jcs@crash.cts.com (John Schultz) (02/22/91)
In a previous post, I mentioned a VR Turing test, where the user cannot distinguish between a VR simulation and reality. I believe we already have systems that can do this. Their domain is limited, but in their domain they work very well. These systems do not use head-coupled displays nor do they use data gloves. The displays are photo-realistic, and the frame rate is is 30Hz. They don't even use stereo displays, but rely solely on depth of field, light intensity, and other monocular visual cues. The display is so convincing to the brain, that if the user is standing and doesn't hold on to anything, that user will fall over just from visual cues from rotation. The user interface is an exact replica of the device being simulated -- you can't get any better than that. What are these systems? Dome-Fusion Flight Simulators. These systems should be used to calibrate/classify other VR systems. These systems are not really portable, nor are they inexpensive, but they do represent true VR. Current head-coupled systems don't come anywhere close to the DFFSs. This is due to low-resolution display technology limitations, lag-time, and bulky design. The head-coupled systems will catch up to the DFFSs when the display technology advances (CRTs/LCDs/Optics). Once the display is conquered, force/tactile feedback must be solved, and how about true motion (hydraulics)? It looks like the DFFSs will be the VR systems to rate against for the near future. John Tom Furness can probably elaborate further on DFFSs, and how close they come to simulating reality in their specific domain.
jmunkki@hila.hut.fi (Juri Munkki) (02/25/91)
In article <17101@milton.u.washington.edu> quasar@neuromancer.leis.bellcore.com (Laurence R. Brothers) writes: >There are no display devices that I am aware of for which the user >cannot distinguish between reality and the display,... Currently one of the easiest methods to distinguish between real images and computer generated (or otherwise displayed) is to check if you can focus on different depths. A stereo display will force your eyes to focus to a single depth while the stereo effect makes the brain think that distance is variable. One way to solve this problem is to monitor the focus of the eye and change the focus of the picture accordingly. Since this is far from easy, I would guess that most systems will ignore the whole thing and hope that a partial illusion is sufficient. IMHO, the human mind is flexible enough to compensate for the defects in virtual reality. All we need is something that will feel real enough, if the user wants to think that it is real. Just think of how well we adapt to the totally unrealistic worlds of video games and think how well a 14 year old kid controls the spaceship (or whatever) with a 9 position joystick. Our minds are the most powerful part of the virtual reality interface and will probably remain so for the next 10-30 years. So, a "virtual reality Turing test" is not necessary for virtual reality to be useful. Full realism has its uses, but the advantages over a slightly unrealistic are really quite minimal. ____________________________________________________________________________ / Juri Munkki / Helsinki University of Technology / Wind / Project / / jmunkki@hut.fi / Computing Center Macintosh Support / Surf / STORM / ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
tsarver@uunet.UU.NET (Tom Sarver) (02/27/91)
In article <1991Feb25.114611.29204@santra.uucp> jmunkki@hila.hut.fi (Juri Munkki) writes: IMHO, the human mind is flexible enough to compensate for the defects in virtual reality. All we need is something that will feel real enough, if the user wants to think that it is real. [Other stuff deleted] Our minds are the most powerful part of the virtual reality interface and will probably remain so for the next 10-30 years. I believe the essence of what he is saying that we need to use technology to allow users to suspend disbelief. We can use "tricks" which people will have to make an effort to evade (ie, discern the underlying technology). I don't believe any external set of equipment can fully deceive a mind looking for the technology artifacts. The only form of input which is fully invisible would be direct neural impulses. Even then, we would have to work at making the synthesis realistic. But the new technology would have "nothing under the covers," so to speak, unlike present technology. Any clue that one wasn't actually experiencing the material would be based purely on the content (eg, "The trees aren't moving, but I feel the wind"). Tom Sarver: tsarver@andersen.com | "Only Amiga makes it possible!" //\ "A real computer has a linear address space. NO 386's!!" \\ //--\
jcs@crash.cts.com (John Schultz) (02/28/91)
In <1991Feb25.114611.29204@santra.uucp> jmunkki@hila.hut.fi (Juri Munkki) writes: >In article <17101@milton.u.washington.edu> quasar@neuromancer.leis.bellcore. com (Laurence R. Brothers) writes: >>There are no display devices that I am aware of for which the user >>cannot distinguish between reality and the display,... I recently spoke to a fellow VR researcher (Mark) who had experienced a military helicopter simulator. A colleague if his was at the controls, and he didn't know what he was doing. After a few moments of flight, the helicopter was headed straight for the ground, out of control. Mark realized that this was a simulator, and wanted to turn around to make sure, but didn't want to look frightened. After all, this was just a simulator, and they were in no real danger. Seconds before impact, Mark could take it no longer and had to look behind to make sure that they were in fact in a simulator... Dome-fusion flight simulation displays don't need to be stereo to work effectively. In flight, most objects are far enough away so that everything appears at infinity (stereoscopically). Further, the dome is far enough away from the viewers' eyes that accommodation problems are minimized. See Computer Graphics: Principles and Practice plate I.5b, for an example of a real-time photo-textured terrain for flight simulation, or visit your local military/commercial flight simulation facility for a first hand look. >So, a "virtual reality Turing test" is not necessary for virtual reality >to be useful. Full realism has its uses, but the advantages over a slightly >unrealistic are really quite minimal. Of course. This whole concept lends itself to total freedom. Make it real, make it unreal. There is no right or wrong. If a specific goal is set, say "train F-16 pilots", then realism is an issue, and rigid realism tests are required. Entertainment applications have no bounds, as long as they are fun. Unless the goal is to torture the player (like viewing "Eraserhead"). There are no "rules" for music or film, but few will listen to "bad" music or "cheap" films. The same will apply to VR. John