From: uselton@nas.nasa.gov (Samuel P. Uselton)
Subject: Re: half silvered lenses (was Re: Direct Neural Input (Was Re: 
Date: 31 Oct 91 05:45:13 GMT
Organization: University of Waterloo



dstamp@watserv1.uwaterloo.ca (Dave Stampe) writes:

>uselton@nas.nasa.gov (Samuel P. Uselton) writes:
>
>>        The first Chapel Hill system also used temple bar mounted CRT's
>>        and 1/2 silvered lenses, superimposing wireframe images on the
>>        real world.
>>
>>        The images look "ghostly" because you do see the world "through"
>>        things rendered.  I don't know how satisfactory that would be,
>>        or how to avoid it.  (The other option, of course, is to digitize
>>        the "real world" and include it in the virtual environment.)
>
>I thought about that kind of system a few years ago, but I had some doubts
>about it.  The problems I felt it had were:
>
>- Any motion of the headset translates into BIG shifts in picture
>location because of the mirror's magnification

Who said anything about magnification?  If you are doing head tracked
rendering, the change in what you see should be determined by the motion.
Whether it is wireframe or raster shouldn't change apparent position.
The CRT's and half-silvered lenses are all part of the same head mounted
rig, so their *relative* positions don't change.  Therefore, whatever
distortion is required is constant.

>- delay between the video and the real world meant that if, say, you
>were moving your hands to manipulate virtual objects, the objects would
>lag behind your hand movements: VERY disconcerting!

This is a problem with (1) the tracking technology and (2) the compute
power to re-render your scene.  This method is independent of (1) and 
(2) should make (2) easier than raster systems.  

>If these objections can be overcome, this system of half-silvered glasses
>has BIG advanteges for the user, as eye-hand coordination is unaffected
>by system delays, and real as well as virtual objects can be worked with.
>Needs a lot less computer power, too.

Of course it depends on the application.  They even tried filled polygon 
rendering and half silvered lenses.  Then the objects *really* look ghostly.

If you want a substitute reality, this doesn't work too well.  If you 
want to add some ghosts to this reality, then it will work.

Opinions, of course.

Sam Uselton 		uselton@nas.nasa.gov
employed by CSC		working for NASA (Ames) 	speaking for myself