Suitability of telescope oculars for stereoscopes

[T3D john bercovitz <bercov@xxxxxxxxxxx>]

This is all Sam's fault.  Entirely.  So blame him.  Sam asked over 
on P3D about curved field lenses.  That reminded me of some posts 
from Dave Martindale of IMAX on P3D a while back.  Dave Martindale 
said that good telescope eyepieces have to have pincushion in them 
to keep objects at the edges of the field from getting squashed in 
a direction radial to the axis of the eyepiece.  According to him, 
this had to do with the telescope looking at things which are so 
far away they essentially form a dome around the observer.  (See 
how it Sam's fault?)  I was highly doubtful at the time Dave wrote 
this but I've been fooling around with perspective and what it 
means to 3D for a few months now, so I thought I'd revisit the 
problem.  

I find that Dave is right and that telescopes (or binoculars for 
that matter) should have pincushion distortion in their eyepieces 
or barrel distortion in their objectives or both.  But not for 
precisely the same reasons as Dave said, IMHO  I find it's the 
effect of magnification, which is equivalent to viewing a scene 
from too close, which distorts the perspective.  If the telescope 
were 1X (I understand that there's little call for 1X astro 
'scopes) there wouldn't be a problem.

Since cameras are more familiar to the members of this group, I've 
drawn some sketches using cameras rather than telescopes to show 
the effect.  The sketches are called cam1x.GIF and cam2x.GIF and 
should be found presently on this group's web site, so kindly 
maintained by Joel, at:
http://www.frii.com/~rkymtmem/tech3d/tech3d.html

The thing to notice is that whether the object field is curved or 
flat, the angles are reproduced correctly if the photo is viewed 
from the center of perspective.  However, if the photo is viewed 
from a distance which is a factor of two too close, the center of 
the photo is enlarged 2x alright but the edges are squished in the 
radial direction because of the extra slanty view you have from 
that location.  (How's that for highly technical jargon?)

To my way of thinking, this is just one more nail in the coffin of 
viewing from the wrong perspective.

John B

=====================================================================