[photo-3d] wobbliing 3d...

["Sergio Baldissara" <winter@xxxxxxxxxx>]

Ciao dall'Ialia!
I heard of a faily recent (1970-80's) and already obsolete 3d standard for
TV: VISIDEP!
I browsed the web but I just found a few tips, like this one:
http://www.islandnet.com/~ianc/dm/23/233.html
In other web pages VISIDEP was mentioned just as a reference, but some
medical AA toke it seriously for a while. (Did anybody in the list ever see
a working sample? What was it like?)
The base idea was physiologically correct (rapid movements of the head can
make depth  perceived in one-eyed vision) but I was doubtful: I couldn't ask
some questions I had in my mind:
1. why switching images from over-under points of view? side-by-side should
anyway be better!
2. if this was true, why field-sequential videos are unwatchable without
spex?

I couldn't find an answer, and visidep shouldn't have proven so exciting,
since any memory of it is nearly lost. I tried a couple experiments by
modifying the original project:

i: I made an AVI to show the 4 frames of a Nimslo in this sequency: 123432.
When looping, I continuosly alternated 4 points of view. The results were
poor: I could just see far object "wobble" less than near ones, but the
complex resulted very ugly.

ii: I toke a shot holding my Nimslo vertical; I kept "scenics" (infinity)
add-ons on lenses 1 & 4 .  My best AVI used just  3 frames: 3242... so the
central image swhowed twice the others... I got the best results at 80fps
(the maximal frequency allowed by my PC hardware)...  this result was less
disturbing than the former, there was some illusion of depth, but not yet
worth the job!.
I'm not set up to process videos, so I just used scans of steady pictures.
Perhaps moving object or scenes (motion pics) may behave differently.
Sergio