3D TV With No Glasses and No Parallax/Focus Issues
Here’s how to make a 3D TV that doesn’t need glasses, and provides truly natural 3D with no parallax-vs-focus problems:
Recording
The camera has no lenses, and doesn’t need to focus. It is a rectangular, movie-screen shaped plane of light-sensing pixels that is aimed at the subject. Current videocameras have a pixel matrix in which each pixel measures the overall intensity of the red, green, or blue light that is hitting it, and sends information about that intensity to the recording computer. In this system, each pixel sends the exact waveform of the light that is hitting it to the recording computer.
Playback
To play back the video, your 3D TV simply reproduces the same waveform from each pixel. This creates wavefronts that exactly mimic the waves of light as they were striking the camera’s receiver, and the result will be that people sitting in front of the TV will see a perfect 3D image. They will be able to focus their eyes on different parts of the image, with parallax and focus always matching. No glasses of any kind will be needed. Off-center viewing angle will not create distortion — everything in the video will appear to be a true 3D object that you can view straight on, or from an off-center angle, just as if it were really there.
Data volume
One obvious problem with this system (besides the technology to create pixels that can emit light of a computer-specified waveform) is the potentially huge amount of data required to represent the image.
This problem might be substantially reduced by the following technique: Eliminate all but three frequencies of light (red, green, and blue), and make your system record (and playback) only those frequencies.
It is not enough for each sub-pixel (say, a red one), to simply generate a red-frequency sine wave of a certain amplitude and phase. For the system to work, each red sub-pixel must be able to generate the sum of multiple red-frequency sine waves, each of a different phase and amplitude. The number of such amplitude-phase pairs that would have to be recorded per sub-pixel per frame would not be immense — no matter how many separate red light sources were hitting this one pixel in the camera, the complexity of a waveform that is formed entirely by adding same-frequency sine waves of different amplitudes and phases is fairly limited. For the duration of a frame (say, a sixtieth of a second), the red-frequency waveform of these combined sine waves can be played back continuously from this sub-pixel.
Pixel synchronization
The playback pixels need to be able to synchronize their emitted waveforms to within a small fraction of the duration of a single wave. The shortest wavelength is blue, and a single cycle of the blue frequency lasts about 1.5 x 10-15 seconds. The pixels need to have their wave output synchronized to within a small fraction of that, so we’ll add three zeros and make it 1.5 x 10-18 seconds — this is the accuracy of synchronization that the pixels need. However, they do not need this degree of synchronization when changing from one frame to the next — they can do just about anything for a small fraction of one sixtieth of a second (but still much more than 1.5 x 10-18 seconds) in-between frames, and the human eye will not notice. But for the bulk of the duration of a frame, the pixels simply need to be able play their light waveforms in near-perfect synchronization with each other.
(Note: Only the sub-pixels of the same color — red, green, or blue — have to be in-sync. In other words, all the red sub-pixels have to be in perfect sync with each other, all the greens do, and all the blues do. But the reds do not have to be in sync with the greens, for example.)
Viewing angle
The viewing angle issue of 3D TV (and even 2D TV, which looks a little odd when viewed off-center) is perfectly eliminated — but this may actually be an undesirable effect. A viewer watching the screen at a 45° angle will actually be able to see things that another viewer does not see, and see things from a different angle, just as two people looking out a window from different angles will see substantially different portions of the exterior view. This may be somewhat disturbing to movie makers, who want greater control over what the user is looking at.
Focus
The same issue described above for viewing angle also exists for focus. In 2D movies, the director often deliberately makes only part of the image in-focus: the part the viewers are intended to be watching. With the above-described technology, the viewers can focus their eyes on anything in the scene at will (unless it is beyond their eyes’ focal range; which might be possible to engineer).
Zoom
I can’t think of how the effect of a zoom lens would be created using this technology. It might not be possible at all.
Polarization
Polarization will be lost when recording this way, just as it is in current forms of video recording.
Pixel max-out
If the incoming waveform of light at a particular pixel has a greater amplitude than can be recorded by the pixel’s mechanism, this could cause severe distortion of large portions of the image. In current video recording technology, the maxed-out area is simply represented with the brightest intensity those pixels can display, and viewers are accustomed to this. But with the 3D technology described here, there would be no way for the recording device to know the difference between a single, over-bright light source coming from one direction, versus multiple light sources coming from different directions that (at this pixel) add up to a max-out situation.
This problem apparently would necessitate pixels that can accurately record light intensities greater than would typically be encountered in a recording session, even taking into account that multiple light sources can add up at a single pixel.
Camera size
Camera size is a serious issue. If you don’t want your subject to look incredibly large, the camera (a flat panel, remember) will need to be quite large — perhaps the size of a typical big-screen TV.
Video conferencing
If a really large screen could be made that is both a camera and a TV, it could be used for superb videoconferencing.
Update 2011.06.24 — Lytro is a camera that may be doing something very similar to what I’m describing here. But it does have a lens system. Interesting.
© 2012 Darel Rex Finley | design by 