Slashdot Mirror
The Future of Holograms
D3 writes "A Slate article talks about the failure of holograms to really catch on and the future of using computers to create true holographic video ala Princess Leia. The article covers some history such as the fact that holograms have been around since 1947. Lots of great geek-pop references as well."
Sega Hologram. I can't believe the article actually went there. At least they pointed out that it was not in fact a 3-D picture. If you don't believe me, try playing one where someone removed the colored blocks.
Speaking of holograms... for generating holograms without a laser (just your PC, a laser printer, and a transparency), check out the MedCosm CGHmaker.
Anyone know of a really hi-res output device?
I'm not sure you would need as much graphical processing power that 2D renditions of 3D scenes. A lot of the math involved is for the "camera" and answering the question "What would element X look like when viewed from angle Y?" If you're dealing with holograms, there is no "camera" angle to worry about, since that's determined by where your eyes are in The Real World.
Actually, DirectX, OpenGL, and other rendering engines used in most games today put a lot of processing power into converting 3d points to 2d screen points (i.e. rendering the 3d scene to a 2d surface).
I really don't think the computational power would be much extra, other than the physical beaming of lights in 3 dimensions rather than 2.
Tech, life, family, faith: Give me a visit
Real time, photorealistic holographic imaging is quite difficult. For one it requires more than just on color. Holograms are produced and re-created using monochromatic light sources. Not only you cannot have multiple colors you cannot even have different shades of the same color! Another complication is that for a sizeable holographic image you'll require substantial amounts of energy focused on relatively confined space. Your fire insurance premium are sure to rise faster than USS 1701D hops across the galaxy at warp 9.
Years ago I saw some work from Stanford (Bert Hesselink's lab, if I remember right) on volumetric displays. Basically they used a crystal as a "screen" for holographic projection. The density of the crystal was better than that of air and it represented a stable medium (compared to water mist of other vapors) to project a hologram. It sounds like smoke and mirrors but it was quite impressive and you could see the hologram in normal light conditions, not only in darkened rooms.
I think that with present technology, holographic imaging is not possible. Holograms, however, are a good basis for developing new kinds of dense data storage systems with true associate recall capabilities. Interesting work on this subject was done by groups at Caltech, Stanford, Colorado State, and UC San Diego in the 1990s. The February 1998 issue of the IEEE Computer magazine features a special section on this kind of technology.
There were some interesting papers at the 2004 Los Angeles SIGGRAPH on using planar arrays of cameras. Lots of people have tried stereo vision- because we have two eyes- but why stop at two? Cameras, projectors, and PCs have been inexpensive enough that you can experiment with redundant arrays of these, much like RAID revolutionized disk storage a decade ago.
Now what can you do with a planar array of cameras? You are seeing one viewpoint, or two, but *all* viewpoints, coarsely sampled. In some respects this is like a realtime hologram.
Marc Levoy's group at Stanford constructed an image "cube" of a scence- all depths of view and points of view. You can pluck out individual objects in a congested space like cocktail party or animals in a cornfield by computer synthesizing the appropriate focus. It almost seems like you can see through objects or arround corners.
Two other groups performed wide-angle realtime 3D TV (without eyeglasses). You have all the viewpoints all the time. Another group used an insect-eye approach using a special lense array and camera on each arrays. Then realtime computing would rearrange the pixels to present a 3D image.
Theres many other ideas to explore out there, if you liberate your thinking from the point of a view of a one or two eye creature.