From: Dave Stampe-Psy+Eng Subject: Re: TECH: Sci Am 5-92 and HMD's Date: Mon, 18 May 92 13:54:09 -0400 > From thinman@netcom.com Mon May 18 12:40:21 1992 > From: thinman@netcom.com (Lance Norskog) > To: rend386@sunee, sci-virtual-worlds@uunet.uu.net > Subject: TECH: Sci Am 5-92 and HMD's > > The Scientific American of May 1992 contains two back-to-back > articles which are important reading for any HMD designer. > > The first article is a "Science In Pictures (TM)" sort of > interactive science demonstration of interesting blind spot > effects. Example: a wheel has spokes between inner & outer > circles. If the inner circle is in your blind spot, you > "fill in" the spokes to the center of the missing circle. > > The second article details a new kind of mirror made via > semiconductor etching techniques. You custom-design a > field of diffraction gratings instead of shaping a smooth > curved shiny surface. Apparently it doesn't have the > restrictions that mirror optics has, and it can do a > lot of things that you can't with mirrors. > > The first is important for helping to figure new tricks > that you can get away with; or if only to remind you that > visual perception is very complex. The second is important > if you're designing HMD's with monitors hanging off the side > and reflected via mirrors in front of the eye. It's quite > possible that you could composite the images seen by each eye > from an invariant background from one monitor and per-eye > views from another monitor... > > Actually the "blind spot" thing isn't at all general. It develops over time in areas of the visual field with no input. For example, I developed a small blind spot in one eye a couple of years ago. For about 6 months, it was just visible when it overlapped the edge between 2 areas of different color, as a gray spot. This is "stablized retinal image", and works in about 3 to 10 seconds. The pattern fill-in took almost a full year to develop, but now can duplicate quite complex patterns, including grids. Point is, these effects only happen for retinally stablized images, so you'd need an eye-tracking display to use them for VR. -Dave Stampe