The 3D format has had something of a renaissance in recent years, but the technology still has some way to go before the potential of "real-life" multiperspective 3D can be realized. The Camera Culture group at the MIT Media Lab is developing a new 3D video projection system that doesn't require glasses and provides different users different perspective angles of the same object. The team sees it not as a final answer, but as a transitional system that sits between current technologies and true holographic video.
In one form or another, 3D projection systems have been around almost as long as the cinema itself. The trick has always been to come up with something practical and economical – preferably without the glasses, nausea, and headaches. Over the decades, the Bijous and Odeons of the world have seen two-color systems, polarized light systems, mechanical shutters and multiple projectors systems come and go as each fell short of the mark.
The MIT approach was to come up with a glasses-free video projection system with a wide vision angle, ultra-high resolution, yet is mechanically simple and doesn't require elaborate installations, as well as being cheaper than conventional holographic systems of comparable quality. The idea is that it will act as a short term, intermediate solution until a more mature technology can be developed, while making it attractive as a transitional technology for users of more conventional 2D systems.
The MIT system doesn't just produce an illusion of parallax – it creates an actual shift in perspective for multiple viewers looking at the image from different angles, as if were looking at real objects. In addition, it provides better resolution and contrast than conventional 2D video.
The heart of the projector is a pair of flat panels of liquid-crystal modulators that act like tiny liquid-crystal displays (LCD) set between the backlight source and the lens. The first bank of LCDs produces light patterns at particular angles. This pattern passes through the second bank, but only at these angles. These then pass through a series of lenses that are arranged in the same manner as a Keplerian telescope. The patterns then pass through a transparent projection screen made up of lines of vertical lenses, a bit like the striated transparent sheets found in children’s books and toys that shift between a pair of images. These resolve the patterns into a 3D image that shifts as one moves from one of eight points of view to another.
The modulators refresh the image 240 times every second, which is less than that of modern televisions, but ten times that of standard-speed film. The system requires a lot of bandwidth to work, but it also opens the possibility of the system displaying ultra-high resolution video. Part of the reason for this is that the MIT system uses data compression algorithms that concentrate on reproducing the edges of the images rather than the body of objects, because edges change more as an object moves or turns. The algorithm also produce a brighter image with stronger contrast by generating something closer to “true black, ” which is not possible for LCDs.
Another way the image is improved is by how the light patterns interfere with one another. By manipulating this, the system can gain higher resolutions, though this requires some prodigious real-time calculations. In addition, the MIT system doesn't simply enlarge the image, which reduces the field of vision in 3D projections. Instead, the system spreads out the pixels from one another.
The team sees the technology as having applications in collaborative design and medical imaging as well as in entertainment.