I've had various ideas for displays over the years. Some are
vaguely practical, so I'll reserve them in case I do build one and
want to describe it on my non-anonymous website. So these are my
silly displays - of little practical value and requiring much
effort to create, but with a novelty that might find value amongst
the public (either as a one-off, or some sort of product) if
someone was in a position to build and publicise them.
SHADOW DOT DISPLAY
I won't deny it, this idea is the product of me standing in a dark
room playing with a bright torch. The observation at the heart of
the idea is that with a opaque ball suspended in front of a light
surface that then points at a screen, the angle from which the
light shines on the ball determines the position of the round
shadow on the screen.
To create a display you need the shadow of the ball (the "dot") to
appear to the eye in multiple places on the screen at once, to form
an image.
The screen would need to have some sort of phosphorescent coating
that would permit (over multiple "exposures" of the projected
image) the areas less exposed to light due to their coverage by the
shadow of the ball, to be visible as areas emitting less light when
the projector light is turned off. The idea is that the average
light absorbed in shadowed areas over the time required for the
scanning of the projector beam to complete, is less than those
areas never behind the shadow. The light source may need to be
switched on/off with a particular duty cycle during projection to
ensure that shadowed areas do not reach saturation of their light
absorbtion capability before scanning completes. Projection with
light wavelengths invisible to the eye, IR, UV, or even X-Ray,
could allow the display on the screen to be observed while the
projector continuously operates, rather than having to observe the
screen only after scanning is complete and the projection light it
turned off. Alternatively, the image could be seen through a
shutter that permits viewing only during the off part of the
projector light's duty cycle.
For an X-Y vector display, the angle of the projector light could
be moved directly to trace out lines as with a laser projector.
Depending of the size of the projector's light beam, additional
scanning without a shadow of areas not meant to be exposed, but not
illuminated by light surrounding any projected dot, may be required.
For a raster display, the angle of the projector light mechanism
could be scanned in a repeating pattern, but with the light only
turned on where the corresponding location of the dot forms a point
(black pixel) in the image. Again, additional exposure without a
dot may be required if not all areas of the screen are exposed due
to the limited number or position of the dots, and the limited
width of the projector light beam.
Additional effects could be possible by replacing the opaque ball
with glass structure exhibiting a lens effect.
Issues:
The exposure of areas surrounding a dot would have to be carefully
controlled where other nearby dots are at risk of being saturated.
The width of the light beam, as well as the exposure timing and,
the qualities of the phosphorescent coating, would be the key means
of controlling this. A limit to the total area that can be
displayed under visible shadow would naturally be reached due to
exposure of previously shadowed areas on each exposure required per
frame.
The update rate would obviously be very slow, though ideally the
transition between frames would fade in/out in a naturalistic
mannor.
WHITEBOARD DISPLAY
This is simple: An X-Y plotter is designed to write on a whiteboard
(nothing that unusual, 3D printable designs for such devices are
described on thingiverse.com), and also to erase parts previously
drawn. It fills an area, a camera takes a pictures, it erases and
redraws parts of the whiteboard, another picture is taken. You end
up with a time-lapse animation of things being drawn on the
whiteboard.
Of course that's already been done by humans (or at least those
vastly more accurate at drawing things than I am) drawing on
whiteboards, I'm pretty sure. What I really want is to create some
sort of whiteboard Matrix effect where maybe the camera moves up
the board so that the symbols look like they're moving down the
screen, while the X-Y machine redraws them randomly between frames.
I guess you could probably do that by hand too, but the machine
would be less frustrating for someone as artless as me.
DOMINO TV
Have a large frame with a board behind. The frame forms rows on
which dominos are positioned, each in front of hole in the board.
The holes are so positioned that when one domino falls, the hole is
uncovered and a light source shining from behind can be seen. But
when a domino on a row behind the front falls, it block s the path
of the light again.
The falling of the dominos, as well as their resetting in an
upright posistion, is mechanically controlled (maybe using
electromagnets?). A computer controls the mechanism, and the
sequence of falling dominos esposes light in a pattern that
projects a raster image onto a screen.
Multiple rows permit lights to be turned on and off again in rapid
succession, preferably so arranged that the time for the first
domino to be stood up again automatically, comes within one frame's
duration of the domino in the last row falling after a light is
alternately turned on/off on every frame (the worst case).
The sound could be quite interesting too, maybe also able to be
used as a sort of musical instrument.
SIMPLIFIED VOLUMETRIC HOLOGRAPHIC DISPLAY
A vertical shaft is spun, along which are vertically strung strands
of wire on sprung, or otherwise stretchy, mounts at both ends (top
and bottom). At the middle of each length of wire is mounted an RGB
LED and possibly some additional weight. As the shaft spins faster,
the LEDs are pulled outwards from the shaft, permited by the
flexible moutings of their connecting wires.
The speed of the shaft alternates rapidly fast/slow, completing a
cycle of fastest-slowest within the duration of each frame.
Individual LEDs are lit at different distances away from the
central shaft, as well as at different points in their circular
trajectory, so as to form a volumetric image. Only one LED is
required for each pixel of vertical resolution, while horizontal
resolution is determined only be the accuracy and speed of the
circuitry controlling whether the LEDs are turned on or off.
Electrically then, this is simpler than other LED-based volumetric
display designs that rely on multiple layers of LEDs mounted in
fixed positions on spinning surfaces.
The variation in speed (slow to fast) during each frame could be
controlled by a cam system, and driven from a stepper motor for
accurate control and synchronisation of the overall speed / frame
rate.
Alternative to the LEDs: Fibre-optics fed out of the shaft, pulled
outwards as it spins like a straight bit of wire being spun.
Wouldn't maintain a fixed vertical location though.
Issues:
The speed required for the shaft to reach in order for the LEDs to
cover a reasonable horizontal resolution at numerous depth levels
would be very high. A rough calculation for 24fps and ten depth
levels suggested a speed of around 14,000RPM. Perhaps there are
solutions that could reduce this required speed, or make it easier
to control?
Derivative Idea:
For the really sci-fi: Ultrasonically levitated tiny LEDs
controlled and powered by induction and/or radio.
