RE: Camera Shift Spacing

[P3D Larry Berlin <lberlin@xxxxxxxxx>]

>Date: Fri, 30 May 1997 
>From: P3D  David W. Kesner inquires:
>
>With all this talk of camera spacing (MAOFD, 1;30, weight-shift,
>etc.) when using a single camera for stereo, I got to thinking:
>wouldn't it be possible to mark your viewfinder (or prism or mirror)
>with two parallel lines once you have figured out the proper spacing?
>Then all you would have to do is make sure that the nearest object in
>the frame doesn't exceed that spacing. It wouldn't matter whether you
>were shooting insects (hypostereo) or the Grand Canyon (hyperstereo).
>Is this just too simple or am I just crazy?
>

****  The Nikon that I use has little rectangular corner markings near the
center of the viewfinder. I do use them in the capacity you mention, but the
amount of shift still varies with the distance of the total scene. It makes
a big difference what part of the scene you align the markings with. I
depend in most cases on a perceptible shift in perspective as I move the
camera. The markings are helpful in getting the framing right, especially if
you are hand holding the camera. By trial and error a sense of the right
amount of shift develops.

Today I saw some interesting clouds and used the markings to align with
identifiable features on the cloud (out the drivers side window) and drove
to the next position. I used the markings to align for the second shot and
prevent rotation of the camera. Since the clouds were distant, I allowed the
second view to shift slightly from where the first one had been positioned.
I was ignoring closer objects due to the hyperstereo positioning. I'll trim
a few treetops off before using the image. The shift was different than
later while taking a close up of a flower using a slide bar.

It pays to study your stereo pair images and learn to discern the
perspective shifts while looking at first one image (2D) and then the other.
Practice seeing the same shifts in the viewfinder.  Use the numbers as a
guide but teach your eyes to recognize the amount of shift that is
desirable. If you can see/sense the right amount of shift you don't worry
about the numbers more than occasionally.

One problem with your proposed idea is that the shift amount isn't the same
thing as the resulting on film deviation of the subject's features. You
would have to reposition your markings in order to determine actual deviation.

The most difficult part of hand holding the camera during the weight shift
method is preventing rotation!

Larry Berlin

Email: lberlin@xxxxxxxxx
http://www.sonic.net/~lberlin/
http://3dzine.simplenet.com/


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