparison functions) (default: 0).

      cmp=<0-2000>
             Sets the comparison function for full pel motion estimation (see
             mbcmp for available comparison functions) (default: 0).

      subcmp=<0-2000>
             Sets the comparison function for sub pel motion estimation  (see
             mbcmp for available comparison functions) (default: 0).

      skipcmp=<0-2000>
             FIXME: Document this.

      nssew=<0-1000000>
             This setting controls NSSE weight, where larger weights will re‐
             sult in more noise.  0 NSSE is identical to  SSE  You  may  find
             this  useful  if  you  prefer to keep some noise in your encoded
             video rather than filtering it away  before  encoding  (default:
             8).

      predia=<-99-6>
             diamond type and size for motion estimation pre-pass

      dia=<-99-6>
             Diamond  type & size for motion estimation.  Motion search is an
             iterative process.  Using a small diamond  does  not  limit  the
             search  to  finding only small motion vectors.  It is just some‐
             what more likely to stop before finding  the  very  best  motion
             vector,  especially when noise is involved.  Bigger diamonds al‐
             low a wider search for the best motion vector, thus  are  slower
             but result in better quality.
             Big  normal diamonds are better quality than shape-adaptive dia‐
             monds.
             Shape-adaptive diamonds are a good tradeoff  between  speed  and
             quality.
             NOTE:  The  sizes of the normal diamonds and shape adaptive ones
             do not have the same meaning.

                -3   shape adaptive (fast) diamond with size 3

                -2   shape adaptive (fast) diamond with size 2

                -1   uneven multi-hexagon search (slow)

                1    normal size=1 diamond (default) =EPZS type diamond
                           0
                          000
                           0

                2    normal size=2 diamond
                           0
                          000
                         00000
                          000
                           0

      trell
             Trellis searched quantization.  This will find the  optimal  en‐
             coding  for  each  8x8  block.  Trellis searched quantization is
             quite simply an optimal quantization in the PSNR versus  bitrate
             sense  (Assuming  that  there would be no rounding errors intro‐
             duced by the IDCT, which is obviously not the case.).  It simply
             finds a block for the minimum of error and lambda*bits.
                lambda
                     quantization parameter (QP) dependent constant
                bits
                     amount of bits needed to encode the block
                error
                     sum of squared errors of the quantization

      cbp
             Rate  distorted  optimal  coded  block pattern.  Will select the
             coded block pattern which minimizes  distortion  +  lambda*rate.
             This can only be used together with trellis quantization.

      mv0
             Try  to  encode each MB with MV=<0,0> and choose the better one.
             This has no effect if mbd=0.

      mv0_threshold=<any non-negative integer>
             When surrounding motion vectors are <0,0> and the motion estima‐
             tion  score  of  the  current  block is less than mv0_threshold,
             <0,0> is used for the motion vector and further  motion  estima‐
             tion is skipped (default: 256).  Lowering mv0_threshold to 0 can
             give a slight (0.01dB) PSNR increase and possibly make  the  en‐
             coded video look slightly better; raising mv0_threshold past 320
             results in diminished PSNR and visual  quality.   Higher  values
             speed up encoding very slightly (usually less than 1%, depending
             on the other options used).
             NOTE: This option does not require mv0 to be enabled.

      qprd (mbd=2 only)
             rate distorted optimal quantization parameter (QP) for the given
             lambda of each macroblock

      last_pred=<0-99>
             amount of motion predictors from the previous frame
                0    (default)
                a    Will  use 2a+1 x 2a+1 macroblock square of motion vector
                     predictors from the previous frame.

      preme=<0-2>
             motion estimation pre-pass
                0    disabled
                1    only after I-frames (default)
                2    always

      subq=<1-8>
             subpel refinement quality (for qpel) (default: 8 (high quality))
             NOTE: This has a significant effect on speed.

      refs=<1-8>
             number of reference frames to consider for  motion  compensation
             (Snow only) (default: 1)

      psnr
             print  the PSNR (peak signal to noise ratio) for the whole video
             after encoding and store the per frame PSNR in  a  file  with  a
             name  like  'psnr_hhmmss.log'.  Returned values are in dB (deci‐
             bel), the higher the better.

      mpeg_quant
             Use MPEG quantizers instead of H.263.

      aic
             Enable AC prediction for MPEG-4 or advanced intra prediction for
             H.263+.  This will improve quality very slightly (around 0.02 dB
             PSNR) and slow down encoding very slightly (about 1%).
             NOTE: vqmin should be 8 or larger for H.263+ AIC.

      aiv
             alternative inter vlc for H.263+

      umv
             unlimited MVs (H.263+ only) Allows encoding of arbitrarily  long
             MVs.

      ibias=<-256-256>
             intra  quantizer  bias (256 equals 1.0, MPEG style quantizer de‐
             fault: 96, H.263 style quantizer default: 0)
             NOTE: The H.263 MMX quantizer cannot handle positive biases (set
             vfdct=1 or 2), the MPEG MMX quantizer cannot handle negative bi‐
             ases (set vfdct=1 or 2).

      pbias=<-256-256>
             inter quantizer bias (256 equals 1.0, MPEG style  quantizer  de‐
             fault: 0, H.263 style quantizer default: -64)
             NOTE: The H.263 MMX quantizer cannot handle positive biases (set
             vfdct=1 or 2), the MPEG MMX quantizer cannot handle negative bi‐
             ases (set vfdct=1 or 2).
             HINT:  A  more positive bias (-32 - -16 instead of -64) seems to
             improve the PSNR.

      nr=<0-100000>
             Noise reduction, 0 means disabled.  0-600 is a useful range  for
             typical  content,  but you may want to turn it up a bit more for
             very noisy content (default: 0).   Given  its  small  impact  on
             speed, you might want to prefer to use this over filtering noise
             away with video filters like denoise3d or hqdn3d.

      qns=<0-3>
             Quantizer noise shaping.  Rather than choosing  quantization  to
             most closely match the source video in the PSNR sense, it choos‐
             es quantization such that noise (usually ringing) will be masked
             by  similar-frequency  content  in the image.  Larger values are
             slower but may not result  in  better  quality.   This  can  and
             should be used together with trellis quantization, in which case
             the trellis quantization (optimal for constant weight)  will  be
             used as startpoint for the iterative search.
                0    disabled (default)
                1    Only lower the absolute value of coefficients.
                2    Only  change coefficients before the last non-zero coef‐
                     ficient + 1.
                3    Try all.

      inter_matrix=<comma separated matrix>
             Use custom inter matrix.  It needs a comma separated  string  of
             64 integers.

      intra_matrix=<comma separated matrix>
             Use  custom  intra matrix.  It needs a comma separated string of
             64 integers.

      vqmod_amp
             experimental quantizer modulation

      vqmod_freq
             experimental quantizer modulation

      dc
             intra DC  precision  in  bits  (default:  8).   If  you  specify
             vcodec=mpeg2video this value can be 8, 9, 10 or 11.

      cgop (also see sc_threshold)
             Close  all GOPs.  Currently it only works if scene change detec‐
             tion is disabled (sc_threshold=1000000000).

      gmc
             Enable Global Motion Compensation.

      (no)lowdelay
             Sets the low delay flag for MPEG-1/2 (disables B-frames).

      vglobal=<0-3>
             Control writing global video headers.
                0    Codec decides where to write global headers (default).
                1    Write global  headers  only  in  extradata  (needed  for
                     .mp4/MOV/NUT).
                2    Write global headers only in front of keyframes.
                3    Combine 1 and 2.

      aglobal=<0-3>
             Same as vglobal for audio headers.

      level=<value>
             Set  CodecContext  Level.   Use  31  or  41  to  play video on a
             Playstation 3.

      skip_exp=<0-1000000>
             FIXME: Document this.

      skip_factor=<0-1000000>
             FIXME: Document this.

      skip_threshold=<0-1000000>
             FIXME: Document this.

  nuv (-nuvopts)
      Nuppel video is based on RTJPEG and LZO.  By default frames  are  first
      encoded with RTJPEG and then compressed with LZO, but it is possible to
      disable either or both of the two passes.  As a result, you can in fact
      output  raw  i420, LZO compressed i420, RTJPEG, or the default LZO com‐
      pressed RTJPEG.
      NOTE: The nuvrec documentation contains some advice and examples  about
      the settings to use for the most common TV encodings.

      c=<0-20>
             chrominance threshold (default: 1)

      l=<0-20>
             luminance threshold (default: 1)

      lzo
             Enable LZO compression (default).

      nolzo
             Disable LZO compression.

      q=<3-255>
             quality level (default: 255)

      raw
             Disable RTJPEG encoding.

      rtjpeg
             Enable RTJPEG encoding (default).

  xvidenc (-xvidencopts)
      There are three modes available: constant bitrate (CBR), fixed quantiz‐
      er and two pass.

      pass=<1|2>
             Specify the pass in two pass mode.

      turbo (two pass only)
             Dramatically speeds up pass one using faster algorithms and dis‐
             abling  CPU-intensive options.  This will probably reduce global
             PSNR a little bit and change individual frame type  and  PSNR  a
             little bit more.

      bitrate=<value> (CBR or two pass mode)
             Sets  the  bitrate  to  be  used in kbits/second if <16000 or in
             bits/second if >16000.  If <value> is negative,  Xvid  will  use
             its  absolute  value as the target size (in kBytes) of the video
             and compute the associated bitrate automagically  (default:  687
             kbits/s).

      fixed_quant=<1-31>
             Switch  to  fixed quantizer mode and specify the quantizer to be
             used.

      zones=<zone0>[/<zone1>[/...]] (CBR or two pass mode)
             User specified quality  for  specific  parts  (ending,  credits,
             ...).   Each  zone  is <start-frame>,<mode>,<value> where <mode>
             may be
                q    Constant quantizer override, where value=<2.0-31.0> rep‐
                     resents the quantizer value.
                w    Ratecontrol  weight  override,  where  value=<0.01-2.00>
                     represents the quality correction in %.

             EXAMPLE:
                zones=90000,q,20
                     Encodes all frames starting with frame 90000 at constant
                     quantizer 20.
                zones=0,w,0.1/10001,w,1.0/90000,q,20
                     Encode  frames  0-10000  at  10%  bitrate, encode frames
                     90000 up to the end at constant quantizer 20.  Note that
                     the  second zone is needed to delimit the first zone, as
                     without it everything up until frame 89999 would be  en‐
                     coded at 10% bitrate.

      me_quality=<0-6>
             This option controls the motion estimation subsystem.  The high‐
             er the value, the more precise the  estimation  should  be  (de‐
             fault:  6).  The more precise the motion estimation is, the more
             bits can be saved.  Precision is gained at the  expense  of  CPU
             time so decrease this setting if you need realtime encoding.

      (no)qpel
             MPEG-4  uses a half pixel precision for its motion search by de‐
             fault.  The standard proposes a mode where encoders are  allowed
             to  use quarter pixel precision.  This option usually results in
             a sharper image.  Unfortunately it has a great impact on bitrate
             and sometimes the higher bitrate use will prevent it from giving
             a better image quality at a fixed bitrate.  It is better to test
             with  and  without this option and see whether it is worth acti‐
             vating.

      (no)gmc
             Enable Global Motion Compensation,  which  makes  Xvid  generate
             special  frames (GMC-frames) which are well suited for Pan/Zoom/
             Rotating images.  Whether or not the use  of  this  option  will
             save bits is highly dependent on the source material.

      (no)trellis
             Trellis  Quantization  is a kind of adaptive quantization method
             that saves bits by modifying quantized coefficients to make them
             more compressible by the entropy encoder.  Its impact on quality
             is good, and if VHQ uses too much CPU for you, this setting  can
             be  a  good  alternative to save a few bits (and gain quality at
             fixed bitrate) at a lesser cost than with VHQ (default: on).

      (no)cartoon
             Activate this if your encoded sequence is an anime/cartoon.   It
             modifies  some Xvid internal thresholds so Xvid takes better de‐
             cisions on frame types and motion vectors for flat looking  car‐
             toons.

      (no)chroma_me
             The  usual  motion  estimation algorithm uses only the luminance
             information to find the best motion vector.   However  for  some
             video  material,  using  the  chroma planes can help find better
             vectors.  This setting toggles the use of chroma planes for  mo‐
             tion estimation (default: on).

      (no)chroma_opt
             Enable a chroma optimizer prefilter.  It will do some extra mag‐
             ic on color information to minimize the stepped-stairs effect on
             edges.   It  will improve quality at the cost of encoding speed.
             It reduces PSNR by nature, as the mathematical deviation to  the
             original picture will get bigger, but the subjective image qual‐
             ity will raise.  Since it  works  with  color  information,  you
             might want to turn it off when encoding in grayscale.

      (no)hq_ac
             Activates  high-quality  prediction of AC coefficients for intra
             frames from neighbor blocks (default: on).

      vhq=<0-4>
             The motion search algorithm is based on a search  in  the  usual
             color  domain  and  tries to find a motion vector that minimizes
             the difference between  the  reference  frame  and  the  encoded
             frame.  With this setting activated, Xvid will also use the fre‐
             quency domain (DCT) to search for a motion vector that minimizes
             not  only the spatial difference but also the encoding length of
             the block.  Fastest to slowest:
                0    off
                1    mode decision (inter/intra MB) (default)
                2    limited search
                3    medium search
                4    wide search

      (no)lumi_mask
             Adaptive quantization allows the macroblock quantizers  to  vary
             inside  each  frame.   This is a 'psychosensory' setting that is
             supposed to make use of the fact that the human eye tends to no‐
             tice  fewer  details  in  very bright and very dark parts of the
             picture.  It compresses those areas more  strongly  than  medium
             ones,  which  will  save  bits  that can be spent again on other
             frames, raising overall subjective quality and possibly reducing
             PSNR.

      (no)grayscale
             Make  Xvid  discard  chroma  planes  so  the  encoded  video  is
             grayscale only.  Note that this does not speed up  encoding,  it
             just  prevents  chroma data from being written in the last stage
             of encoding.

      (no)interlacing
             Encode the fields of interlaced video material.  Turn  this  op‐
             tion on for interlaced content.
             NOTE: Should you rescale the video, you would need an interlace-
             aware   resizer,   which   you    can    activate    with    -vf
             scale=<width>:<height>:1.

      min_iquant=<0-31>
             minimum I-frame quantizer (default: 2)

      max_iquant=<0-31>
             maximum I-frame quantizer (default: 31)

      min_pquant=<0-31>
             minimum P-frame quantizer (default: 2)

      max_pquant=<0-31>
             maximum P-frame quantizer (default: 31)

      min_bquant=<0-31>
             minimum B-frame quantizer (default: 2)

      max_bquant=<0-31>
             maximum B-frame quantizer (default: 31)

      min_key_interval=<value> (two pass only)
             minimum interval between keyframes (default: 0)

      max_key_interval=<value>
             maximum interval between keyframes (default: 10*fps)

      quant_type=<h263|mpeg>
             Sets  the type of quantizer to use.  For high bitrates, you will
             find that MPEG quantization preserves more detail.  For low  bi‐
             trates,  the  smoothing of H.263 will give you less block noise.
             When using custom matrices, MPEG quantization must be used.

      quant_intra_matrix=<filename>
             Load a custom intra matrix file.  You can build such a file with
             xvid4conf's matrix editor.

      quant_inter_matrix=<filename>
             Load a custom inter matrix file.  You can build such a file with
             xvid4conf's matrix editor.

      keyframe_boost=<0-1000> (two pass mode only)
             Shift some bits from the pool for other  frame  types  to  intra
             frames,  thus improving keyframe quality.  This amount is an ex‐
             tra percentage, so a value of 10 will give  your  keyframes  10%
             more bits than normal (default: 0).

      kfthreshold=<value> (two pass mode only)
             Works  together  with  kfreduction.  Determines the minimum dis‐
             tance below which you consider that two  frames  are  considered
             consecutive  and  treated  differently  according to kfreduction
             (default: 10).

      kfreduction=<0-100> (two pass mode only)
             The above two settings  can  be  used  to  adjust  the  size  of
             keyframes  that  you consider too close to the first (in a row).
             kfthreshold sets the range in which keyframes are  reduced,  and
             kfreduction determines the bitrate reduction they get.  The last
             I-frame will get treated normally (default: 30).

      max_bframes=<0-4>
             Maximum number of B-frames to put between  I/P-frames  (default:
             2).

      bquant_ratio=<0-1000>
             quantizer  ratio between B- and non-B-frames, 150=1.50 (default:
             150)

      bquant_offset=<-1000-1000>
             quantizer offset between B- and non-B-frames, 100=1.00 (default:
             100)

      bf_threshold=<-255-255>
             This setting allows you to specify what priority to place on the
             use of B-frames.  The higher the value, the higher the probabil‐
             ity  of B-frames being used (default: 0).  Do not forget that B-
             frames usually have a higher quantizer, and therefore aggressive
             production of B-frames may cause worse visual quality.

      (no)closed_gop
             This  option  tells  Xvid  to close every GOP (Group Of Pictures
             bounded by two I-frames), which makes GOPs independent from each
             other.   This just implies that the last frame of the GOP is ei‐
             ther a P-frame or a N-frame but not a B-frame.  It is usually  a
             good idea to turn this option on (default: on).

      (no)packed
             This  option  is meant to solve frame-order issues when encoding
             to container formats like AVI that cannot cope with out-of-order
             frames.  In practice, most decoders (both software and hardware)
             are able to deal with frame-order themselves, and may  get  con‐
             fused  when this option is turned on, so you can safely leave if
             off, unless you really know what you are doing.
             WARNING: This will generate an illegal bitstream, and  will  not
             be decodable by ISO-MPEG-4 decoders except DivX/libavcodec/Xvid.
             WARNING: This will also store a fake DivX version in the file so
             the bug autodetection of some decoders might be confused.

      frame_drop_ratio=<0-100> (max_bframes=0 only)
             This setting allows the creation  of  variable  framerate  video
             streams.   The  value of the setting specifies a threshold under
             which, if the difference of the following frame to the  previous
             frame is below or equal to this threshold, a frame gets not cod‐
             ed (a so called n-vop is placed in the  stream).   On  playback,
             when reaching an n-vop the previous frame will be displayed.
             WARNING:  Playing with this setting may result in a jerky video,
             so use it at your own risks!

      rc_reaction_delay_factor=<value>
             This parameter controls the number of frames the CBR  rate  con‐
             troller will wait before reacting to bitrate changes and compen‐
             sating for them to obtain a constant bitrate over  an  averaging
             range of frames.

      rc_averaging_period=<value>
             Real  CBR  is hard to achieve.  Depending on the video material,
             bitrate can be variable, and hard to  predict.   Therefore  Xvid
             uses  an averaging period for which it guarantees a given amount
             of bits (minus a small variation).  This settings expresses  the
             "number  of frames" for which Xvid averages bitrate and tries to
             achieve CBR.

      rc_buffer=<value>
             size of the rate control buffer

      curve_compression_high=<0-100>
             This setting allows Xvid to take a certain  percentage  of  bits
             away  from  high  bitrate  scenes  and give them back to the bit
             reservoir.  You could also use this if you have a clip  with  so
             many  bits allocated to high-bitrate scenes that the low(er)-bi‐
             trate scenes start to look bad (default: 0).

      curve_compression_low=<0-100>
             This setting allows Xvid to give a certain percentage  of  extra
             bits  to  the low bitrate scenes, taking a few bits from the en‐
             tire clip.  This might come in handy if you have a  few  low-bi‐
             trate scenes that are still blocky (default: 0).

      overflow_control_strength=<0-100>
             During  pass one of two pass encoding, a scaled bitrate curve is
             computed.  The difference between that expected  curve  and  the
             result  obtained during encoding is called overflow.  Obviously,
             the two pass rate controller tries to compensate for that  over‐
             flow,  distributing  it over the next frames.  This setting con‐
             trols how much of the overflow is distributed every  time  there
             is  a  new  frame.   Low values allow lazy overflow control, big
             rate bursts are compensated for more slowly (could lead to  lack
             of  precision for small clips).  Higher values will make changes
             in bit redistribution more abrupt, possibly too  abrupt  if  you
             set it too high, creating artifacts (default: 5).
             NOTE:  This setting impacts quality a lot, play with it careful‐
             ly!

      max_overflow_improvement=<0-100>
             During the frame bit allocation, overflow control  may  increase
             the frame size.  This parameter specifies the maximum percentage
             by which the overflow control is allowed to increase  the  frame
             size, compared to the ideal curve allocation (default: 5).

      max_overflow_degradation=<0-100>
             During  the  frame bit allocation, overflow control may decrease
             the frame size.  This parameter specifies the maximum percentage
             by  which  the overflow control is allowed to decrease the frame
             size, compared to the ideal curve allocation (default: 5).

      container_frame_overhead=<0...>
             Specifies a frame average overhead per frame, in bytes.  Most of
             the time users express their target bitrate for video w/o taking
             care of the video container overhead.  This small  but  (mostly)
             constant overhead can cause the target file size to be exceeded.
             Xvid allows users to set the amount of overhead  per  frame  the
             container  generates  (give only an average per frame).  0 has a
             special meaning, it lets Xvid use its own  default  values  (de‐
             fault: 24 - AVI average overhead).

      profile=<profile_name>
             Restricts options and VBV (peak bitrate over a short period) ac‐
             cording to the Simple, Advanced Simple and DivX  profiles.   The
             resulting videos should be playable on standalone players adher‐
             ing to these profile specifications.
                unrestricted
                     no restrictions (default)
                sp0
                     simple profile at level 0
                sp1
                     simple profile at level 1
                sp2
                     simple profile at level 2
                sp3
                     simple profile at level 3
                sp4a
                     simple profile at level 4a
                sp5
                     simple profile at level 5
                sp6
                     simple profile at level 6
                asp0
                     advanced simple profile at level 0
                asp1
                     advanced simple profile at level 1
                asp2
                     advanced simple profile at level 2
                asp3
                     advanced simple profile at level 3
                asp4
                     advanced simple profile at level 4
                asp5
                     advanced simple profile at level 5
                dxnhandheld
                     DXN handheld profile
                dxnportntsc
                     DXN portable NTSC profile
                dxnportpal
                     DXN portable PAL profile
                dxnhtntsc
                     DXN home theater NTSC profile
                dxnhtpal
                     DXN home theater PAL profile
                dxnhdtv
                     DXN HDTV profile
             NOTE: These profiles should be used in conjunction with  an  ap‐
             propriate -ffourcc.  Generally DX50 is applicable, as some play‐
             ers do not recognize Xvid but most recognize DivX.

      par=<mode>
             Specifies the Pixel Aspect Ratio mode (not to be  confused  with
             DAR,  the  Display Aspect Ratio).  PAR is the ratio of the width
             and height of a single pixel.  So both are  related  like  this:
             DAR = PAR * (width/height).
             MPEG-4  defines 5 pixel aspect ratios and one extended one, giv‐
             ing the opportunity to specify a specific pixel aspect ratio.  5
             standard modes can be specified:
                vga11
                     It is the usual PAR for PC content.  Pixels are a square
                     unit.
                pal43
                     PAL standard 4:3 PAR.  Pixels are rectangles.
                pal169
                     same as above
                ntsc43
                     same as above
                ntsc169
                     same as above (Do not forget to give the exact ratio.)
                ext
                     Allows you to specify your own pixel aspect  ratio  with
                     par_width and par_height.
             NOTE:  In  general,  setting  aspect  and  autoaspect options is
             enough.

      par_width=<1-255> (par=ext only)
             Specifies the width of the custom pixel aspect ratio.

      par_height=<1-255> (par=ext only)
             Specifies the height of the custom pixel aspect ratio.

      aspect=<x/y | f (float value)>
             Store movie aspect internally, just like MPEG files.  Much nicer
             solution  than  rescaling,  because  quality  is  not decreased.
             MPlayer and a few others players will play these files  correct‐
             ly,  others will display them with the wrong aspect.  The aspect
             parameter can be given as a ratio or a floating point number.

      (no)autoaspect
             Same as the aspect option, but  automatically  computes  aspect,
             taking into account all the adjustments (crop/expand/scale/etc.)
             made in the filter chain.

      psnr
             Print the PSNR (peak signal to noise ratio) for the whole  video
             after  encoding  and  store  the per frame PSNR in a file with a
             name like 'psnr_hhmmss.log' in the current directory.   Returned
             values are in dB (decibel), the higher the better.

      debug
             Save  per-frame  statistics  in ./xvid.dbg. (This is not the two
             pass control file.)

      The following options are only available in Xvid 1.1.x and later.

      bvhq=<0|1>
             This setting allows vector candidates for B-frames  to  be  used
             for the encoding chosen using a rate distortion optimized opera‐
             tor, which is what is done for P-frames by the vhq option.  This
             produces  nicer-looking  B-frames while incurring almost no per‐
             formance penalty (default: 1).

      vbv_bufsize=<0...> (two pass mode only)
             Specify the video buffering verifier (VBV) buffer size  in  bits
             (default:  0 - VBV check disabled).  VBV allows restricting peak
             bitrate to make the video play  properly  on  hardware  players.
             For  example, the Home profile uses vbv_bufsize=3145728.  If you
             set vbv_bufsize you should  set  also  vbv_maxrate.   Note  that
             there  is  no vbv_peakrate because Xvid does not actually use it
             for bitrate controlling; the other VBV options are enough to re‐
             strict the peak bitrate.

      vbv_initial=<0...vbv_bufsize> (two pass mode only)
             Specify the initial fill of the VBV buffer in bits (default: 75%
             of vbv_bufsize).  The default is probably what you want.

      vbv_maxrate=<0...> (two pass mode only)
             Specify the maximum processing rate in bits/s (default: 0).  For
             example, the Home profile uses vbv_maxrate=4854000.

      The following option is only available in Xvid 1.2.x and later.

      threads=<0-n>
             Create n threads to run the motion estimation (default: 0).  The
             maximum number of threads that can be used is the picture height
             divided by 16.

  x264enc (-x264encopts)
      bitrate=<value>
             Sets  the  average  bitrate to be used in kbits/second (default:
             off).  Since local bitrate may vary, this average may be inaccu‐
             rate  for very short videos (see ratetol).  Constant bitrate can
             be achieved by combining this with vbv_maxrate,  at  significant
             reduction in quality.

      qp=<0-51>
             This selects the quantizer to use for P-frames.  I- and B-frames
             are offset from this value by ip_factor and  pb_factor,  respec‐
             tively.  20-40 is a useful range.  Lower values result in better
             fidelity, but higher bitrates.  0 is lossless.  Note that  quan‐
             tization  in  H.264  works  differently from MPEG-1/2/4: H.264's
             quantization parameter (QP) is on a logarithmic scale.  The map‐
             ping  is  approximately H264QP = 12 + 6*log2(MPEGQP).  For exam‐
             ple, MPEG at QP=2 is equivalent to H.264 at  QP=18.   Generally,
             this  option should be avoided and crf should be used instead as
             crf will yield better visual results for the same size.

      crf=<1.0-50.0>
             Enables constant quality mode, and  selects  the  quality.   The
             scale  is similar to QP.  Like the bitrate-based modes, this al‐
             lows each frame to use a different QP based on the frame's  com‐
             plexity.  This option should generally be used instead of qp.

      crf_max=<float>
             With  CRF  and VBV, limit RF to this value (may cause VBV under‐
             flows!).

      pass=<1-3>
             Enable 2 or 3-pass mode.  It is recommended to always encode  in
             2  or  3-pass  mode as it leads to a better bit distribution and
             improves overall quality.
                1    first pass
                2    second pass (of two pass encoding)
                3    Nth pass (second and third passes of three  pass  encod‐
                     ing)
             Here is how it works, and how to use it:
             The  first  pass  (pass=1)  collects statistics on the video and
             writes them to a file.  You might want to deactivate  some  CPU-
             hungry options, apart from the ones that are on by default.
             In  two pass mode, the second pass (pass=2) reads the statistics
             file and bases ratecontrol decisions on it.
             In three pass mode, the second pass (pass=3, that is not a typo)
             does  both: It first reads the statistics, then overwrites them.
             You can use all encoding options,  except  very  CPU-hungry  op‐
             tions.
             The  third  pass (pass=3) is the same as the second pass, except
             that it has the second pass' statistics to work from.   You  can
             use all encoding options, including CPU-hungry ones.
             The  first pass may use either average bitrate or constant quan‐
             tizer.  ABR is recommended, since it does not require guessing a
             quantizer.  Subsequent passes are ABR, and must specify bitrate.

      profile=<name>
             Constrain options to be compatible with an H.264 profile.
                baseline
                     no8x8dct  bframes=0  nocabac cqm=flat weightp=0 nointer‐
                     laced qp>0
                main no8x8dct cqm=flat qp>0
                high qp>0 (default)

      preset=<name>
             Use a preset to select encoding settings.
                ultrafast
                     no8x8dct aq_mode=0 b_adapt=0 bframes=0  nodeblock  nomb‐
                     tree  me=dia  nomixed_refs  partitions=none ref=1 scene‐
                     cut=0 subq=0 trellis=0 noweight_b weightp=0
                superfast
                     nombtree me=dia nomixed_refs partitions=i8x8,i4x4  ref=1
                     subq=1 trellis=0 weightp=0
                veryfast
                     nombtree nomixed_refs ref=1 subq=2 trellis=0 weightp=0
                faster
                     nomixed_refs rc_lookahead=20 ref=5 subq=4 weightp=1
                fast rc_lookahead=30 ref=2 subq=6
                medium
                     Default settings apply.
                slow b_adapt=2   direct=auto   me=umh  rc_lookahead=50  ref=5
                     subq=8
                slower
                     b_adapt=2  direct=auto  me=umh  partitions=all  rc_look‐
                     ahead=60 ref=8 subq=9 trellis=2
                veryslow
                     b_adapt=2 b_frames=8 direct=auto me=umh me_range=24 par‐
                     titions=all ref=16 subq=10 trellis=2 rc_lookahead=60
                placebo
                     bframes=16 b_adapt=2  direct=auto  nofast_pskip  me=tesa
                     me_range=24    partitions=all   rc_lookahead=60   ref=16
                     subq=10 trellis=2

      tune=<name,[name,...]>
             Tune the settings for a particular type of source or  situation.
             All  tuned  settings  are  overridden by explicit user-settings.
             Multiple tunings are separated by commas, but only one psy  tun‐
             ing can be used at a time.
                film (psy tuning)
                     deblock=-1,-1 psy-rd=<unset>,0.15
                animation (psy tuning)
                     b_frames={+2}       deblock=1,1       psy-rd=0.4:<unset>
                     aq_strength=0.6 ref={double if >1 else 1}
                grain (psy tuning)
                     aq_strength=0.5  nodct_decimate  deadzone_inter=6  dead‐
                     zone_intra=6  deblock=-2,-2 ipratio=1.1 pbratio=1.1 psy-
                     rd=<unset>,0.25 qcomp=0.8
                stillimage (psy tuning)
                     aq_strength=1.2 deblock=-3,-3 psy-rd=2.0,0.7
                psnr (psy tuning)
                     aq_mode=0 nopsy
                ssim (psy tuning)
                     aq_mode=2 nopsy
                fastdecode
                     nocabac nodeblock noweight_b weightp=0
                zerolatency
                     bframes=0  force_cfr   rc_lookahead=0   sync_lookahead=0
                     sliced_threads

      slow_firstpass
             Disables  the  following  faster  options with pass=1: no_8x8dct
             me=dia partitions=none ref=1 subq={2 if >2 else unchanged} trel‐
             lis=0 fast_pskip.  These settings significantly improve encoding
             speed while having little or no impact on the quality of the fi‐
             nal pass.
             This option is implied with preset=placebo.

      keyint=<value>
             Sets maximum interval between IDR-frames (default: 250).  Larger
             values save bits, thus improve quality, at the cost  of  seeking
             precision.   Unlike  MPEG-1/2/4,  H.264 does not suffer from DCT
             drift with large values of keyint.

      keyint_min=<1-keyint/2>
             Sets minimum interval between IDR-frames  (default:  auto).   If
             scenecuts appear within this interval, they are still encoded as
             I-frames, but do not start a new GOP.  In H.264, I-frames do not
             necessarily  bound a closed GOP because it is allowable for a P-
             frame to be predicted from more frames than just the  one  frame
             before it (also see frameref).  Therefore, I-frames are not nec‐
             essarily seekable.  IDR-frames restrict subsequent P-frames from
             referring to any frame prior to the IDR-frame.

      scenecut=<-1-100>
             Controls  how  aggressively  to  insert extra I-frames (default:
             40).  With small values of scenecut,  the  codec  often  has  to
             force  an  I-frame  when it would exceed keyint.  Good values of
             scenecut may find a better location for the I-frame.  Large val‐
             ues  use  more  I-frames  than necessary, thus wasting bits.  -1
             disables scene-cut detection, so I-frames are inserted only once
             every  other  keyint frames, even if a scene-cut occurs earlier.
             This is not recommended and wastes bitrate as scenecuts  encoded
             as  P-frames  are  just as big as I-frames, but do not reset the
             "keyint counter".

      (no)intra_refresh
             Periodic intra block refresh instead of keyframes (default: dis‐
             abled).   This  option disables IDR-frames, and, instead, uses a
             moving vertical bar of intra-coded blocks. This reduces compres‐
             sion  efficiency  but  benefits  low-latency  streaming  and re‐
             silience to packet loss.

      frameref=<1-16>
             Number of previous frames used as predictors in B- and  P-frames
             (default:  3).   This  is effective in anime, but in live-action
             material the improvements usually drop off very rapidly above  6
             or  so  reference frames.  This has no effect on decoding speed,
             but does increase the memory needed for decoding.  Some decoders
             can only handle a maximum of 15 reference frames.

      bframes=<0-16>
             maximum  number  of consecutive B-frames between I- and P-frames
             (default: 3)

      (no)b_adapt
             Automatically decides when to use B-frames and how many,  up  to
             the  maximum  specified  above (default: on).  If this option is
             disabled, then the maximum number of B-frames is used.

      b_bias=<-100-100>
             Controls the decision performed by  b_adapt.   A  higher  b_bias
             produces more B-frames (default: 0).

      b_pyramid=<normal|strict|none>
             Allows  B-frames  to  be used as references for predicting other
             frames.  For example, consider 3 consecutive B-frames: I0 B1  B2
             B3 P4.  Without this option, B-frames follow the same pattern as
             MPEG-[124].  So they are coded in the order I0 P4 B1 B2 B3,  and
             all  the  B-frames  are predicted from I0 and P4.  With this op‐
             tion, they are coded as I0 P4 B2 B1  B3.   B2  is  the  same  as
             above,  but  B1 is predicted from I0 and B2, and B3 is predicted
             from B2 and P4.  This usually results in slightly improved  com‐
             pression,  at almost no speed cost.  However, this is an experi‐
             mental option: it is not fully tuned and may  not  always  help.
             Requires  bframes  >= 2.  Disadvantage: increases decoding delay
             to 2 frames.
                normal
                     Allow B-frames as references  as  described  above  (not
                     Blu-ray compatible).
                strict
                     Disallow P-frames referencing B-frames. Gives worse com‐
                     pression, but is required for Blu-ray compatibility.
                none
                     Disable using B-frames as references.

      (no)open_gop
             Use recovery points to close GOPs; only available with bframes.

      (no)bluray_compat
             Enable compatibility hacks for Blu-Ray support.

      (no)fake_interlaced
             Flag stream as interlaced but encode progressive. Makes it poss‐
             sible  to  encode 25p and 30p Blu-Ray streams. Ignored in inter‐
             laced mode.

      frame_packing=<0-5>
             Define frame arrangement for stereoscopic videos.
                0    Checkerboard - pixels are alternately from L and R.
                1    Column alternation - L and R are interlaced by column.
                2    Row alternation - L and R are interlaced by row.
                3    Side by side - L is on the left, R is on the right.
                4    Top-bottom - L is on top, R is on the bottom.
                5    Frame alternation - one view per frame.

      (no)deblock
             Use deblocking filter (default: on).  As it  takes  very  little
             time compared to its quality gain, it is not recommended to dis‐
             able it.

      deblock=<-6-6>,<-6-6>
             The first parameter  is  AlphaC0  (default:  0).   This  adjusts
             thresholds for the H.264 in-loop deblocking filter.  First, this
             parameter adjusts the maximum amount of change that  the  filter
             is  allowed to cause on any one pixel.  Secondly, this parameter
             affects the threshold for difference across the edge being  fil‐
             tered.   A  positive  value reduces blocking artifacts more, but
             will also smear details.
             The second parameter is Beta (default: 0).  This affects the de‐
             tail  threshold.   Very  detailed blocks are not filtered, since
             the smoothing caused by the filter would be more noticeable than
             the original blocking.
             The  default behavior of the filter almost always achieves opti‐
             mal quality, so it is best to either leave it alone, or make on‐
             ly  small adjustments.  However, if your source material already
             has some blocking or noise which you would like  to  remove,  it
             may be a good idea to turn it up a little bit.

      (no)cabac
             Use  CABAC (Context-Adaptive Binary Arithmetic Coding) (default:
             on).  Slightly slows down encoding and decoding, but should save
             10-15%  bitrate.  Unless you are looking for decoding speed, you
             should not disable it.

      qp_min=<1-51> (ABR or two pass)
             Minimum quantizer, 10-30 seems to be a  useful  range  (default:
             10).

      qp_max=<1-51> (ABR or two pass)
             maximum quantizer (default: 51)

      qp_step=<1-50> (ABR or two pass)
             maximum  value  by which the quantizer may be incremented/decre‐
             mented between frames (default: 4)

      (no)mbtree
             Enable macroblock tree ratecontrol (default:  enabled).   Use  a
             large lookahead to track temporal propagation of data and weight
             quality accordingly.  In multi-pass mode, this writes to a sepa‐
             rate stats file named <passlogfile>.mbtree.

      rc_lookahead=<0-250>
             Adjust the mbtree lookahead distance (default: 40).  Larger val‐
             ues will be slower and cause x264 to consume  more  memory,  but
             can yield higher quality.

      ratetol=<0.1-100.0> (ABR or two pass)
             allowed  variance  in average bitrate (no particular units) (de‐
             fault: 1.0)

      vbv_maxrate=<value> (ABR or two pass)
             maximum local bitrate, in kbits/second (default: disabled)

      vbv_bufsize=<value> (ABR or two pass)
             averaging period for vbv_maxrate, in kbits (default: none,  must
             be specified if vbv_maxrate is enabled)

      vbv_init=<0.0-1.0> (ABR or two pass)
             initial buffer occupancy, as a fraction of vbv_bufsize (default:
             0.9)

      ip_factor=<value>
             quantizer factor between I- and P-frames (default: 1.4)

      pb_factor=<value>
             quantizer factor between P- and B-frames (default: 1.3)

      qcomp=<0-1> (ABR or two pass)
             quantizer compression (default: 0.6).  A lower value  makes  the
             bitrate  more constant, while a higher value makes the quantiza‐
             tion parameter more constant.

      cplx_blur=<0-999> (two pass only)
             Temporal blur of the estimated frame  complexity,  before  curve
             compression  (default:  20).   Lower  values allow the quantizer
             value to jump around more, higher values force it to  vary  more
             smoothly.   cplx_blur ensures that each I-frame has quality com‐
             parable to the following P-frames, and ensures that  alternating
             high  and  low complexity frames (e.g. low fps animation) do not
             waste bits on fluctuating quantizer.

      qblur=<0-99> (two pass only)
             Temporal blur of the quantization parameter,  after  curve  com‐
             pression (default: 0.5).  Lower values allow the quantizer value
             to jump around more, higher values force it to vary more smooth‐
             ly.

      zones=<zone0>[/<zone1>[/...]]
             User  specified  quality  for  specific  parts (ending, credits,
             ...).  Each zone is <start-frame>,<end-frame>,<option> where op‐
             tion may be
                q=<0-51>
                     quantizer
                b=<0.01-100.0>
                     bitrate multiplier
             NOTE: The quantizer option is not strictly enforced.  It affects
             only the planning stage of ratecontrol, and is still subject  to
             overflow compensation and qp_min/qp_max.

      direct_pred=<name>
             Determines  the  type  of motion prediction used for direct mac‐
             roblocks in B-frames.
                none Direct macroblocks are not used.
                spatial
                     Motion vectors are extrapolated from neighboring blocks.
                     (default)
                temporal
                     Motion  vectors  are  extrapolated from the following P-
                     frame.
                auto The codec selects between spatial and temporal for  each
                     frame.
             Spatial  and temporal are approximately the same speed and PSNR,
             the choice between them depends on the video content.   Auto  is
             slightly  better,  but slower.  Auto is most effective when com‐
             bined with multipass.  direct_pred=none is both slower and lower
             quality.

      weightp
             Weighted P-frame prediction mode (default: 2).
                0    disabled (fastest)
                1    weighted refs (better quality)
                2    weighted refs + duplicates (best)

      (no)weight_b
             Use weighted prediction in B-frames.  Without this option, bidi‐
             rectionally predicted macroblocks give equal weight to each ref‐
             erence  frame.   With this option, the weights are determined by
             the temporal position of the B-frame relative to the references.
             Requires bframes > 1.

      partitions=<list>
             Enable     some     optional    macroblock    types    (default:
             p8x8,b8x8,i8x8,i4x4).
                p8x8 Enable types p16x8, p8x16, p8x8.
                p4x4 Enable types p8x4, p4x8, p4x4.  p4x4 is recommended only
                     with subq >= 5, and only at low resolutions.
                b8x8 Enable types b16x8, b8x16, b8x8.
                i8x8 Enable  type  i8x8.  i8x8 has no effect unless 8x8dct is
                     enabled.
                i4x4 Enable type i4x4.
                all  Enable all of the above types.
                none Disable all of the above types.
             Regardless of this option, macroblock types p16x16, b16x16,  and
             i16x16 are always enabled.
             The  idea is to find the type and size that best describe a cer‐
             tain area of the picture.  For example, a global pan  is  better
             represented by 16x16 blocks, while small moving objects are bet‐
             ter represented by smaller blocks.

      (no)8x8dct
             Adaptive spatial transform size: allows  macroblocks  to  choose
             between  4x4 and 8x8 DCT.  Also allows the i8x8 macroblock type.
             Without this option, only 4x4 DCT is used.

      me=<name>
             Select fullpixel motion estimation algorithm.
                dia  diamond search, radius 1 (fast)
                hex  hexagon search, radius 2 (default)
                umh  uneven multi-hexagon search (slow)
                esa  exhaustive search (very slow, and no better than umh)

      me_range=<4-64>
             radius of exhaustive or multi-hexagon  motion  search  (default:
             16)

      subq=<0-11>
             Adjust subpel refinement quality.  This parameter controls qual‐
             ity versus speed tradeoffs involved in the motion estimation de‐
             cision  process.   subq=5  can  compress  up  to 10% better than
             subq=1.
                0    Runs fullpixel precision motion estimation on all candi‐
                     date  macroblock types.  Then selects the best type with
                     SAD metric (faster than subq=1, not  recommended  unless
                     you're looking for ultra-fast encoding).
                1    Does  as 0, then refines the motion of that type to fast
                     quarterpixel precision (fast).
                2    Runs halfpixel precision motion estimation on all candi‐
                     date  macroblock types.  Then selects the best type with
                     SATD metric.  Then refines the motion of  that  type  to
                     fast quarterpixel precision.
                3    As 2, but uses a slower quarterpixel refinement.
                4    Runs  fast  quarterpixel  precision motion estimation on
                     all candidate macroblock types.  Then selects  the  best
                     type  with  SATD metric.  Then finishes the quarterpixel
                     refinement for that type.
                5    Runs best quality quarterpixel precision motion  estima‐
                     tion on all candidate macroblock types, before selecting
                     the best type.  Also refines the two motion vectors used
                     in  bidirectional  macroblocks  with SATD metric, rather
                     than reusing  vectors  from  the  forward  and  backward
                     searches.
                6    Enables rate-distortion optimization of macroblock types
                     in I- and P-frames.
                7    Enables rate-distortion optimization of macroblock types
                     in all frames (default).
                8    Enables  rate-distortion  optimization of motion vectors
                     and intra prediction modes in I- and P-frames.
                9    Enables rate-distortion optimization of  motion  vectors
                     and intra prediction modes in all frames.
                10   QP-RD;   requires  trellis=2  and  aq_mode=1  or  higher
                     (best).
                11   Full RD; disable all early terminations.
             In the above, "all candidates" does not exactly mean all enabled
             types: 4x4, 4x8, 8x4 are tried only if 8x8 is better than 16x16.

      (no)chroma_me
             Takes  into  account  chroma  information during subpixel motion
             search (default: enabled).  Requires subq>=5.

      (no)mixed_refs
             Allows each 8x8 or 16x8 motion partition to independently select
             a reference frame.  Without this option, a whole macroblock must
             use the same reference.  Requires frameref>1.

      trellis=<0-2> (cabac only)
             rate-distortion optimal quantization
                0    disabled
                1    enabled only for the final encode (default)
                2    enabled  during  all  mode  decisions  (slow,   requires
                     subq>=6)

      psy-rd=rd[,trell]
             Sets the strength of the psychovisual optimization.
                rd=<0.0-10.0>
                     psy  optimization  strength (requires subq>=6) (default:
                     1.0)
                trell=<0.0-10.0>
                     trellis (requires trellis, experimental) (default: 0.0)

      (no)psy
             Enable psychovisual optimizations that hurt PSNR  and  SSIM  but
             ought to look better (default: enabled).

      deadzone_inter=<0-32>
             Set  the  size  of the inter luma quantization deadzone for non-
             trellis quantization (default: 21).  Lower values help  to  pre‐
             serve fine details and film grain (typically useful for high bi‐
             trate/quality encode), while higher values help filter out these
             details  to  save  bits  that  can  be spent again on other mac‐
             roblocks and frames (typically useful  for  bitrate-starved  en‐
             codes).   It  is  recommended  that  you start by tweaking dead‐
             zone_intra before changing this parameter.

      deadzone_intra=<0-32>
             Set the size of the intra luma quantization  deadzone  for  non-
             trellis  quantization  (default:  11).  This option has the same
             effect as deadzone_inter except that it  affects  intra  frames.
             It  is recommended that you start by tweaking this parameter be‐
             fore changing deadzone_inter.

      (no)fast_pskip
             Performs early skip detection in  P-frames  (default:  enabled).
             This  usually  improves  speed  at no cost, but it can sometimes
             produce artifacts in areas with no details, like sky.

      (no)dct_decimate
             Eliminate dct blocks in P-frames containing only a small  single
             coefficient  (default: enabled).  This will remove some details,
             so it will save bits that can be spent again  on  other  frames,
             hopefully  raising  overall subjective quality.  If you are com‐
             pressing non-anime content with a high target bitrate,  you  may
             want to disable this to preserve as much detail as possible.

      nr=<0-100000>
             Noise  reduction,  0 means disabled.  100-1000 is a useful range
             for typical content, but you may want to turn it up a  bit  more
             for  very noisy content (default: 0).  Given its small impact on
             speed, you might want to prefer to use this over filtering noise
             away with video filters like denoise3d or hqdn3d.

      chroma_qp_offset=<-12-12>
             Use  a different quantizer for chroma as compared to luma.  Use‐
             ful values are in the range <-2-2> (default: 0).

      aq_mode=<0-2>
             Defines how adaptive quantization (AQ) distributes bits:
                0    disabled
                1    Avoid moving bits between frames.
                2    Move bits between frames (by default).

      aq_strength=<positive float value>
             Controls how much adaptive quantization  (AQ)  reduces  blocking
             and blurring in flat and textured areas (default: 1.0).  A value
             of 0.5 will lead to weak AQ and less details, when  a  value  of
             1.5 will lead to strong AQ and more details.

      cqm=<flat|jvt|<filename>>
             Either  uses  a predefined custom quantization matrix or loads a
             JM format matrix file.
                flat
                     Use the predefined flat 16 matrix (default).
                jvt
                     Use the predefined JVT matrix.
                <filename>
                     Use the provided JM format matrix file.
             NOTE: Windows CMD.EXE users may experience problems with parsing
             the command line if they attempt to use all the CQM lists.  This
             is due to a command line length limitation.  In this case it  is
             recommended the lists be put into a JM format CQM file and load‐
             ed as specified above.

      cqm4iy=<list> (also see cqm)
             Custom 4x4 intra luminance matrix, given as a list of  16  comma
             separated values in the 1-255 range.

      cqm4ic=<list> (also see cqm)
             Custom 4x4 intra chrominance matrix, given as a list of 16 comma
             separated values in the 1-255 range.

      cqm4py=<list> (also see cqm)
             Custom 4x4 inter luminance matrix, given as a list of  16  comma
             separated values in the 1-255 range.

      cqm4pc=<list> (also see cqm)
             Custom 4x4 inter chrominance matrix, given as a list of 16 comma
             separated values in the 1-255 range.

      cqm8iy=<list> (also see cqm)
             Custom 8x8 intra luminance matrix, given as a list of  64  comma
             separated values in the 1-255 range.

      cqm8py=<list> (also see cqm)
             Custom  8x8  inter luminance matrix, given as a list of 64 comma
             separated values in the 1-255 range.

      level_idc=<10-51>
             Set the bitstream's level as defined by annex  A  of  the  H.264
             standard  (default:  51  - level 5.1).  This is used for telling
             the decoder what capabilities it needs to support.  Use this pa‐
             rameter  only  if you know what it means, and you have a need to
             set it.

      (no)cpu_independent
             Ensure exact reproducibility across different  CPUs  instead  of
             chosing  different algorithms when available/better (default:en‐
             abled).

      threads=<0-16>
             Spawn threads to encode in parallel on multiple  CPUs  (default:
             0).   This  has  a  slight penalty to compression quality.  0 or
             'auto' tells x264 to detect how many CPUs you have and  pick  an
             appropriate number of threads.

      (no)sliced_threads
             Use  slice-based  threading  (default: disabled).  Unlike normal
             threading, this option adds no encoding latency, but is slightly
             slower and less effective at compression.

      slice_max_size=<0 or positive integer>
             Maximum  slice size in bytes (default: 0).  A value of zero dis‐
             ables the maximum.

      slice_max_mbs=<0 or positive integer>
             Maximum slice size in number of  macroblocks  (default:  0).   A
             value of zero disables the maximum.

      slices=<0 or positive integer>
             Maximum number of slices per frame (default: 0).  A value of ze‐
             ro disables the maximum.

      sync_lookahead=<0-250>
             Adjusts the size of the threaded lookahead buffer (default:  0).
             0 or 'auto' tells x264 to automatically determine buffer size.

      (no)deterministic
             Use only deterministic optimizations with multithreaded encoding
             (default: enabled).

      (no)global_header
             Causes SPS and PPS to appear only once, at the beginning of  the
             bitstream  (default:  disabled).  Some players, such as the Sony
             PSP, require the use of this option.  The default behavior caus‐
             es SPS and PPS to repeat prior to each IDR frame.

      (no)tff
             Enable interlaced mode, top field first (default: disabled)

      (no)bff
             Enable interlaced mode, bottom field first (default: disabled)

      nal_hrd=<none|vbr|cbr>
             Signal HRD information (requires vbv_bufsize) (default: none).

      (no)pic_struct
             Force pic_struct in Picture Timing SEI (default: disabled).

      (no)constrained_intra
             Enable  constrained  intra prediction (default: disabled).  This
             significantly reduces compression, but is required for the  base
             layer of SVC encodes.

      output_csp=<i420|i422|i444|rgb>
             Specify output colorspace (default: i420).

      (no)aud
             Write  access unit delimeters to the stream (default: disabled).
             Enable this only if your target container format requires access
             unit delimiters.

      overscan=<undef|show|crop>
             Include  VUI  overscan  information in the stream (default: dis‐
             abled).  See doc/vui.txt in the x264 source code for more infor‐
             mation.

      videoformat=<component|pal|ntsc|secam|mac|undef>
             Include  VUI  video  format  information in the stream (default:
             disabled).  This is a purely informative setting for  describing
             the  original  source.   See doc/vui.txt in the x264 source code
             for more information.

      (no)fullrange
             Include VUI full range information in the stream (default:  dis‐
             abled).   Use this option if your source video is not range lim‐
             ited.  See doc/vui.txt in the x264 source code for more informa‐
             tion.

      colorprim=<bt709|bt470m|bt470bg|smpte170m|smpte240m|film|undef>
             Include  color  primaries information (default: disabled).  This
             can be used for color correction.  See doc/vui.txt in  the  x264
             source code for more information.

      transfer=<bt709|bt470m|bt470bg|lin‐
      ear|log100|log316|smpte170m|smpte240m>
             Include VUI transfer characteristics information in  the  stream
             (default:  disabled).   This  can  be used for color correction.
             See doc/vui.txt in the x264 source code for more information.

      colormatrix=<bt709|fcc|bt470bg|smpte170m|smpte240m|GBR|YCgCo>
             Include VUI matrix coefficients in  the  stream  (default:  dis‐
             abled).  This can be used for color correction.  See doc/vui.txt
             in the x264 source code for more information.

      chromaloc=<0-5>
             Include VUI chroma sample location  information  in  the  stream
             (default: disabled).  Use this option to ensure alignment of the
             chroma and luma  planes  after  color  space  conversions.   See
             doc/vui.txt in the x264 source code for more information.

      log=<-1-3>
             Adjust the amount of logging info printed to the screen.
                -1   none
                 0   Print errors only.
                 1   warnings
                 2   PSNR  and other analysis statistics when the encode fin‐
                     ishes (default)
                 3   PSNR, QP, frametype, size, and other statistics for  ev‐
                     ery frame

      (no)psnr
             Print signal-to-noise ratio statistics.
             NOTE:  The  'Y',  'U', 'V', and 'Avg' PSNR fields in the summary
             are not mathematically sound (they are  simply  the  average  of
             per-frame  PSNRs).   They are kept only for comparison to the JM
             reference codec.  For all other purposes, please use either  the
             'Global' PSNR, or the per-frame PSNRs printed by log=3.

      (no)ssim
             Print  the Structural Similarity Metric results.  This is an al‐
             ternative to PSNR, and may be better correlated  with  the  per‐
             ceived quality of the compressed video.

      (no)visualize
             Enable x264 visualizations during encoding.  If the x264 on your
             system supports it, a new window will be opened during  the  en‐
             coding  process,  in which x264 will attempt to present an over‐
             view of how each frame gets encoded.  Each block type on the vi‐
             sualized movie will be colored as follows:

      dump_yuv=<file name>
             Dump YUV frames to the specified file.  For debugging use.
                red/pink
                     intra block
                blue
                     inter block
                green
                     skip block
                yellow
                     B-block
             This  feature  can  be  considered  experimental  and subject to
             change.  In particular, it depends on x264 being  compiled  with
             visualizations  enabled.   Note  that  as  of writing this, x264
             pauses after encoding and visualizing each  frame,  waiting  for
             the  user  to press a key, at which point the next frame will be
             encoded.

  xvfw (-xvfwopts)
      Encoding with Video for Windows codecs is mostly  obsolete  unless  you
      wish to encode to some obscure fringe codec.

      codec=<name>
             The name of the binary codec file with which to encode.

      compdata=<file>
             The name of the codec settings file (like firstpass.mcf) created
             by vfw2menc.

  MPEG muxer (-mpegopts)
      The MPEG muxer can generate 5 types of streams, each of which has  rea‐
      sonable default parameters that the user can override.  Generally, when
      generating MPEG files, it is advisable to disable MEncoder's frame-skip
      code  (see  -noskip, -mc as well as the harddup and softskip video fil‐
      ters).

      EXAMPLE:
                format=mpeg2:tsaf:vbitrate=8000

      format=<mpeg1 | mpeg2 | xvcd | xsvcd | dvd | pes1 | pes2>
             stream format (default: mpeg2).  pes1 and pes2 are  very  broken
             formats  (no  pack header and no padding), but VDR uses them; do
             not choose them unless you know exactly what you are doing.

      size=<up to 65535>
             Pack size in bytes, do not change unless you know  exactly  what
             you are doing (default: 2048).

      muxrate=<int>
             Nominal  muxrate  in  kbit/s  used in the pack headers (default:
             1800 kb/s).  Will be updated as necessary in the case  of  'for‐
             mat=mpeg1' or 'mpeg2'.

      tsaf
             Sets  timestamps  on  all  frames, if possible; recommended when
             format=dvd.  If dvdauthor complains with a message like "..audio
             sector  out  of  range...", you probably did not enable this op‐
             tion.

      interleaving2
             Uses a better algorithm to interleave audio and  video  packets,
             based  on  the  principle that the muxer will always try to fill
             the stream with the largest percentage of free space.

      vdelay=<1-32760>
             Initial video delay time, in milliseconds (default: 0),  use  it
             if  you  want  to delay video with respect to audio.  It doesn't
             work with :drop.

      adelay=<1-32760>
             Initial audio delay time, in milliseconds (default: 0),  use  it
             if you want to delay audio with respect to video.

      drop
             When used with vdelay the muxer drops the part of audio that was
             anticipated.

      vwidth, vheight=<1-4095>
             Set the video width and height when video is MPEG-1/2.

      vpswidth, vpsheight=<1-4095>
             Set pan and scan video width and height when video is MPEG-2.

      vaspect=<1 | 4/3 | 16/9 | 221/100>
             Sets the display aspect ratio for MPEG-2 video.  Do not  use  it
             on  MPEG-1  or  the  resulting  aspect  ratio will be completely
             wrong.

      vbitrate=<int>
             Sets the video bitrate in kbit/s for MPEG-1/2 video.

      vframerate=<24000/1001 | 24 | 25 | 30000/1001 | 30 | 50 | 60000/1001  |
      60 >
             Sets  the framerate for MPEG-1/2 video.  This option will be ig‐
             nored if used with the telecine option.

      telecine
             Enables 3:2 pulldown soft telecine mode: The muxer will make the
             video stream look like it was encoded at 30000/1001 fps.  It on‐
             ly  works  with  MPEG-2  video  when  the  output  framerate  is
             24000/1001  fps,  convert it with -ofps if necessary.  Any other
             framerate is incompatible with this option.

      film2pal
             Enables FILM to PAL and NTSC to PAL soft telecine mode: The mux‐
             er  will  make  the  video stream look like it was encoded at 25
             fps.  It only works with MPEG-2 video when the output  framerate
             is 24000/1001 fps, convert it with -ofps if necessary.  Any oth‐
             er framerate is incompatible with this option.

      tele_src and tele_dest
             Enables arbitrary telecining  using  Donand  Graft's  DGPulldown
             code.   You need to specify the original and the desired framer‐
             ate; the muxer will make the video stream look like it  was  en‐
             coded at the desired framerate.  It only works with MPEG-2 video
             when the input framerate is smaller than  the  output  framerate
             and the framerate increase is <= 1.5.

             EXAMPLE:
                tele_src=25,tele_dest=30000/1001
                     PAL to NTSC telecining

      vbuf_size=<40-1194>
             Sets  the size of the video decoder's buffer, expressed in kilo‐
             bytes.  Specify it only if the bitrate of the  video  stream  is
             too  high  for  the chosen format and if you know perfectly well
             what you are doing.  A too high value may lead to an  unplayable
             movie, depending on the player's capabilities.  When muxing HDTV
             video a value of 400 should suffice.

      abuf_size=<4-64>
             Sets the size of the audio decoder's buffer, expressed in  kilo‐
             bytes.  The same principle as for vbuf_size applies.

  FFmpeg libavformat demuxers (-lavfdopts)
      analyzeduration=<value>
             Maximum length in seconds to analyze the stream properties.

      format=<value>
             Force a specific libavformat demuxer.

      o=<key>=<value>[,<key>=<value>[,...]]
             Pass  AVOptions  to  libavformat demuxer.  Note, a patch to make
             the o= unneeded and pass all unknown options through  the  AVOp‐
             tion  system  is welcome.  A full list of AVOptions can be found
             in the FFmpeg manual.  Note that some options may conflict  with
             MPlayer/MEncoder options.

             EXAMPLE:
                o=ignidx

      probesize=<value>
             Maximum  amount of data to probe during the detection phase.  In
             the case of MPEG-TS this value identifies the maximum number  of
             TS packets to scan.

      cryptokey=<hexstring>
             Encryption  key  the demuxer should use.  This is the raw binary
             data of the key converted to a hexadecimal string.

  FFmpeg libavformat muxers (-lavfopts) (also see -of lavf)
      delay=<value>
             Currently only meaningful for  MPEG[12]:  Maximum  allowed  dis‐
             tance,  in  seconds,  between  the reference timer of the output
             stream (SCR) and the decoding timestamp  (DTS)  for  any  stream
             present (demux to decode delay).  Default is 0.7 (as mandated by
             the standards defined by MPEG).  Higher  values  require  larger
             buffers and must not be used.

      format=<container_format>
             Override which container format to mux into (default: autodetect
             from output file extension).
                mpg
                     MPEG-1 systems and MPEG-2 PS
                asf
                     Advanced Streaming Format
                avi
                     Audio Video Interleave file
                wav
                     Waveform Audio
                swf
                     Macromedia Flash
                flv
                     Macromedia Flash video files
                rm
                     RealAudio and RealVideo
                au
                     SUN AU format
                nut
                     NUT open container format (experimental)
                mov
                     QuickTime
                mp4
                     MPEG-4 format
                ipod
                     MPEG-4 format with extra header flags required by  Apple
                     iPod firmware
                dv
                     Sony Digital Video container
                matroska
                     Matroska

      muxrate=<rate>
             Nominal  bitrate of the multiplex, in bits per second; currently
             it is meaningful only for MPEG[12].   Sometimes  raising  it  is
             necessary in order to avoid "buffer underflows".

      o=<key>=<value>[,<key>=<value>[,...]]
             Pass  AVOptions to libavformat muxer.  Note, a patch to make the
             o= unneeded and pass all unknown options  through  the  AVOption
             system is welcome.  A full list of AVOptions can be found in the
             FFmpeg manual.  Note that some options may  conflict  with  MEn‐
             coder options.

             EXAMPLE:
                o=packetsize=100

      packetsize=<size>
             Size,  expressed  in bytes, of the unitary packet for the chosen
             format.  When muxing to  MPEG[12]  implementations  the  default
             values are: 2324 for [S]VCD, 2048 for all others formats.

      preload=<distance>
             Currently  only  meaningful  for  MPEG[12]: Initial distance, in
             seconds, between the reference timer of the output stream  (SCR)
             and  the  decoding timestamp (DTS) for any stream present (demux
             to decode delay).

ENVIRONMENT VARIABLES
      There are a number of environment variables that can be used to control
      the behavior of MPlayer and MEncoder.

      MPLAYER_CHARSET (also see -msgcharset)
             Convert  console messages to the specified charset (default: au‐
             todetect).  A value of "noconv" means no conversion.

      MPLAYER_HOME
             Directory where MPlayer looks for user settings.

      MPLAYER_VERBOSE (also see -v and -msglevel)
             Set the initial verbosity level across all message modules  (de‐
             fault:  0).  The resulting verbosity corresponds to that of -ms‐
             glevel 5 plus the value of MPLAYER_VERBOSE.

  libaf:
      LADSPA_PATH
             If LADSPA_PATH is set, it searches for the specified  file.   If
             it  is  not  set,  you  must  supply a fully specified pathname.
             FIXME: This is also mentioned in the ladspa section.

  libdvdcss:
      DVDCSS_CACHE
             Specify a directory in which to store title  key  values.   This
             will  speed up descrambling of DVDs which are in the cache.  The
             DVDCSS_CACHE directory is created if it does not  exist,  and  a
             subdirectory  is created named after the DVD's title or manufac‐
             turing date.  If DVDCSS_CACHE is not set or is empty,  libdvdcss
             will  use  the  default  value which is "${HOME}/.dvdcss/" under
             Unix and "C:\Documents and Settings\$USER\Application Data\dvdc‐
             ss\" under Win32.  The special value "off" disables caching.

      DVDCSS_METHOD
             Sets  the  authentication  and  decryption method that libdvdcss
             will use to read scrambled discs.  Can be one of title,  key  or
             disc.
                key
                     is the default method.  libdvdcss will use a set of cal‐
                     culated player keys to try and get the disc  key.   This
                     can  fail  if  the  drive  does not recognize any of the
                     player keys.
                disc
                     is a fallback method when key has  failed.   Instead  of
                     using player keys, libdvdcss will crack the disc key us‐
                     ing a brute force algorithm.  This process is CPU inten‐
                     sive and requires 64 MB of memory to store temporary da‐
                     ta.
                title
                     is the fallback when all other methods have failed.   It
                     does  not rely on a key exchange with the DVD drive, but
                     rather uses a crypto attack to guess the title key.   On
                     rare cases this may fail because there is not enough en‐
                     crypted data on the disc to perform  a  statistical  at‐
                     tack,  but  in  the other hand it is the only way to de‐
                     crypt a DVD stored on a hard disc, or  a  DVD  with  the
                     wrong region on an RPC2 drive.

      DVDCSS_RAW_DEVICE
             Specify  the raw device to use.  Exact usage will depend on your
             operating system, the Linux utility to set  up  raw  devices  is
             raw(8)  for  instance.   Please note that on most operating sys‐
             tems, using a raw device requires highly aligned buffers:  Linux
             requires a 2048 bytes alignment (which is the size of a DVD sec‐
             tor).

      DVDCSS_VERBOSE
             Sets the libdvdcss verbosity level.
                0    Outputs no messages at all.
                1    Outputs error messages to stderr.
                2    Outputs error messages and debug messages to stderr.

      DVDREAD_NOKEYS
             Skip retrieving all keys on startup.  Currently disabled.

      HOME   FIXME: Document this.

  libao2:
      AO_SUN_DISABLE_SAMPLE_TIMING
             FIXME: Document this.

      AUDIODEV
             FIXME: Document this.

      AUDIOSERVER
             Specifies the Network Audio System server to which the nas audio
             output  driver  should  connect and the transport that should be
             used.  If unset DISPLAY is used instead.  The transport  can  be
             one  of  tcp  and  unix.   Syntax  is tcp/<somehost>:<someport>,
             <somehost>:<instancenumber> or [unix]:<instancenumber>.  The NAS
             base port is 8000 and <instancenumber> is added to that.

             EXAMPLES:
                AUDIOSERVER=somehost:0
                     Connect to NAS server on somehost using default port and
                     transport.
                AUDIOSERVER=tcp/somehost:8000
                     Connect to NAS server on somehost listening on TCP  port
                     8000.
                AUDIOSERVER=(unix)?:0
                     Connect to NAS server instance 0 on localhost using unix
                     domain sockets.

      DISPLAY
             FIXME: Document this.

  vidix:
      VIDIX_CRT
             FIXME: Document this.

      VIDIXIVTVALPHA
             Set this to 'disable' in order to stop  the  VIDIX  driver  from
             controlling  alphablending settings.  You can then manipulate it
             yourself with 'ivtvfbctl'.

  osdep:
      TERM   FIXME: Document this.

  libvo:
      DISPLAY
             FIXME: Document this.

      FRAMEBUFFER
             FIXME: Document this.

      HOME   FIXME: Document this.

  libmpdemux:
      HOME   FIXME: Document this.

      HOMEPATH
             FIXME: Document this.

      http_proxy
             FIXME: Document this.

      LOGNAME
             FIXME: Document this.

      USERPROFILE
             FIXME: Document this.

  GUI:
      DISPLAY
             The name of the display to which the GUI should connect.

      HOME   The home directory of the current user.

  libavformat:
      AUDIO_FLIP_LEFT
             FIXME: Document this.

      BKTR_DEV
             FIXME: Document this.

      BKTR_FORMAT
             FIXME: Document this.

      BKTR_FREQUENCY
             FIXME: Document this.

      http_proxy
             FIXME: Document this.

      no_proxy
             FIXME: Document this.

FILES
      /usr/local/etc/mplayer/mplayer.conf
             MPlayer system-wide settings

      /usr/local/etc/mplayer/mencoder.conf
             MEncoder system-wide settings

      ~/.mplayer/config
             MPlayer user settings

      ~/.mplayer/mencoder.conf
             MEncoder user settings

      ~/.mplayer/input.conf
             input bindings (see '-input keylist' for the full list)

      ~/.mplayer/gui.conf
             GUI configuration file

      ~/.mplayer/gui.gain
             for audio files not containing ReplayGain data, add a line  with
             replay gain and filename separated by a space character, e.g.

             +1.50 /home/me/music/Song.mp3

      ~/.mplayer/gui.history
             GUI directory history

      ~/.mplayer/gui.pl
             GUI playlist

      ~/.mplayer/gui.url
             GUI URL list

      ~/.mplayer/font/
             font  directory  (There  must be a font.desc file and files with
             .RAW extension.)

      ~/.mplayer/DVDkeys/
             cached CSS keys

EXAMPLES OF MPLAYER USAGE
      Quickstart Blu-ray playing:
      mplayer br:////path/to/disc
      mplayer br:// -bluray-device /path/to/disc

      Quickstart DVD playing:
      mplayer dvd://1

      Play in Japanese with English subtitles:
      mplayer dvd://1 -alang ja -slang en

      Play only chapters 5, 6, 7:
      mplayer dvd://1 -chapter 5-7

      Play only titles 5, 6, 7:
      mplayer dvd://5-7

      Play a multiangle DVD:
      mplayer dvd://1 -dvdangle 2

      Play from a different DVD device:
      mplayer dvd://1 -dvd-device /dev/dvd2

      Play DVD video from a directory with VOB files:
      mplayer dvd://1 -dvd-device /path/to/directory/

      Copy a DVD title to hard disk, saving to file title1.vob :
      mplayer dvd://1 -dumpstream -dumpfile title1.vob

      Play a DVD with dvdnav from path /dev/sr1:
      mplayer dvdnav:////dev/sr1

      Stream from HTTP:
      mplayer http://mplayer.hq/example.avi

      Stream using RTSP:
      mplayer rtsp://server.example.com/streamName

      Convert subtitles to MPsub format:
      mplayer dummy.avi -sub source.sub -dumpmpsub

      Convert subtitles to MPsub format without watching the movie:
      mplayer /dev/zero -rawvideo pal:fps=xx -demuxer rawvideo -vc null -vo null -noframedrop -benchmark -sub source.sub -dumpmpsub

      input from standard V4L:
      mplayer tv:// -tv driver=v4l:width=640:height=480:outfmt=i420 -vc rawi420 -vo xv

      Playback on Zoran cards (old style, deprecated):
      mplayer -vo zr -vf scale=352:288 file.avi

      Playback on Zoran cards (new style):
      mplayer -vo zr2 -vf scale=352:288,zrmjpeg file.avi

      Play DTS-CD with passthrough:
      mplayer -ac hwdts -rawaudio format=0x2001 -cdrom-device /dev/cdrom cdda://
      You can also use -afm hwac3 instead of -ac hwdts.  Adjust  '/dev/cdrom'
      to  match  the CD-ROM device on your system.  If your external receiver
      supports decoding raw DTS streams, you can directly play it via cdda://
      without setting format, hwac3 or hwdts.

      Play a 6-channel AAC file with only two speakers:
      mplayer -rawaudio format=0xff -demuxer rawaudio -af pan=2:.32:.32:.39:.06:.06:.39:.17:-.17:-.17:.17:.33:.33 adts_he-aac160_51.aac
      You  might  want to play a bit with the pan values (e.g multiply with a
      value) to increase volume or avoid clipping.

      checkerboard invert with geq filter:
      mplayer -vf geq='128+(p(X\,Y)-128)*(0.5-gt(mod(X/SW\,128)\,64))*(0.5-gt(mod(Y/SH\,128)\,64))*4'

EXAMPLES OF MENCODER USAGE
      Encode DVD title #2, only selected chapters:
      mencoder dvd://2 -chapter 10-15 -o title2.avi -oac copy -ovc lavc -lavcopts vcodec=mpeg4

      Encode DVD title #2, resizing to 640x480:
      mencoder dvd://2 -vf scale=640:480 -o title2.avi -oac copy -ovc lavc -lavcopts vcodec=mpeg4

      Encode DVD title #2, resizing to 512xHHH (keep aspect ratio):
      mencoder dvd://2 -vf scale -zoom -xy 512 -o title2.avi -oac copy -ovc lavc -lavcopts vcodec=mpeg4

      The same, but with bitrate set to 1800kbit and optimized macroblocks:
      mencoder dvd://2 -o title2.avi -oac copy -ovc lavc -lavcopts vcodec=mpeg4:mbd=1:vbitrate=1800

      The same, but with MJPEG compression:
      mencoder dvd://2 -o title2.avi -oac copy -ovc lavc -lavcopts vcodec=mjpeg:mbd=1:vbitrate=1800

      Encode all *.jpg files in the current directory:
      mencoder "mf://*.jpg" -mf fps=25 -o output.avi -ovc lavc -lavcopts vcodec=mpeg4

      Encode from a tuner (specify a format with -vf format):
      mencoder -tv driver=v4l:width=640:height=480 tv:// -o tv.avi -ovc raw

      Encode from a pipe:
      rar p test-SVCD.rar | mencoder -ovc lavc -lavcopts vcodec=mpeg4:vbitrate=800 -ofps 24 -

BUGS
      Don't panic.  If you find one, report it to us, but  please  make  sure
      you  have  read all of the documentation first.  Also look out for smi‐
      leys. :) Many bugs are the result of incorrect setup or  parameter  us‐
      age.  The bug reporting section of the documentation (http://www.mplay‐
      erhq.hu/DOCS/HTML/en/bugreports.html) explains how to create useful bug
      reports.

AUTHORS
      MPlayer  was initially written by Arpad Gereoffy.  See the AUTHORS file
      for a list of some of the many other contributors.

      MPlayer is (C) 2000-2022 The MPlayer Team

      This man page was written mainly by Gabucino, Jonas Jermann  and  Diego
      Biurrun.   It  is maintained by Diego Biurrun.  Please send mails about
      it to the MPlayer-DOCS mailing list.  Translation specific mails belong
      on the MPlayer-translations mailing list.



The MPlayer Project               2022-02-27                        MPlayer(1)