John Phillips' recent posting about punch sizes should open up an
interesting discussion.  In an email exchange a little while ago
John brought up his findings to me, so I got out the original roll
in question and measured the single punch dimensions.  It was clear
that the original punch was not round, but a round punch 0.081 inches
in diameter with flats ground on two sides to 0.071 inches apart.
(Sorry John, but for the U.S. audience this discussion will not be
in metric dimensions!)

Due to the tendency of the paper to push away from the punch in
proportion to the number of layers, the usual production punch run
of 8+ paper sheets can't be "nibbled" to produce a hole like that.
With perhaps up to 3 layers, the hole could be "nibbled" to 0.081
inches long, but not in the usual production run.  To say nothing of
the problem of advancing the paper in increments less than the usual
row advance for a given perforator.

As a practical matter, today's perforators are limited to round
punches and fixed row advances.  This means that single perforations
are the diameter of the punches.  The Tonnesen's perforator, which is
the one used to make our Rock Soup rolls, uses punches 0.071 inches in
diameter.  So, in the case of the Schumann sonata set John bought from
us, even though the starting point of the notes is accurate in relation
to the original, the duration is not.  Or perhaps more importantly, the
area of the hole is less than the original.

Referring to the 1911 roll standards, the minimum single perforation
diameter is 0.0595 inches, but the perforation must be at least
0.004475 square inches in area.  This works out to be a perforation
0.083 inches long, at the minimum diameter of 0.0595 inches.  The
equivalent diameter for this area is about 0.0755 inches.  The original
Schumann roll single punch is 0.00503 square inches, but in the copy it
is only about 0.004 square inches.  This must be contributing to John's
single note response problem, though it is only one factor.

The Schumann sonata set was intended for the late "C" Recordo system,
which was only found in grand pianos (at least as far as I know).
So the rolls were designed for the faster response/repetition of the
grand.  The "B" Recordo test roll expects a repetition of 7 notes/
second at the zero intensity, and was intended for use on upright
pianos.  Properly regulated grands should be able to do at least 10
notes/second at zero intensity, and in my experience can do better.
This is contributing factor number two.

The third factor is that the roll is probably pushing the limits
of what most pianos can handle well, even in the best of condition.
Our upright Cable Recordo can't play this roll well, though the
Gulbransen baby grand Recordo we once had could have done so.

One could improve the situation by increasing the punch diameter to
0.076 inches, or perhaps more, but then horizontal positioning becomes
more critical, both in punching and playing, and bridging is more
difficult to match.

This discussion is only about Recordo, but it applies to the other
makes as well, with all their variants.  I'd like to hear about the
experiences others have had in this area.  John's comments about the
points at which the notes/valves operate certainly highlights the
importance of having valves and bleeds closely matched in performance.

Bob Billings