Re: Stereogram on the "Stereo Window"

[P3D Gregory J. Wageman <gjw@xxxxxxxxxx>]

George teaches:

>The stereo window is controlled by changing the borders of the images. 
>This can be done by either trimming the images (in unmounted prints) or by
>shifting the images with respect to a fixed mask (in masked prints or
>slides).  In slides where a fixed mask is used (as in Realist slides)
>bringing the chips closer together pulls the window towards the observer. 
>Pulling the chips apart pushes the window further back.

I think you left out some key information, George.  Once I understood
the following, manipulating the window became very simple for me.

As you say, when the borders of the two images are aligned (zero
disparity), they appear to be at the same depth plane as the screen.
What you don't say is that what you are doing when you "adjust the
window" is that you are changing which objects in the scene *also*
have zero disparity, and therefore appear to be "at the window".
This works because your eyes have converged the proper amount to
fuse the window borders; you can shift the chips relative to the
borders without having your eyes change convergence along with
the chips, which would defeat the whole purpose.

You are not "moving the window"; to do that you would have to shift the
masks closer or farther apart.  You are in fact manipulating the plane
of zero disparity in the image, which can make the image appear closer
to, or farther behind, the fixed window.  I know it's a minor distinction,
but I really understood "mounting to the window" when I grasped this
concept.  You simply make the closest object have zero disparity to place
it at the window.

Why is "window violation" undesirable?

Most people recognize that there's something "funny" when shown a
stereograph with a window violation, and will tend not to spend much
time looking at the edges because of the sense of discomfort it causes,
even if they can't tell you exactly why or what bothers them about it.

It is due to the conflict between the depth cue provided by binocular
disparity, which "says" the object is in front of the window, and the
depth cue of the object being cut off by the window (occlusion), which
"says" it must be behind the window.  This tends to cause the stereo effect
to 'collapse' at the window edge, or the object to appear to oscillate
from behind to in front of the window, as the brain tries in vain to
reconcile the conflictory cues, in either case causing an unnatural
situation which is best avoided for comfortable viewing.

	-Greg W.


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