P3D Stereo base vs. focal length

["Dr. George A. Themelis" <DrT-3d@xxxxxxxxxxxxxxxxxxxxxxxxxxx>]

Mark Hatfield (fellow OSPS member) asks a few questions regarding
focal lengths:

>Isn't it true that there is a relationship between the focal length of 
>the lens and the diagonal of the film plane? 

A better way to put this is to say that there is a rule of thumb that
a lens is considered "normal" for a given film format if the focal
length is approximately equal to the diagonal of the film frame.
For full frame 35 mm (24x36) this is close to 45 mm.  For Realist-
format (24x23) it is close to 33 mm.

>When one uses a lens with a focal length greater than the film plane 
>hypotenuse the images foreground and afterground tend to look compressed, 
>or if one were to use a lens with a focal length less than
>the film plane hypotenuse the image tends to look elongated. 

That's correct, except I have to make a very important clarification:
It is not only the focal length of the camera that counts but also the
focal length of the viewer, or, to be more exact, the ratio of the 
viewing distance divided by the diagonal of the image.

The assumption is made that this viewing ratio is close to one.  That's
why the normal lens is defined as one having the same ratio.  This
is a situation that I can call "orthoviewing" (note: we are still talking
about 2d photography here!)  If you use a longer focal length lens and
still keep the viewing ratio of one, then you get this compression effect.

The situation can be expanded (and it is more dramatic!) when you add
depth as in stereo photography.

>So if one were to
>use say a pair of 500mm lenses on a pair of 35mm cameras mounted on a
>horizontal bar, might the apparent compression of the image tend to be
>relieved somewhat by increasing the stereo base? 

YES!  You have just expressed the PePax principle that we were discussing
earlier!  If you increase your stereo base in proportion to the increase
of the focal length then there will be a tendency for the two effects
(depth compression due to longer FL and depth expansion due to longer
stereo base) to cancel each other and result in something that looks
like normal stereo, i.e. fool the observer into thinking that he or she
came closer to the subject.

The only catch is that there are still ways to tell that this is not
the case, i.e. the observer has not really come closer to the subject.

I was planning to post some more discussion on this but have already
posted WAY TOO MUCH in P3d this weekend!

-- George Themelis


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