P3D Re: Realist mounting and the "stereo window"

[fj834@xxxxxxxxxxxxxxxxxxxxx (Dr. George A. Themelis)]

>Could someone define for me what the stereo window is, exactly?

"Exactly"?  No.  There is no mathematical definition.  So I will try
an inexact, emperical definition.  Every picture (even 2D) has a 
frame associated with it.  It ends somewhere.  These are the boundaries
of the picture.  Normally, you don't think about it.  Sure, it ends
somewhere... So, what is the big deal?  

Consider now a 3D picture.

A 3D picture viewed stereoscopically is placed in a 3-dimensional
space.  This picture also has a boundary.  It ends somewhere in space.
It is very natural to consider the boundaries of a 3D picture as a
"window" and the scene generally "existing" behind this window.
What creates this window is the stereoscopic fusion of the frame
around each member of the stereo pair.  This window can be black
(as in the case of slides viewed in a viewer or in projection) or
light (as in the case of prints mounted in white backing and viewed
with reflected light).

To make this "window" illusion (I hate this word!) look real you have
to consider how real windows work.  If you sit in front of a window you
will notice that your left eye sees less in the left side of the window
(as compared to the right eye).  The same must be true for the stereo
window.  When mounting slides you can control the placement of the
window in space by sliding them horizontally behind a mask.  The
stereo window must be placed in front of any object intersecting it.
That's true both for real and stereo windows.  You can have objects
coming through the window but they cannot go beyond the edges.

The window can also be controlled by trimming the images (used for
prints).  It does not have to rectangular or even parallel to the
sight of view (it will not be if the sizes of each R/L masks or
images are unequal.)

>As I'm mounting my Realist slides from Greece, all I shoot for is 
>vertical and horizontal alignment as close as I can get them...

What do you mean by horizontal alignment?  There is no such thing.
Alignment is only vertical.  Trying to set all equivalent points
in one straight line.  Horizontally you have placement to worry
about, not alignment.  This horizontal placement (with respect to 
a fixed mask) is the key to setting the stereo window.

>By stereo window, are people talking about a function of when you're 
>actually framing a shot (shooting the picture)? In which case it's fixed 
>from the time you release the shutter? Or is it a mounting thing?

Both.  Most stereo cameras have a built-in stereo window.  This is done
by shifting the centers of the film gates inwards with respect to the 
center of the lenses.  This stereo window is normally positioned at
around 7 feet from the camera.  If you view the film chips unmounted
(or if you make unmasked prints of the entire area) you will see that
this is the case.

However, most pictures are either placed in masks or trimmed.  Masking
on trimming will affect the stereo window.  When shooting a picture
you should be aware that the window will be placed at 7ft and be ready
to use different masks if you are trying extreme close ups.  (If you are
shooting with one camera or two cameras and parallel axes then the
window is at infinity and you can fix it by mounting/masking).

Enough with theory!  Here is an experiment almost everyone can do:

Take a stereo pair (two film chips).  Put them in an Albion or Realist
(or equivalent) aluminum mount.  Center them.  Take a look.  You should
see a scene evolving behind an imaginary "window" set at about 7 feet.
Move the chips closer to each other.  The "window" is moving back 
(away from the observer) and becoming larger in size (or the scene
is moving closer to the observer - the two perceptions are equivalent).
Move the chips away from each other.  The opposite is observed.  Note
that only about 1.5 mm of movement is enough to take the window from
7ft to infinity so use very small shifts with a pair of twizers.  If
you are using prints, experiment with Q-V mounts and slide the prints
under the mounts.

I suspect that you are using heat-seal mounts.  There is very little
room to move the chips horizontally.  Sometimes you might need to 
trim the edges (or corners) to get more freedom and place the window
where it should be.  But where should it be?  It is up to you...
There are different philosophies here, possibly a subject of another 
(equally long) posting!

-- George Themelis

PS.  I should add RBT in the list of mounts to experiment with.  They might
be the best for such task as they allow you to slide the chips perfectly
horizontally.  If you can freeview the pair as you move the chips then
you will have a moving stereo window shifting experience!!!


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