[MF3D.FORUM:490] Re: Exact fl match for paired lenses.

[Greg Erker <erker@xxxxxxxxxxxxxx>]

>I recently bought a second Mamiya 80mm lens. Now I can now shoot with 2
>bodies side by side with a dual cable release.  I realize that lenses must
>be identical fl on each camera for proper stereo viewing.  And the labeled
>fl can vary from batch to batch.  These were purchased 2 years apart.
>
>1.  What is the best way to determine if this lens matches my other 80mm
>lens?
>2.  How close does the match have to be?  +/- 1%
>3.  What is the effect in the viewer when lenses are not identical, but are
>within say 1%?  Is there some break point where things become noticeable,
>like greater than 2% for example...

  In a window-less stereo system you
could probably tolerate 3% or more of
size missmatch if you line up the centers
of the images. Many people with differing
eye-glass prescriptions for their two eyes
have larger image differences for their
two eyes.

  But when you take your two different
sized images and mount them in a slide
mount (with two same sized openings)
then the image size differences become much
more visible.

  If you align the bottoms and there
is no detail at the top (sky for example)
then the differences are hidden. So 1%
should be no problem.

  But if there is fine detail at the top
and bottom then the different cropping
on L versus R will be very visible.

  When I was building my stereo Ricohmatic
225 (SR225) I had 5 sets of lenses to pick
the best matching pair. The best ones matched
to 0.35% but I wasn't completely happy* with this
so I shimmed one lens out a few thou to reduce
the size difference to 0.13%. Then I was happy.

* the mismatch at the top was visible when
the bottoms were aligned (test roll shot of
my brick fireplace).

  Also FYI the RBT 35mm mounts have offset
adjustments of 0 to 0.4 mm in 0.1mm steps.
The image height is 20-some mm so the steps
are less than 0.5%. So RBT thinks aligning
images to this small amount is worthwhile.

  My SR225 page:
http://www.angelfire.com/ca/erker/twin225.html

Greg E.