[comp.graphics.visualization] Crystal Eyes

andyrose@batcomputer.tn.cornell.edu (Andy Rose) (11/15/90)

Regarding C. Upson's call for wish list, any software which can
drive the CrystalEyes setup or any other stereo pair setup would be great.

When visualizing "regular" shapes in 3D, the spatial interpretation is
helped greatly by the brain's ability to recognize straight lines and
angles.  As we begin to see more data which is amorphous, the lack of
straight lines makes stereo viewing essential.

For example, viewing a city street and discerning depth with one eye is
not too tough since there are plenty of other visual cues, especially
convergence of straight lines (perspective).  Look at fire or a cloud
with one eye though and there is little for the brain to "lock onto".

Since a common viz technique is the display of particles (ex. 1000 small
spheres animated), many of the tricks we have relied on in the past to
fool the brain into "seeing" (feeling?) 3D don't work.  Lighting models?
Phong, Gouroud (sp), Blinn, Hall, ray trace, radiosity, whatever don't
help because the surfaces are minimal and the balls may be moving 
quickly.  Relative sizes (perspective transform) hardly helps unless you 
are right in the action. 
Shadows? Adding 1000 shadows on a flat surface behind the action
will just confuse things.  Reflections? Forget it.  Visual occlusion (basic
z-sort) and flat shading is fast and as good as more computationally 
extreme methods.  

However, if you have stereo vision (call it binocular disparity, or what you
will) you "see" much better.  A swirl of dots will jump out at you.  Words,
as usual, don't do justice to the effect so see it yourself.

Because you can do much better with stereopairs than with all the fancy
rendering stuff, it is more COST-EFFECTIVE to use your compute power to
generate a left eye view and a right eye view in a simple model rather
than a monocular view with all the bells and whistles.

... I heard (SIGGRAPH) that powerstation SGIs have been shipped with
stereo ready monitors for some time now.  I think that means they 
can display 120HZ.

V for vis.
  


-- 
Andrew Newkirk Rose '91 Department of Visualization CNSF/Theory Center
632 E & T Building, Hoy Road Ithaca, NY 14583  
607 254 8686  andy@cornellf.tn.cornell.edu

andyrose@batcomputer.tn.cornell.edu (Andy Rose) (11/15/90)

Workstation News November 1990
p.9
PCs go 3D with goggles

Users of PCs and Macs can now wear a special set of stereoscopic goggles
to enhance 3-D images on the screens of their desktops.  StereoGraphics
Corp. is marketing a new, less pricey version of its CrystalEyes stereo
eyewear for use with low-end desktops and PCs.
  CrystalEyes/PC works with PCs equipped with "stereo-ready" graphic
boards from various vendors.  StereoGraphics will send you a list of
third-party products if you want one.
Price: $995
Contact: StereoGraphics Corp., 2171-H E. Francisco Blvd, San Rafael CA 94901
415-459-4500
---     
I imagine you will need a 60Hz monitor to get a good flicker-free effect.

This magazine, BTW, has an article on the facing page titled:
More Visualization from Tektronix,
The XD88/700 reaffirms Tektronix's commitment to integrated graphics, imaging,
and video.

So I don't know how good an info source this magazine is.  Subscriptions
are free though.

V for vis.

-- 
Andrew Newkirk Rose '91 Department of Visualization CNSF/Theory Center
632 E & T Building, Hoy Road Ithaca, NY 14583  
607 254 8686  andy@cornellf.tn.cornell.edu

bernhold@qtp.ufl.edu (David E. Bernholdt) (11/15/90)

In article <1990Nov14.201940.23433@batcomputer.tn.cornell.edu> andyrose@batcomputer.tn.cornell.edu (Andy Rose) writes:
>However, if you have stereo vision (call it binocular disparity, or what you
>will) you "see" much better.  A swirl of dots will jump out at you.  Words,
>as usual, don't do justice to the effect so see it yourself.
>
>Because you can do much better with stereopairs than with all the fancy
>rendering stuff, it is more COST-EFFECTIVE to use your compute power to
>generate a left eye view and a right eye view in a simple model rather
>than a monocular view with all the bells and whistles.
>
>... I heard (SIGGRAPH) that powerstation SGIs have been shipped with
>stereo ready monitors for some time now.  I think that means they 
>can display 120HZ.

You might be interested in the CAChe (Computer Aided Chemistry) system
from Tektronix.  It is a software/hardware package intended, as the
name says, at chemists, and makes use of a stereo monitor for
visualization.

The hardware is based on a MacII.  A special graphics display board is
used and one or more 88000-based co-processor baords may be used for
extra CPU power.  The display is a Tektronix 120Hz stereo monitor with
a liquid crystal shutter.  The user wears polarized glasses.  A
trackball system is used to give ther user the ability to manipulate
objects easily in 3D.

The software, combined with the aforementioned hardware allows you to
build and study molecules; computing quantities of interest, and
displaying them in 3D. The system works wonderfully, and I think is
unbeatable for _manipulation_ of complex 3D objects (like molecules).

Admittedly the software is aimed at chemists, but I think it is
excellently executed and if you have the opportunity, its worth taking
one out for a spin.  I _think_ they may now sell pieces separately, so
that you wouldn't have to get the chemistry software if you don't want
it.  Anybody interested in more information should contact me by
email.
-- 
David Bernholdt			bernhold@qtp.ufl.edu
Quantum Theory Project		bernhold@ufpine.bitnet
University of Florida
Gainesville, FL  32611		904/392 6365

shenkin@cunixf.cc.columbia.edu (Peter S. Shenkin) (11/16/90)

In article <1990Nov14.201940.23433@batcomputer.tn.cornell.edu> andyrose@batcomputer.tn.cornell.edu (Andy Rose) writes:
>
>... I heard (SIGGRAPH) that powerstation SGIs have been shipped with
>stereo ready monitors for some time now.  I think that means they 
>can display 120HZ.

The way SGI's stereo-ready monitors work is that in stereo mode, the "upper 
half" of the raster gets displayed not on the upper half of the screen but
rather interlaced with the lower half.  Thus, if the screen only had 8
raster lines the display would put them out as follows in stereo and
mono mode:

	mono	stereo
	0	0
	1	4
	2	1
	3	5
	4	2
	5	6
	6	3
	7	7

Thus, you lose half your vertical resolution in stereo mode, but the screen
refreshes at the same rate.  An IR LED emitter sends signals to a receiver
in the goggles to sync liquid crystal shutters in the goggle lenses with the
refresh rate.  

I have this on my personal iris, and I love it.  It does work with anti-
aliasing.

	-P.
************************f*u*cn*rd*ths*u*cn*gt*a*gd*jb**************************
Peter S. Shenkin, Department of Chemistry, Barnard College, New York, NY  10027
(212)854-1418  shenkin@cunixc.cc.columbia.edu(Internet)  shenkin@cunixc(Bitnet)
***"In scenic New York... where the third world is only a subway ride away."***

brucec@phoebus.labs.tek.com (Bruce Cohen;;50-662;LP=A;) (11/17/90)

There's another technique which might help some, which Jim Kajiya at
CalTech was working on about two years ago, when he described it to me.
The idea is to take a stereo display like the CrystalEyes (we happened to
be looking at the Tektronix display on a Tek 4330 workstation, since that's
who I work for), and add a gadget which tracks the position of the glasses
relative to the display.  We talked about  using a scanner, like what you see
in a supermarket checkout counter, and a reflective strip on the bridgpiece
of the glasses.  Then change the viewing parameters to follow the change in
position of the viewer's eyes.  This allow you to bob you head up and down
and right and left to get changes in parallax.  Unforunately, it only works
for one viewer per display.

I haven't heard if Jim has continued work on this idea.  If you're out
there Jim, can you tell us?

--
------------------------------------------------------------------------
Speaker-to-managers, aka
Bruce Cohen, Computer Research Lab        email: brucec@tekchips.labs.tek.com
Tektronix Laboratories, Tektronix, Inc.                phone: (503)627-5241
M/S 50-662, P.O. Box 500, Beaverton, OR  97077