When we talk about “3DTV” we generally mean “stereoscopic” images. This approach presents different views to the left and right eyes. The brain then deciphers information from the small differences between the two images, and constructs a three-dimensional image from the combination. This can deliver the illusion of depth, where some objects appear to be closer and others farther away.
This does not successfully mimic all aspects of real-world vision, however. The most glaring omission is that it does not provide “motion parallax” effects. This is the effect that you see when you move your head from side to side; closer objects will “move” relative to the background, and elements of the background will either be covered up or revealed by this motion. It turns out that this is a very strong effect; it allows you to “see around” objects so that you can see their sides.
You can’t do this with stereoscopic imagery because there are only two images available. If you have more images from additional angles, you could create a hologram that would appear to have volume, but this is expensive and difficult and generally requires lasers. But now a group from Osaka University has demonstrated a system that uses simple front projectors to show multiple images. Instead of projecting onto a flat surface, the images are projected into a cloud of water droplets. Depending on your viewing angle, you will see a different image. As you walk around the cloud of mist, the image changes and it appears as though you are looking at a three-dimensional object.
Unlike stereoscopic imagery, you can capture this effect in a simple video. Here’s a recording of the demonstration:
This approach certainly could be applied to larger, full color displays. A stable “cloud screen” would be required, and it would probably have to rely on interpolation to create enough different images to give it a natural look, but it would definitely solve the multiple-viewer problems inherent in auto-stereoscopic displays.