Re: NuVision 3-D Spex

[P3D Jonathan R. Gross <catalyst@xxxxxxxxxxxx>]

Those of you that know what I'm up to are probably not surprised that I
am responding to Harold R. Baize's and Larry Berlins informal
evaluation of NuVision's 3D Spex.  I do not own a pair, but I might buy
some just so that I can claim to have actually compared them to
VR-Vision.  In the mean time, without even using them I can tell that,
IMO, there are some misunderstandings intermixed with the objective
observations in these two postings.

> I'll confirm that these glasses are pretty good.
> 
The glasses may be very good.  The cost of quality LCS glasses continues
to drop.  The misconception is that the quality of the image is
dependent primarily on the LCS glasses.  Although poor quality glasses
will reduce the quality of the stereopsis and comfort of the viewer,
good quality glasses do not ensure good quality stereoscopic images. 
The software, the refresh technique, the LCS drivers, and the monitor
are all as important, or taken together, more important than the quality
of the glasses.  The reduction in the price of the glasses simply means
that the cost PER VIEWER will go down quickly. Good news.

> I've been working with a small DOS based viewer so putting my own stereo images in the viewer is possible and operational. Better viewers are on the way. The big advantage this board offers for stereo photography is that it provides page flipping at 120Hz which appears really steady and presents the full resolution of the images within the limits of this simple program (640 x 480 at 32 bit color). 

Refreshing the fields at 120HZ is necessary to obtain a truly flicker
free image,  but the resolution mentioned here is inadequate for high
quality images.  The horizontal resolution of 640 pixels is particularly
marginal because human vision is very sensitive to horizontal
displacement (if it were not so, then retinal disparity would not be so
effective at inducing stereopsis through parallax).  This sensativity
one reason why electronic monitors scan back and forth, not up and
down.  IMO, 1280 pixels is the minimum horizontal resolution for quality
stereo.

Color resolution, to some extent, can replace spatial resolution.  In
other words, if each pixel can have a greater range of colors, then
fewer pixels are needed to convey the image at an acceptable quality. 
If this were not so, the then TV would never look as good as it does. 
Here again, the specifications above fall short of what would satisfy
most serious stereographers.  I suspect that the 32-bit color (4 Billion
colors) referred to above is actually 15-bit color (32 Thousand
colors).  A simple mistake, this makes a huge difference.  The
requirements for spatial and color resolution are why VR-Vision supports
1024 horizontal pixels with up to 2Million different colors, and refresh
rates up to about 160HZ (depending on your monitors capability). 
Remember, doubling the spatial resolution provides FOUR times as many
pixels, and  every BIT added to the pixel depth DOUBLES the number of
colors.

> The 3D games give a spectacular experience combining full steroscopic images with lots of fast action. 3D slides are a totally separate category by comparison to both a stereo view and motion with interactivity. These will definitely promote the use of and interest in stereoscopic 3D. Worth having the experience if you're into 3D, even if you aren't particularly a *gamer* by nature. I used the games in the standard mode which uses a 60Hz rate and they are fun but with a noticable flicker.
> 
ALL of the inexpensive systems are designed for use with games, and not
for those interested in high quality stereoscopic images.  Its fine
when you want to play games, but dont expect to be satisfied with the
quality of the images.  There is a reason that VR-Vision sells for $295
instead of $59.95 or even $159.95 .

In anticipation of those of you who might feel that this posting is
commercial and self-serving, I partially agree.  You tell me how I
should respond.  Electronic stereoscopic displays have their own sets of
issues and terminology, just like photographic systems do.  In the same
way that its of interest to know why a Realist with f2.8 German Tessar
lenses sells for more than a Realist with f3.5 D.C. White triplets, I am
trying to explain some differences in the products mentioned above.  
These differences are directly related to fundamental issues of
(electronic)stereoscopy, and are not either product or vendor related
except to the extent that some products incorporate certain
technologies, and others do not.  Without understanding the issues and
technologies, it is impossible to evaluate some of the (mis)statements
made in previous postings.

Jon Gross
http://www.skypoint.com/~catalyst/stereoscopic.html


------------------------------