Re: Where is the window - proposed experiment

[P3D Larry Berlin <lberlin@xxxxxxxxx>]

>Date: Fri, 20 Jun 1997 
>From: P3D Dr. George A. Themelis writes:
>.........
>
>As I've said, for me the stereo window is at 7 feet or wherever I decided
>to put it during mounting.  I cannot reach it with my hands!

****  Your measurement is entirely subjective. That makes it next to
impossible to describe accurately. I propose you use a picture with some
kind of meaningful measurement within it. The window is wherever you place
it, and you should be able to identify precisely where it is according to
the measurements within the image itself. Since you like using *feet* set up
a series of objects that are one foot in size and stack them up nicely
forming a series of steps into the image.

>
>I am really puzzled by the reactions that this discussion has generated. 
>..........  Now it is
>revealed that stereo photography does not hold much realism for most here.

*****  I didn't get that impression from these discussions at all!


>.......  I like them because it is "like being there".  It is the
>incredible realism of good stereoscopic photography that has me hooked for
>life.
>

*****  You look for realism according to your pattern of enjoyment. I don't
see realism as exclusionary of reality, the reality of having an image
instead of a real person. Realistic is, well, realistic. I think that it is
enjoyable. I don't have to hide real physical factors relating to images in
order to get a sense of the realism in a stereo view, even if I'm
freeviewing them.

Recently while using an RB viewer, I became aware that the lenses involved
create a bowl-like distortion in the reconstruction of reality. I was able
to see this by comparing spatial locations of points on a planar surface
within the middle distance of the image. It's easy to go beyond that
distortion and experience the image relative to it's contained references.
However, that bowl like distortion significantly alters the level of
absolute reality available within the RB viewer. Perhaps that is why you
have trouble noticing where the *window* is within it. The answer to your
question of *where the window is*, is that it is fairly close to the viewer
in that the reconstructed space stretches back towards the viewing device
and the person viewing. The closer you are looking towards the edges of
visible space the closer it's spatial location becomes to the viewer in a
non-linear fashion. 

The frames holding the slides are way off near the sides of that space
(especially for full frame mounts!). The mount is the defined location of
the stereo window in a hand held viewer. It's considered sort of *outside
the field of view* but beware it still exists and can affect the viewability
of your images!. This window location factor could be why there is so much
resistance to any stereo images that have elements significantly in front of
the window. For the RB viewer, that puts it almost literally *in your face*
and hard even with (or because of ...) optics to look at. The use of 5 perf
mounts causes most of the image to be in the central region of the optics
and subject to less of the distortion. The distortion is persent however. 

I know several won't believe this, so all I can say in advance is get your
demo slides together and check it out very carefully by obervation in both a
viewer and projected and freeviewed. I submit that freeviewing provides the
best evaluation of flat surfaces seen stereoscopically due to the stretched
perspective from which a flat surface could be observed. At such
perspectives it would be least subject to any distortions of any kind. You
need sharp awareness of triangulation convergence factors to see this
distortion in the RB. Using a full frame slide mount helps to explore the
distortion since it is more visible at the extreme sides than near the
center. The distortion is continuous, not flat in the center then suddenly
distorted, therefore it must exist to some degree in all views seen in an RB.

The RB viewer seems to have been designed to look through the window rather
than include the window. That significantly affects the total range of
stereo experience available in that or similar devices. It is significantly
different from LCS viewing on the computer screen, where the window isn't
*on your nose* but is a ways in front of you and more easily a part of the
viewing experience. I propose that LCS viewing is in this sense a more
complete stereo experience than a hand-held viewer.

I seek realism within constraints of each method. I experience no less
realism from freeviewing than from the RB viewer, though each is distinctly
different.

The presence of this distortion in the RB does not interfere with my
enjoyment of using it.

Larry Berlin

Email: lberlin@xxxxxxxxx
http://www.sonic.net/~lberlin/
http://3dzine.simplenet.com/


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