I ran some tests on pouch leather sealers about one month ago.
It was primarily done to tell me which product sealed the best, and
if contamination of the wood was an issue in how the sealers were
applied to the leather.

I was having problems determining how to measure the flexibility of the
leather after being sealed.  You seem to have come up with a simple way
of doing a comparison.

In making up my leather samples, I sealed half of the leather with the
sealer and the other half was untouched.  I was using the untouched
part of the leather as a control sample.  This gave me an idea about
how leaky and how stiff the leather was before the sealer was applied.

I ran these tests using a 50 year RTV, 111 Silicone valve sealant
(grease), 100% acrylic polymer emulsion (Goldens GAC-100 Mediums),
and Ross Rubber Cement.  The GAC-100 sealer was the only one that
seemed to change the stiffness of the leather.  I was only using my
fingers for the comparison, but there was an ever-so-slight change
between the sealed and unsealed leather.  All of the other samples
didn't seem to change the stiffness at all if thinned properly.

I didn't try the egg white and water sealer.  I should go ahead and add
this to my samples.

I only ran my tests because of curiosity of using RTV as a sealer and
comparing the sealing properties of each product.

I'm not completely convinced that the slight changes in stiffness that
we are testing for contribute to how sensitive or how fast the valve
will work.  There are so many other factors that enter into the
equation.  Logic says that the stiffer the pouch, the less sensitive
the valve.  Is this difference so small that it doesn't even matter
when the valve is first rebuilt.  The big difference may be in how the
sealer ages.  How stiff does the pouch gets with age?  How porous does
the pouch get with age?

Has somebody run tests on a special valve that could have the pouch
leather changed without gluing in the samples?  This way the pouch
would be the only variable in the test fixture.  The pouch dish would
also be a second variable, but all of the other variables such as
bleed, valve gap, and value pick presser would be the same.  These
tests could then be done using a variable vacuum source so to see how
the different pouch materials affect the performance of the single
valve.

Pete Knobloch

 [ Pouch stiffness isn't important in an orchestrion, but it could
 [ be very critical in a reproducing piano.  The biggest concern
 [ expressed in MMD by old-timer techs is longevity.  Thus far we
 [ haven't learned how to test this in a short time.  ;)  -- Robbie