Amica Ampico Article: Fifth and final installment of comments.

9. "Indeed, the Model B device is an amplifier; but it is not a true
   crescendo.  This single amplifier affects the suction supply to both
   'C' pouch regulators simultaneously." p 71, para 7.

ANS: In the first place, Jeffrey isn't aware of the fact that the regulator
pouch or curtain in the final Model B Ampico was never called a "C" pouch
(stands for control pouch) by the factory.  That terminology was reserved for
an earlier prototype built about 1925 or so.  See pg 239, The Ampico
Reproducing Piano, by Richard J. Howe.

   The purpose of the comment is actually a rebuttal to something I had
said in an earlier article anyway, and doesn't really apply to his
thesis.  He wants to believe that unless the "crescendo" function can
directly change piano loudness without the need for intensities, it
isn't a "true" crescendo.

   The determination of whether something is a true crescendo or not is
dependent solely on the way it was programmed by the roll, and it
doesn't matter if one piano required an angelic presence and the other
didn't.  All that matters is the roll itself.  And if both the Ampico 1A
and the Model B used a crescendo dependent on the intensities while the
Stoddard and Model A had a different setup but they were all used in
the same way and the roll made them do exactly the same thing, then
what could he possibly mean by saying that one piano had a true
crescendo and the other didn't?

   Don't take my word for it.  Take the word of Angelico Valerio and
Nelson Barden who discussed the issue of crescendo oriented rolls on p
128 of The Ampico Reproducing Piano.  Valerio: "atrocious." Barden:
"Terrible, Just awful." I think they are talking about some of the
rolls from the Hupfeld roll series here, because very few Ampico rolls
were "crescendo oriented." That is a provable fact that can be
literally measured.  But some of the recoded rolls by  Suskind in a
"make-do" desperate situation required, like 10" long slow crescendos,
and stuff like that.  And the rolls were almost humorous, they were so
melodramatic.

   The main reason Ampico didn't use the so-called "crescendos" without
intensities was because it made the roll too tempo dependent.  The
loudness of the piano escalated on an exponential scale while the
crescendo progressed linearly.  That isn't too friendly when your tempo
could be off, say, 15%.  As a matter of fact, approaching the correct
tempo number from the left side, as opposed to passing the number and
then returning to it from the right creates about a 10% change, in some
cases.

   The term, Crescendo was given to these analog ramping devices or
"volume controls," from its inception-- long before they were actually
used in that way.  It was a convenient label that described in a word
what they could do, not that they would ever expect to create realistic
musical crescendos with them.

   Even the 1919 Inspector's Instruction Book explains the use of
crescendos only as a "platform" to develop intensities on, and not as a
separate musical effect, although earlier rolls made use of them on
occasion more than they did after the 1921-22 Standardization.
Remember, though, before 1922, they were using a variety of methods for
coding Ampico rolls.  That means, some would play on a "B' better than
others.  Later, they recoded all of the top sellers before 1922 so that
they would be compatible with what was coming in their new pianos.
Since not too many Ampicos were sold before 1920 anyway, there aren't
too many early rolls to worry about.

   I have ben asked by Linda Bird, presently V.P. of Amica
International, to write a correction article to the Jeffrey Morgan
Ampico article.  It is a little different than these inserts, but
technically, about the same.  There is a bit more detail in the article
itself.

   I'm sure I do not have all the mistakes found, yet, but the basic
error underlying Jeffrey's "comparisons" is just that he was not aware
that the crescendo in both instruments is powered directly by the pump,
and that there is no feedback or regulated supply going to it from the
stack.  From this basic error comes all of his erroneous conclusions.  I
hope he hasn't been hooking up all his crescendos to the regulated
stack supply!

   In any kind of a regulator, a reference signal has to be measured
against an output signal before any kind of regulation can be
accomplished.  The reference signal (or desired intensity) must be
maintained with close tolerances until the output can be brought up (or
down) to equal it.  Unless the crescendo pressure can be maintained
independently of the intensities and the stack pressure changes, there
can be no regulation at all, and therefore, no reproducer.

   Without this basic understanding of any kind of regulator, whatever
else Jeffrey might say is moot.

Craig B.