Covergence and keystone distortion

[P3D Dr. George A. Themelis <DrT-3d@xxxxxxxxxxxxxxxxxxxxxxxxxxx>]

>John R writes: 
>    
>> A "splitter" with a single input aperture introduces keystone distortion

Gary Nored replies:

>I've heard this myself, but used a beamsplitter for several years, 
>delighted with the results, and never noticed any "distortions." 

That's because you are using the viewer that came with the beamplitter
to view the slides, correct?  If you try to cut the slides and mount
them for viewing in a "parallel" viewing system like a Realist red
button viewer, then the distorition will be evident.  Greg Erkers
has some experience with this when he tried to remount his beamsplitter
wedding slides.

The "beamsplitter viewer" uses mirrors in a way similar to the one used
to record the pictures and the distorition is cancelled.

A similar situation is found in SLR pairs taken with considerable 
convergence of the cameras.  There is keystone distortion in the pair 
if viewed (or projected) with a "parallel" system, i.e. a system where 
the viewing/projection lenses are parallel. If these lenses are 
converged to the same angle as the recording lenses, then the distortion 
is cancelled.

Going one step further, I think the reason we don't see any keystone
distortion when we converge our eyes is that the retina ("film" that
records/views the image) is glued in the back of the eyes and rotating
with them.  So both the recording and viewing systems are converging.
But I am getting in Jim C.'s area...

-- George Themelis


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