P3D Re: Single SLR Macro 3-D - Part II

[Larry Berlin <lberlin@xxxxxxxxx>]

>Date: Sat, 28 Mar 1998 
>From: "Andrea Blair" writes:
>................................................ If 
>you work from right to left, position the subject more to the left side 
>of the frame (not too close to edge of the slide if you will be 
>remounting into 5p or 7p Realist format mounts). Take first picture. Now 
>move the camera slightly to the left via the rail (no toe-in - just 
>straight across) and take another picture. I usually take 3-4 pictures 
>of every set-up, moving the camera slightly to the left each time. This 
>means the left image will have more of the subject (or background) on 
>the left and the right image will have more of the subject (or 
>background) on the right. ***I have been told this is opposite what 
>stereo pairs are supposed to be (which has resulted in some of my shots 
>being reversed by other people when published, and are thus unviewable), 
>but all of my shots are like this and are viewed perfectly. If someone 
>would like to explain this to me, please do. I don't know WHY it works, 
>I just know it WORKS!*** 

*****  It works because ... First of all, you are accurately describing your
work environment and movement directions. It makes sense that starting with
the Right image you will have most of your subject somewhat to the left of
center, and that this reverses when you move to the left image. What
confuses certain others is a misapplied rule of thumb. It has to do with the
fundamental geometry rules relating to the stereo window. You are working in
front of the stereo window so nothing reverses or crosses over. In scenics
or shots where your scene goes beyond the stereo window, the shifts are
reversed, and it is this infinity related shifting that the rule of thumb is
based on. It's for this reason that I seldom bother using this particular
rule. There are other cues that are less able to be confused. Freeviewing
will reveal the truth with no ambiguity!

You may need to re-adjust your stereo window in the results and that will
affect whether or not the rule of thumb works.

The persons publishing your work should by rights, use the facility of
freeviewing to double check the image placement before publishing. That the
images end up reversed is because failing to *look at the stereo aspect* is
like asking a blind person to set up the images by touch. Of course since so
many persons have not taken the opportunity to learn freeviewing, prominent
labeling of L and R can help avoid the problem. Ordinary viewing skills
should be able to decipher the labeling, if they will only believe what it
says. ;-)


Now for the extremely tricky technical part of 
>how far to move the camera each time. Based on the camera to subject 
>distance ratio (do I need to move to Tech 3-D at this point {:>) ?), the 
>answer is......about that much. What? How much? Did you miss it? I'm 
>sorry to disappoint a lot of you "techies," but I don't have a formula. 
>I just let the picture "happen." That's why I take more than two shots. 
>I can match the best two images which give me the best 3-D view. 

*****  In making multiple images it is more important to be consistent in
the amount you move each time than in precisely how far you moved. The
results will speak for themselves. I use a similar technique and the
dependence on a developed sense of the right distance is far more useful
than a calculator.

Since this is your technique, you may be missing an additional use for your
images. Lenticulars. You can arrange to have your images turned into very
nice lenticulars with a bit of planning. The typical lenticular with
lenticular cameras use three views, a L and R and a Center. Better
lenticulars can provide a smoother viewing experience by including several
more in between views that are spaced equally between the outside views. You
could plan for a 3, 5, 8, or more, and send your results to be processed
into lenticulars. I don't have all the lenticular companies info at hand so
won't try to supply that. 

It does involve some image editing and careful placement of the results on a
new strip of film. In one scenario, you arrange your images so that they
fall exactly on the film like happens in a lenticular camera, which is
essentially half frames. You will have to research your own alignments and
methods. This is one area in which digital imagery is very useful combined
with printing the finished results back to film. Carefully arranged three
piece images can be sent in for processing as if it were a standard strip of
film from a lenticular camera. For more than *three image set ups* it will
cost extra for special handling and processing. The results are fun though.

Larry Berlin

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


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