Convergence and ortho seating

[P3D Ray Hannisian <ray3d@xxxxxxxxx>]

I enjoyed the recent discussion about finding the ortho-seat in a theater by
looking through the range finder of the camera which took the shots to be
projected.  

However, because I shoot 3D video, it is convenient for me to adjust the
'stereo widow' by manipulating convergence.  (Until I can mechanically or
electronically 'side-slip' the images into an artificail 'coincidence',
convergence solves more problems than it creates.)  

Moving the convergence point closer increases stretch, and causes distant
objects to diverge.  Therefore, it seems that there will be no true
ortho-seat for videos which use different convergence points from scene to
scene.

Perhaps, if I must choose a default convergence point for a projection
situation, I should move my cameras to the distance from the screen which
fills my video frame, then set the screen itself at my convergence point.
(?)  This, however, places me too close for a functional convergence point,
for shots which show distant objects.  (The situation gets worse for the
average TV screen.)

When all is said and done, there are so many types of distortions in
creating video, (which act as 'off-setting penalties'), that I have simply
learned the circumstances which yield results that 'feel' right to me.  I've
never had anyone react to the "3 inch tall King Kong effect" which was
discussed here several weeks back.  People 'accept' the images, and often
comment about "how real they look".

(Too bad Andrew Woods is off-line this week!  This seems like a subject with
which he may have some experience.)

Ray Hannisian
(RAY3D)


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