NAME
   Audio::Analyzer::ToneDetect - Detect freq of tones in an audio file or
   stream

SYNOPSIS
     use Audio::Analyzer::ToneDetect;
     my $tone_detect = Audio::Analyzer::ToneDetect->new( source => \*STDIN );
     my $tone = $tone_detect->get_next_tone();
     say "I heard $tone!";

DESCRIPTION
   Consider this alpha software. It is still under fairly active
   development and the interface may change in incompatible ways.

   Audio::Analyzer::ToneDetect is a module for detecting single frequency
   tones in an audio stream or file. It supports mono PCM data and defaults
   to STDIN. For supporting other formats, eg MP3, you can pipe things
   through sox.

USAGE
 new (%opts)
   Takes the following named parameters:

   source \*FH or $path
       The audio source. Only Mono PCM is supported. You can pass the path
       to a WAV file or a file handle for an open file or stream. Defaults
       to STDIN.

   sample_rate 16000
       Source sample rate, results will be orders of magnitude off if set
       incorrectly. Defaults to 16000.

   chunk_size 1024
       Number of samples to analyze at once. Corresponds to dft_size in
       Audio::Analyzer. Must be a power of 2. Defaults to 1024.

   chunk_max 70
       Maximum number of chunks to process before returning. Returns false
       if it reaches this number of chunks without detecting a tone. With
       default chunk_size and sample_rate, the default of 70 equates to
       about 4.5 seconds of audio.

   min_tone_length 0.5
       Minimum durration of a tone, in seconds, before we consider it
       detected. Due to sample rate, chunk size, and integer math, with
       defaults this ends up being 0.448 seconds. The formula for actual
       seconds is int( min_length * sample_rate / chunk_size ) * chunk_size
       / sample_rate. Default to 0.5

   valid_tones undef, 'builtin', or ARRAYREF
       A list of valid (expected) tones. If supplied, the closest expected
       tone for a given detected tone is returned. Call get_next_tone() in
       list context or supply the following call back if you want both
       values. A value of 'builtin' uses a builtin list of valid classic
       Motorola Minitor tones. Defaults to unset.

   valid_error_cb
       A callback that if provided and valid_tones is set will be called
       just before get_next_tone or find_closest_valid returns. Arguments
       are the closest valid tone, the actual detected tone, the diference
       between the two in Hertz.

       Example:

         valid_error_cb => sub { printf "VF %s DF %s EF %.2f\n", @_; return }

       Return value is expected to be one of three possibilities.

       undef
           Has no effect on program flow, if you don't want your call back
           changing stuff make sure you have an explicit 'return' as the
           last line.

       Zero (the number)
           A return value of 0 discards the tone and continues the
           get_next_tone loop.

       N   Any other value replaces the valid detected tone with the return
           value from the call back.

   rejected_freqs undef or ARRAYREF
       If specified, a reference to an array of frequencies that will be
       ignored. e.g roger beeps, repeater beeps, etc. Note, if you use
       valid tone detection, then this is the raw detected tone, not the
       closest match. Defaults to empty list.

 valid_tones
   Returns the arraref of valid tones currently being used. Optionally
   takes a reference to an array of new tones to use that will be copied to
   replace the current valid list.

 get_next_tone
   Returns the next detected tone in the stream. Will return false if we go
   through chunk_max without detecting a tone but the buffer will be
   preserved between calls if the a tone begins just before hitting
   chunk_max. If valid_tones was supplied, returns the result of passing
   the tone to find_closest_valid(), following it's list vs scalar
   semantics.

 get_next_two_tones
   Calls get next tone twice. Will return false if either tone returns
   false.

 find_closest_valid
   In scalar context, returns the closest valid tone in valid_tones. In
   list context returns the closest valid tone and the delta from detected
   tone.

AUTHOR
   Mike Greb <[email protected]>

COPYRIGHT
   Copyright 2013 - Mike Greb

LICENSE
   This library is free software; you can redistribute it and/or modify it
   under the same terms as Perl itself.

SEE ALSO
   Audio::Analyzer

   Math::FFT