Re: [photo-3d] Spicer mounting problems

[Paul Talbot <list_post@xxxxxxxxx>]

Ralph Johnston/Linda Sherman wrote:
> 
> I mounted a few recent rolls of RBT 8-perf slides in Spicer wide 8perf
> mounts.
> They look great in a viewer, but when projected in a Brackett fader projector,
> had vertical alignment problems.  About 1/4 of the slides had an inch of
> vertical error on the screen.  (The image was about 20 x 30") [snip]
> Then I measured the slide  with a
> digital caliper and found that it was tapered!  One end was 40.0mm while the
> other end was 39.9mm.  Or one end was 40.0 and the other 40.1mm.

Ralph, the observed error appears to be due to something other than the
mounts.

First let me point out that manufacturing tolerances are going to be
some non-zero value for any mount.  If you've ever had some of those
RBT mounts that don't securely hold the pin bars in place, you know
that even the most expensive mount on the market is not made to zero
tolerance.  (I submitted an RBT-mounted slide to a competition and by
the time it arrived one chip was not even completely visible inside the
aperture window anymore, due to the pin bar shifting inside the mount.)
I previously mentioned that our new mounts were spec'd to .003" precision
(about .076 mm).  So 0.1mm manufacturing deviation is plausible for any
cardboard mount.  Deviations for plastic or aluminum could conceivably be
a bit less, but that comes at a significant cost!

Second, there seems to be a problem with the math of the above analysis.
Spicer extra-wide mounts are full height (23 mm).  If the image was
projected to a height of twenty inches, the magnification factor was
about 22x.  Mutliplying the magnification factor by the measured size
differential of 0.1mm yields a possible on-screen effect from the
manufacturing tolerance of only 2.2mm.  That is less than 9% of the
on-screen error that you observed!  There clearly must be some other
factor at work here to result in eleven times as much deviation as can
be attibutable to manufacturing variations in the mounts.

Perhaps you could investigate further and let us know what else is going
on to cause the observed error.

Paul Talbot