From: deering@deering.Eng.Sun.COM (Michael F. Deering) Subject: Re: Viewing 3D images in a Head Mounted Display Date: 23 Jan 1992 21:47:22 GMT Message-ID: Organization: Sun Microsystems, Mt. View, Ca. In article <1992Jan23.171655.29643@watserv1.waterloo.edu> dstamp@watserv1. waterloo.edu (Dave Stampe-Psy+Eng) writes: >deering@deering.Eng.Sun.COM (Michael F. Deering) writes: > >> (My application of high accuracy superimposing physical and virtual worlds) > >An idea I had awhile back to fix this would be to use a pupil-tracker >on the display to compensate for mirror movement, as well as giving >feedback on what areas of the picture are important. This would also >allow you to track the real entrance pupil position of the eye, so you >could fix up secondary effects caused by eye motion in high-magnification >systems. This would take a fair amount of research and development, and >the exact compensation formula would depend on display optics. You need >to keep drift caused by head mounting down to a degree or less (about >1/2 inch at arm's length) so this is critical. In order to get the accuracy we were after, we decided to correct for the secondary effects caused by eye motion, even though our display doesn't use high magnification optics. This was only one of several secondary effects that become significant when going for high accuracy. All the gory details of our work will appear (hopefully) in a SIGGRAPH paper this year. On a related topic, in article 3106 Chris Shaw mentioned hyperstereo as an example where the eye separation need not be perfect. This is misleading. Yes, your agent in a virtual or physical world can have quite different interocular spacing, e.g. the scale of an ant or a giant. However, if you at the receiving end are to view the virtual world with head tracked stereo, then your local display needs to know your individual interocular spacing to make things look right as you move your head about. Most hyperstereo displays are not headtracked; my comments applied to headtracked systems. I'm glad to see a discussion of some of the fundumentals of the physical aspects of displaying virtual reality taking place. I hope that those who haven't yet experienced these effects will understand that they are important to the R portion of VR.