From: sweberg@garfield.wes.army.mil (Scott D. Weberg)
Subject: Re: HARDWARE: Full VR on your PC for $1500
Date: Thu, 1 Apr 1993 17:28:00 GMT
Organization: USACE Waterways Experiment Station


I saw the Subject and instantly was skeptical.  Autodesk tried to build a  
vr system for the pc, and it took 2 Matrox graphics boards at $5000 each  
to get "real-time" rendering.  A VPL Dataglove and Eyephones (used to)  
cost almost $10,000 each.  And now we can have "full VR" on our pc for  
just $1500!  A little too good to be true, I think.

I must admit, after reading the article it does seem like you will get  
quite a bit for your $1500, but I would hardly call it "full vr".  That's  
the kind of advertising that makes my job really tough - my boss reads  
this and says, "Why can't we do that with our system?  I want it to work  
like that - by the end of the month!"

> Until recently, Virtual Reality systems were beyond the reach of the
> home PC user. Happily, this situation has changed, owing to a range of
> PC add-on components released in America this week. 

This statement is misleading.  The kind of vr systems that are currently  
"beyond the reach" of pc users ARE NOT going to be replaced by this  
system.  The product that is advertised here as a *general* vr system  
should actually be called a "vr gaming system".  There is no mention of  
software that allows you to build your own vr worlds, or import existing  
CAD models into the vr system.  It comes with 3 *games* - that's it.
Now if it really does have full blown model-building and real-time model  
interaction and rendering software, then that's something.  But from the  
article this doesn't seem to be the case.

> The VPS has an inbuilt graphics and maths co-processor, enabling the
> system to produce smooth real-time surface renderings and animations
> with a barely noticeable 500 millisecond timelag. Most impressive. 

Mmmm.  500 milliseconds is a fancy way of saying "half a second".  Now I  
know from experience that a half-second time lag is very noticeable.  It  
means that you have to turn your head slowly, so that the computer image  
doesn't get too far behind.  Some people using our systems have actually  
experienced a kind of motion sickness caused by the time lag - and our  
time lag was less than half a second (we don't run on pc's :-).

There is also no mention made of frame rate vs. number of polygons.  What  
kind of frame rate does it get for a 5000 polygon world?  Now, if all it  
does is games, then you can hardcode some animation stuff and get a lot  
more speed, but that is no longer general purpose vr.

> An advanced version of the glove controller interface is designed for
> use with Psynapse's own dataglove tracking system. Known as the
> Fibre-Optic Orientation Link, this device not only allows manipulation
> of VR objects but also provides tactile feedback to the user.

I bet this costs a bit extra - so much for the $1500 threshold.

My purpose here is not just to pick this apart, but to note one example of  
"misleading" advertising about vr.  This occurs all the time in this new  
field.  I saw a nintendo game (Faceball 2000) recently advertised as  
"Virtual Reality for your Super NES!"  It was just a 1st person, 3-D,  
pac-man shoot-the-bad-guys-in-a-maze type game - certainly NOT virtual  
reality!

Actually this product sounds like a pretty fun vr game; if you have $1500  
to throw away, have a blast!  Just don't think that you are getting the  
same kind of vr as those "out of reach" systems.  And I might add, I have  
not been overwhelmingly impressed by even the "out of reach" systems that  
I have worked with.  "Full blown VR" is not nearly as impressive (IMHO) as  
it is often made out to be.  Now 10 or 15 years from now things will begin  
to get really fun, but today it can be frustrating.

Scott D. Weberg