P3D Re: Stereo Boot Camp Drills (was formats & focal lengths)

[Bruce Springsteen <bsspringsteen@xxxxxxxxx>]

Thumbs up to Gabriel for his do-it-yourself exercise in seeing the effects
of perspective (viewing position) change vs magnification (focal length)
change.  Now, if you set up a small tabletop scene with assorted objects
in the fore, middle, and backgrounds, you can do a more photographic
experiment!

Using a short FL lens on your mono camera, shoot the objects from six feet
away.  Take a second shot with the same lens at four feet away.  Take a
third shot at two feet away.

Now put the camera back at six feet and take three shots, all from that
distance, first with the same short FL lens, then with a medium FL lens,
then with a long lens.

When you put the first set of three shots in a row, you will see what
change in perspective does.  The differences in size between the
foreground objects and the background objects will become more exaggerated
with each move towards the table, and the proportions of objects
themselves may show distortion as well, like the big noses and tiny ears
seen on close-up portraits taken with very wide angle lenses.

When you put the second set of three shots in a row (below the first set
is a nice arrangement) you will see what focal length changes do.  Now
there is no change in the *comparative* sizes of objects in the scene, and
no perceived distortion, just an enlargement of everything, like "blowing
up" a picture in a copy machine.

If you study the two sets laid out together as suggested, the difference
will fuse itself into your brain and you will never forget it.  I first
saw this demonstration in a photography book, and studying it was the
clearest lesson for a beginner that I've seen.  At least it worked for me!
 
In stereo, I insisted on the pure principle that changes in perspective
(shooting/viewing position) and parallax (the effect of stereo base
changes) are quite separate and unrelated things.  Non-ortho perspective
is about squash and stretch.  Non-ortho base is about changes in scale. 
But Doctor T explains that one can still be used to "sort of" compensate
for the other - a technique the stereo author Herbert McKay called
"PerPax".

> Agreed.  But let's be practical.  If you shoot stereo with longer lenses
> because you want to increase your image size (as in wildlife
> photography) should you also increase the stereo base, or not?
> 
> Only by increasing the stereo base you get anything that does not have
> obvious "compression" and cardboarding built into it.  The two actions
> do not "cancel" each other but even Ferwerda states that the
> deformations will not be noticeable if you isolate the subject and it
> does not have much depth to it (which is how this technique is normally
> used).

Agreed!  I'm very eager to dispel the misconception that increasing base
puts "stretch" in a picture, so that's what I zeroed in on.  But treating
base and focal length "as if" they reciprocate each other is a useful
trick (I think that's a fair word for it. ;-) ) in the practical
circumstance Dr T describes.  In fact he has a good example of it in his
*new* Realist camera book - a picture of a bear taken, of necessity, at a
distance, but with enlarged base to stereoscopically bring the bear
"smaller and closer", the real effect of base change.  But the average
person seeing the picture doesn't assume a small bear, if the base
increase is not too great.  It assumes a normal bear taken from a closer
position, and so a "happy compromise" is achieved.  I only ask that
everyone remember that the increased depth perceived in the bear is not
due to "stretch" or distortion - that's a different animal altogether!

Now how might we do a self-demonstration, similar to the one I described
above, to firmly and finally illustrate the stereo principles of
perspective and parallax, and the differences?  Ideas?

Bruce
 



__________________________________________________
Do You Yahoo!?
Bid and sell for free at http://auctions.yahoo.com