cs161agc@sdcc10.ucsd.EDU (John Schultz) (01/30/89)
In article <3069@pixar.UUCP> good@pixar.uucp (Go ahead. Make my day.) writes: [stuff deleted] >...there's nothing they can do to get information to you at a >faster rate. (On the other hand, ShowScan, a film process, *can* do that. >And it's a lot better than 3-D.) > --Craig I saw ShowScan at the Reuben H. Fleet Space theater; it was incredible to say the least. Some women in front of us were crying because the view of the mountain ranges was so breathtaking. It really stimulates the brain. They stated that they couldn't do a film like Predator in ShowScan as it would be too powerful of an effect for people with health problems. At 60 frames of new information a second, very large reels are needed for ShowScan films. They stated that is would cost on the order of ~$100,000 to convert a theater to ShowScan. The only change to the screen is to make the whole front of the theater the screen. I wonder if they have tried ShowScan in 3D [120 frames/sec, 60 per eye]. To date, that would be the ultimate... John Schultz
jbn@glacier.STANFORD.EDU (John B. Nagle) (01/31/89)
Clearly the next step after HDTV will have to approach Showscan
resolution. We are also going to need computer displays with that kind
of bandwidth. How much bandwidth is it? Well, a 70mm film frame is
generally considered to be about 6000 by 8000 pixels. So 60fps x
256 levels x 3 colors x 6000 x 8000 = 2.2*10^12 bits/sec, or an
uncompressed data rate of about two terabaud. Hardware with this
kind of speed should be achievable by 1995 or so.
For flight simulators and such, one really needs an image that covers
a hemisphere. What fraction of a sphere is Showscan?
The amount of compute power necessary to generate high-quality
images in real time at these densities will be rather large. If we
take a goal of obtaining the quality of Pixar's "Tin Toy", and want a
hemispherical view, how much compute power will be required, and,
based on projections that progress continues at current rates, when do we
get it?
John Naglechris@spock (Chris Ott) (01/31/89)
cs161agc@sdcc10.ucsd.EDU (John Schultz) writes: > [Lots of stuff about how wonderful ShowScan is] > I wonder if they have tried ShowScan in 3D [120 frames/sec, 60 per > eye]. To date, that would be the ultimate... Actually, I've heard that someone (George Lucas maybe?) has experimented with a setup where showing 60 frames/sec allows the audience to see in 3D _without_ glasses. Supposedly, they just alternate between eyes, i.e. show a frame for the left eye for 1/60th of a second then a frame for the right eye for another 1/60th of a second, and so on, so the effective frame rate is actually 1/30th of a second. Still, 3D without glasses sounds pretty nice. --Chris ------------------------------------------------------------------------------- Chris Ott Computational Fluid Mechanics Lab Infatuation is blind, not love. A University of Arizona person in love can see the other's faults, but loves them anyway. Internet: chris@spock.ame.arizona.edu UUCP: {allegra,cmcl2,hao!noao}!arizona!amethyst!spock!chris -------------------------------------------------------------------------------
cs161agc@sdcc10.ucsd.EDU (John Schultz) (01/31/89)
In article <883@amethyst.ma.arizona.edu> chris@spock.ame.arizona.edu (Chris Ott) writes: > Actually, I've heard that someone (George Lucas maybe?) has >experimented with a setup where showing 60 frames/sec allows the audience >to see in 3D _without_ glasses. Supposedly, they just alternate between >eyes, i.e. show a frame for the left eye for 1/60th of a second then a >frame for the right eye for another 1/60th of a second, and so on, so the >effective frame rate is actually 1/30th of a second. Still, 3D without >glasses sounds pretty nice. How can this work? Both eyes will see both images. ShowScan itself has an almost 3D quality, but alternating left/right images at 60hz will produce a noticeable "double image" if the left/right eyes are not masked from the opposite right/left display frames. They must be doing something else other than alternating stereo pairs. I generate stereo 3D on the Amiga by the same technique as described above, but we use an LCD shutter the mask off the appropriate images... John Schultz
myers@hpfcdj.HP.COM (Bob Myers) (02/01/89)
> Clearly the next step after HDTV will have to approach Showscan >resolution. We are also going to need computer displays with that kind >of bandwidth. How much bandwidth is it? Well, a 70mm film frame is >generally considered to be about 6000 by 8000 pixels. So 60fps x >256 levels x 3 colors x 6000 x 8000 = 2.2*10^12 bits/sec, or an >uncompressed data rate of about two terabaud. Hardware with this >kind of speed should be achievable by 1995 or so. However, I strongly suspect that we will *not* be seeing the computer graphics industry moving toward this resolution; it simply will be too difficult to get the display device itself at a reasonable cost, at least in this timeframe. If you're talking color displays at 1k x 1k or higher, you are by necessity talking about CRT-based displays. An 8k x 6k display, even with a 0.20 mm dot pitch, would require a tube approaching 2 meters on the diagonal; finer dot pitches would make the tube smaller, but probably are not achievable anytime soon. (And it still wouldn't be THAT small - care to work out the numbers for 0.15 mm or even 0.10 mm dot pitches?) For that matter, how much resolution is usable? A 4k x 3k image, on a 19" tube, gives around 300 dpi, and is already comfortably beyond the color resolution limit of the eye. I suspect that something in this range will be the limit for "desk-size personal workstation" displays, especially in color. ShowScan is nice, but don't forget that it IS for the "big screen"! (One other frightening number to contemplate - the video clock for a 8k x 6k, 60 Hz non-interlaced display!) Bob Myers KC0EW HP Graphics Tech. Div.| Opinions expressed here are not Ft. Collins, Colorado | those of my employer or any other {the known universe}!hplabs!hpfcla!myers | sentient life-form on this planet.
rick@hanauma (Richard Ottolini) (02/01/89)
Nagle recently talks about high resolution ShowScan requiring terrabaud tranmission speeds and GIPs of computing. Perhaps not. At some point we reach the perceptual bandwidth of the human visual system and don't have to compute everything. Just compute more wisely. There are tradeoffs in spatial, color, and temporal resolution. Flight simulator people take advantage of these and compute very realistic motion pictures these days.
chris@spock (Chris Ott) (02/01/89)
> cs161agc@sdcc10.ucsd.EDU (John Schultz) writes: > > chris@spock.ame.arizona.edu (Chris Ott) writes: > > Actually, I've heard that someone (George Lucas maybe?) has > >experimented with a setup where showing 60 frames/sec allows the audience > >to see in 3D _without_ glasses. Supposedly, they just alternate between > >eyes, i.e. show a frame for the left eye for 1/60th of a second then a > >frame for the right eye for another 1/60th of a second, and so on, so the > >effective frame rate is actually 1/30th of a second. Still, 3D without > >glasses sounds pretty nice. > > How can this work? Both eyes will see both images. ShowScan > itself has an almost 3D quality, but alternating left/right images > at 60hz will produce a noticeable "double image" if the left/right > eyes are not masked from the opposite right/left display frames. > They must be doing something else other than alternating stereo > pairs. I generate stereo 3D on the Amiga by the same technique as > described above, but we use an LCD shutter the mask off the > appropriate images... It's some sort of mind trick. You know, like mixing red and green to get yellow. You think you're seeing yellow, but you're really seeing red and green mixed. Your eye can't tell the difference. What I'm talking about works because the brain can only process information at 1/60th of a second or something like that. Your visual processing center automatically figures which frame goes with which eye and you can see in 3D. However, I've heard that it doesn't work for everyone, only about 95% of the people who have seen it. > John Schultz --Chris ------------------------------------------------------------------------------- Chris Ott Computational Fluid Mechanics Lab Infatuation is blind, not love. A University of Arizona person in love can see the other's faults, but loves them anyway. Internet: chris@spock.ame.arizona.edu UUCP: {allegra,cmcl2,hao!noao}!arizona!amethyst!spock!chris -------------------------------------------------------------------------------
jv0l+@andrew.cmu.edu (Justin Chris Vallon) (02/01/89)
Sorry about the ignorance to follow. What is ShowScan? What principle is it based on (projection & perception)? What is the film medium? Thanks -Justin Arpa: Justin.Vallon@andrew.cmu.edu Bitnet: Justin.Vallon@andrew
cs161agc@sdcc10.ucsd.EDU (John Schultz) (02/01/89)
In article <EXtZ8Fy00VoBMI6VQ2@andrew.cmu.edu> jv0l+@andrew.cmu.edu (Justin Chris Vallon) writes: >What is ShowScan? What principle is it based on (projection & perception)? >What is the film medium? According to the ShowScan people, regular film is 24 frames/sec shown twice at 48 frames/sec (24 new images a sec). ShowScan is 60 frames/sec with 60 new images/sec. That's it. They found that the brain could perceive 60 frames a second optimally; less frames a sec was worse, and more than 60 produced no improvement in perception. With almost three times more information a second, images look much more realistic, almost 3D. The effect is like looking out of a large window at the front of the theater. John Schultz
jbn@glacier.STANFORD.EDU (John B. Nagle) (02/01/89)
Showscan is a straightforward movie system using 70mm film at about 60 frames per second. At this speed, the illusion of motion achieved at 24 frames per second is much improved. The screen is also made sufficiently large to cover the entire human field of view. The overall effect is said to approximate reality. It can be considered a benchmark as to how good a display system has to be before it disappears and becomes a virtual reality. John Nagle
sanders@tut.cis.ohio-state.edu (Steve Anderson) (02/01/89)
Having just returned from a tour of the Showscan studios in Culver City California, here is my 2 cents: The Showscan process, developed by special effects god Douglas Trumbull (2001:A Space Odyssey, Close Encounters, Blade Runner, Brainstrorm,...) features 60 different frames/second using 70MM film, compared with the ususal 24 different frames (each imaged twice) on 35MM film. The technicians explained that 60 fps was chosen because they felt it was a number which would provide them some degree of 'compatibility' with techniques that do or will exist. (HDTV for example.) They also said that they could do up to 72 fps, but the difference between 60 and 72 was not noticable. Another factor they said contributed to the 60 fps decision was it made the modifications to the camera/projector systems 'easier', with respect to gear drive changes. The Showscan process is actually a complete projection environment, with carefully planned seat placement, a big screen, and awesome sound systems. They even installed a system in France with seats that moved around on hydraulics, capable of some incredible g-forces. I think I remember them saying they had about 20-30 sites installed around the world, and are of course working on a few more. They had a recently announced 10-movie deal signed with a major studio to produce featrue films using Showscan. They are then planning on 'converting' some number of theaters in major markets which will show those films. They have a chicken- egg problem; no one wants to be the first to film using Showscan because there are no theaters to show it, and no theaters want to convert untill there are some movies. The conversion process wasn't really that difficult, they exchange some gears in the projector, and add a "little more hardware" I forget what they said it cost, but I remember thinking it wasn't all that much given the big business of the movies. (and the $5 tickets :-) ) Speaking of the movies, they don't have that many, but I saw two of them and they were pretty awesome. The 3D quality was pretty impressive. I won't address the idiotic issue of viewing movies through stupid plastic glasses. The crispness and clarity of the images were truly breathtaking. Besides the usual gratuitous roller coaster, mountain flying, motorcycle, etc. shots, things like close ups of people were amazing. Many of the scenes seemed so real you'd swear you were there. For anybody that wants to get more information from the real source here is there address: Showscan Film Corporation 3939 Landmark St. Culver City, CA 90232-2315 (213) 558-0150 (213) 559-7984 (FAX) -- H. Stephen Anderson | e-mail: The Ohio Supercomputer Graphics Project | steve@hobbes.cgrg.ohio-state.edu 1224 Kinnear Rd. | phone: Columbus OH 43212 | (614) 292-3416
charl@pnet01.cts.com (Charl Phillips) (02/02/89)
60 frames per second?!?? Mygawd! Aren't there any persistence problems? Charl Phillips San Diego, CA
jbn@glacier.STANFORD.EDU (John B. Nagle) (02/04/89)
In article <17670007@hpfcdj.HP.COM> myers@hpfcdj.HP.COM (Bob Myers) writes: >However, I strongly suspect that we will *not* be seeing the computer >graphics industry moving toward this resolution; it simply will be too >difficult to get the display device itself at a reasonable cost, at least >in this timeframe. Maybe not. There are multigun monochrome CRTs with resolution in that range, sometimes used for displaying X-rays for doctor interpretation. I believe that Tektronix licences that technology from the startup company that developed it. Color will be more difficult. Small projection systems can, of course, be constructed. At really high resolutions, though, beaded or ground-glass screens will have to be replaced by something with a finer structure, such as a matte white surface. Aligning the system will be a pain, but automated means have already been developed for compensating for optical misalignment in projection TV systems for military applications, so that problem is not overwhelming. It will be a long, hard road to reality-quality TV. But it will happen. John Nagle