P3D Re: Hummingbirds, telestereo close-up

[abram klooswyk <abram.klooswyk@xxxxxx>]

Dave Williams wrote (11 Jun 1999, P3D 3348):
>I'd like to photograph hummingbirds in flight in stereo.

George A. Themelis answered (11 Jun 1999, P3D 3349):
>I'll tell you how the experts do it... They use twin SLRs with
>longer lenses (say 135 or 200 mm lenses). (...) 
>I am aware of the work by Paul Milligan and Bernard Stell (...)

Rudy Wisnevski (Germany), Paul Wing (USA) and Allan Griffin 
(Australia) are also among the experts using this technique.
Rudy made a magnificent series of stereoslides of kingfishers.

Dave Williams again:
>How would I calculate the base for something like this? 

Tom Hubin answered (12 Jun 1999, P3D 3350):
>If you want the same size image and stereo perspective then 
>the object distance and stereo base are proportionate to the 
>lens focal length.
(...)
>the results are not identical.

Other things remaining unchanged (format, mounting, viewing)
*only* enlarging the base gives lilliputism, while *only* increasing
the focal length gives larger images, but with the depth dimension
squeezed (the Z-coordinate). 

Just for the theory, imagine stereographing a cube.
A large base makes a medium cube into a small cube, and
a long focal length makes a medium cube into a large "match box",
or better a box for ten 5 1/4'' floppy disks, as seen from the 
largest surface.

Combination of the two will reduce the disk box format from e.g.
5 1/4'' to 3 1/2'', giving (I assume) the right size of height and 
width (X and Y coordinates), but still squeezed in depth 
(Z-coordinate).

The solution is to enlarge the base *more* than proportionally,
and to move the chips outwards in the mounting . The latter will 
automatically be done when you mount "to the window".
The "moving-out" adds, in viewing, some stretch in depth, which to 
some degree compensates for the squeeze introduced by the longer 
focal length.
But as Tom said, the result isn't identical, theory says that
you will end up not with a cube, but with a frustum of a 4-sided
pyramid (assuming the original was a cube). But in practice this 
effect is *not* perceptible when viewing the slides.

(I have written on this subject before, stating that the "Pepax 
principle is incomplete", 22 August 1998, PHOTO-3D Digest 2917.
The theory can be proved mathematically, which has been done,
and Koo Ferwerda also has tried to illustrate the effects in his
book "The World of 3-D" - still available from 3d book productions
at www.stereoscopy.com/3d-books)

If this sounds complicated, just ignore theory and make a few
tests. I know that several of the experts also did this, with
splendid results and no mathematics. I only suggest to include
some longer base lengths in these tests.
Incidentally, by lack of a better name I call this technique 
"telestereo close-up".

Abram Klooswyk


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