Mounting to window vs. infinity

[P3D Dr. George A. Themelis <fj834@xxxxxxxxxxxxxxxxxxxxx>]

Mounting to window vs. infinity

Mounting to infinity:  Mount so that the infinity objects have a 
constant separation.  Ideally, this separation is equal to the lens' 
spacing in a viewer or 65 mm (average eye spacing) on the screen, so 
that the eyes are looking at infinity with parallel axes, as in real 
life.

Mounting to window:  Mount so that the closest object is at window level 
and let the infinity objects fall wherever they may.

Claimed advantages of mounting to infinity:  Consistency and better 
simulation of real life.  OK with the first, but the second is 
questionable, simply because viewing stereo slides is by nature 
different that real-life viewing.  In real life the accommodation and 
convergence change together depending on where we are looking.  In 
viewing stereo images accommodation is fixed and only convergence 
changes.  One argument for mounting to infinity is that by varying the 
infinity separation the eyes "get tired" while shifting convergence to 
view the infinity.  That cannot be true because the eyes change 
convergence all the time as they explore different depth levels in a 
scene.  Convergence carries little depth information and change in 
convergence takes place constantly and should not be a problem within 
limits.

My stongest objection to infinity mounting is that most of my slides do 
not have a real infinity.  I take lots of close-ups, hyperstereos, 
abstract slides, etc., not your typical 7ft. to infinity scenics.  Where 
is the infinity in a close-up of a flower?  There is no infinity.  
Mounting to the window is the only valid alternative for a pleasing 
image, IMO.

Dr. T's modified "mounting to window" approach:  Place the window 
wherever it is more pleasing only if the infinity separation is below a 
maximum value.  This is the "window control" approach.  Mounting to 
infinity gives you no control over the stereo window.  Most impact in 
stereo projection is carried by objects close to the window.  If they 
are pushed too far back, they get lost.  Many good stereo slides do not 
have anything close to the camera.  Bringing the window back will help 
improve those pictures.  Many times I want to bring an object through 
the window.  This is fine and very effective in 3d.  I do that by 
pushing the window forward.  WARNING:  When you do that make sure that 
you don't exceed the maximum separation.  This is when things can go 
wrong.  But if one consistenly shoots pairs with below maximum allowable 
deviation, and pushes the window within limits then there should not be 
any problem.

I would like to hear the other side... Am I doing something wrong and 
why?  I am going with whatever looks good.  I have seen many slides (in 
a viewer) that could be improved (IMO) by moving the window and have 
never heard anyone complaining about window placement in my slides.

-- George Themelis


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