From: Michael B. Johnson Subject: TECH: Real holographic 3D displays at MIT, with citations Date: Sat, 17 Apr 93 19:52:01 -0400 Someone asked about 3D displays for VR. Someone else mentioned work that had gone here at the Lab concerning holograms. Alan Wexelblat responded that they were referring to work done by the Spatial Imaging Group under the direction of Professor Steve Benton. He then mentioned one of the current research goals of that group: the holographic equivalent of today's laser printer. Finally, he made a general aside that he felt holographic displays of any kind would suffer from a lack of immersion. In my original post I had the quote; now I just have the memory. If I misconstrued, I'm sure Alan can correct me. Either way, I wanted to respond in general to some misinformation flying around here about real 3D displays. Finally, someone else then posted the fact that they thought the idea of a real-time holographic display was blue sky. Okay, so that's what I'm responding to... Disclaimer: I'm not in the Spatial Imaging Group, but I'm very interested in what the folks in that group do, and bother them a lot, trying to understand what they do. Any misconceptions are my fault, not theirs, cause they know a lot more about optics than I do... Now: Alan failed to mention the other work that the Spatial Imaging Group has been working on for some time: a real-time holographic display. Several years ago they successfully demonstrated an image that was 64 lines by about 50,000 pixels, which ends up being about a 2" square. The working prototype was monochrome, red, and used a 16K CM2, a 5GB datavault, and a specially modified framebuffer (along with lots of optics and electronics) to drive it. New frames could be recomputed in a reasonable time, and movie loops could be played from disk. Since then, they have achieved full color (red, green, and blue lasers), upped the resolution to 128 lines, and are now working on a much larger scale system. Mark Lucente even hooked up a dial box and made a pretty cool 3D Etch-A-Sketch system. More exciting, from my point of view, Mark Lucente and Tinsley Galyean demonstrated how you can get an SGI to drive the system (as opposed to a CM2). This point is important for my response to Alan. Since Tinsley and Mark can use an SGI to drive Pierre/Mark/Steve/ etc's holographic display, we can display any of the scene databases that we normally display on our SGI monitors or HDTV monitors. And, since we can (with head mounted input) correct the display to to be viewpoint dependent (see Mike McKenna's paper in the SIGGRAPH 92 Symposium on Interactive 3D Graphics), there's no reason that you can't have a real 3D display that feels pretty darn immersive. In my experience, low latency and comfort (and sound!) have a lot more to do with immersive experience than having the display head-mounted. Holographic displays are, in my estimation, clearly where things will be, and not that far in the future (i.e. 10 or 20 years). The bandwidth requirements of such displays can be staggering, which makes them all the more interesting from my point of view. Anyway, I've asked some folks in SPI to give me some references, so y'all will please read these papers instead of asking me questions I probably don't know the answer to. Enjoy: --------------------- S. A. Benton, ``Experiments in holographic video imaging'', Proceedings of the SPIE, (ed. P. Greguss, 1991, Bellingham, WA), vol. IS#08, pages 247-267. Pierre St. Hilaire, Stephen A. Benton and Mark Lucente, ``Synthetic aperture holography: a novel approach to three-dimensional displays'', Journal of the Optical Society of America, vol. 1, #11, Nov 1992, pages 1969-1977. Mark Lucente, ``Interactive computation of holograms using a look-up table'', Journal of Electronic Imaging, vol. 2, #1, Jan 1993, pages 28-34. And for computer generated holography background stuff: W. J. Dallas, Topics in Applied Physics, The Computer in Optical Research, 6: "Computer-Generated Holograms", Vol. 41, ed. B. R. Frieden, {Springer-Verlag}, {New York}, 1980, pages = 291-366. --------------------------- Enjoy. --> Michael B. Johnson --> MIT Media Lab -- Computer Graphics & Animation Group --> (617) 253-0663 -- wave@media-lab.media.mit.edu