P3D Re: barrier grid review

[aifxtony@xxxxxxx (Tony Alderson)]

John R (digest 3093) wrote:
>(...) I believe lenticular and barrier strip-type displays are usually
>made with a particular interocular in mind. (...)<

Not exactly. Autostereos are designed to a particular viewing distance for
the average interocular. Persons with narrower spacing will find the "sweet
spot" a little closer, persons with wider spacing will find it a little
further back. (Think about the angles from a "top" view.)

I just got a chance to see two of Jason Shen's barrier strip images over at
Ray Zones'; the two tigers and the pool room. They strike me as pretty good
examples of the art.  The viewing distance seems to be about 6-7 feet. In
my opinion, this is a little far back for images of this size (about a foot
square), but I realize that can be adjusted. One thing I notice is that
barrier strip images are significantly sharper and contrastier than
lenticulars, this improves the impact.

It's clear there are parallax limits to autostereos, but as far as I can
tell, practitioners determine this by trial and error, or just hope for the
best. If Shen's pool room image had a little less parallax, the image
doubling at the far range would not occur, and the slight loss of depth
would not be noticed, probably improving the effect overall. It might be
helpful to have something like MAOFD for autostereos. I have never seen
such an analysis published, and haven't been sufficiently motivated to
apply my limited intellect to the problem. This is a multilevel problem:
first one has to determine the limits for a given barrier or lens, then
figure out out to get that into the image array.  But that is a topic for
Tech 3D.

Anyway, I would be interested in hearing how the tigers were photographed.
I seem to recall Jason saying he is using a rail camera, how the heck did
you prevent the animals from following the motion of the camera? Was this
done some other way?

Tony Alderson
aifxtony@xxxxxxx
www.aifx.com



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