Re: PHOTO-3D digest 1375

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

Neil Harrington writes:

>One minor correction:  The lens axes ARE parallel, I believe.  You're
>correct that they aren't centered on the film apertures, so that the viewing
>angles of the lenses are slightly toed in, so to speak.  But if the lens
>axes were not parallel there would be some keystoning.  Granted, the
>keystoning would be very small with small angles of toe-in, but no purpose
>would be served by doing this anyway.  As long as the frame apertures are
>spaced slightly farther apart than the lenses, the camera will have slightly
>converging angles of view (taking each view as a whole) while the lens axes
>remain parallel. 

Are you sure that's correct?  I thought the reason for the lenses not
being centered on the aperture was for window placement, by causing each
frame to include a slightly offset view of the scene.  I do not believe
there is any intention to converge the angles of view as you say.

If the apertures were centered, wouldn't the stereo "window" be at
infinity, with the entire scene appearing to be in front of it?  (That
would be a strange effect if true, wouldn't it?)  This is based on the
assumption that binocular disparity is greatest for nearer objects and
approaches zero for objects at "infinity".  If the frame borders were
coincident (as would be the case for centered lenses) there would be
no disparity between them and they'd appear at infinity, right?

	-Greg


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