Dave Saul brought up some good questions about A and B "compatibility"
that still need to be answered. Granted, there will be no way to
satisfactorily answer someone who is basically saying, "Well, to my
ears, the B plays the A rolls poorly." I'll leave that alone. But there
were a few salient, legitimate technical points that can still be
addressed, so I will quote them from Dave's letter, yesterday.
> Some Ampico "A" rolls yield artistically passable performances on a
> "B" piano. Other "A" rolls, however, do not satisfy even moderately
> critical ears. That much is a fact, even without going into the
> technical reasons, which are manifold and have been well documented
> in the AMICA bulletin and elsewhere.
"Artistically passable" cannot be discussed objectively and is not a
"fact," (factoid, perhaps) since we do not know what his piano sounded
like, what rolls he was playing, or what problems the piano(s) may have
had. Still, I can comment on the last part in which he says the
technical reasons for model B's being poor reproducers are "Manifold"
and "Well documented in the AMICA Bulletin and elsewhere." I will
challenge that statement. The three exceptions to its "compatibility"
with the A piano have already been given. Aside from those three
exceptions, there are no others of which I am aware. So, "manifold"
and "well-documented" is, in my opinion, possibly exaggerated and
I do, however, remember Dave Saul writing in an AMICA article entitled
"A New Valve For Reproducing Pianos," that he had a terrible time
rebuilding his own Ampico valves-- to the point that he finally threw
them away, if I recall, and designed his own. (The first paragraph was
all about how critical the valves really are to a reproducer). So I
realized that the performance of the Ampico B had been a disappointment
to Dave until he built and installed a totally different valve of his
own design. He related the story of throwing away all the old original
Ampico valves, after not being able to get them to perform as intended
by the test roll. The purpose of the article was how one rebuilder
solved the problem created by badly designed valves which were not able
to be rebuilt correctly.
However, I think perhaps there was still more to his problem than the
stack valves, after hearing his present objections. In his book,
Rebuilding the Ampico B, Dave impugns that it is optional to replace
the expression curtains, since they are sealed and seldom show
deterioration. It really isn't a judgment call. I ran some simple tests
with my first Ampico B, and discovered these old curtains were
considerably stiffer than new material. I was really delighted at the
musical difference after replacing them. More differences in
performance ensued as I later on replaced the (original and seemingly
perfect) expression pouches, and as the work progressed, realized that
the secret of doing Ampicos right-- all the way through-- was the only
So if you notice that the Ampico valve blocks in your stack (for
example) are still coated with that thick, original dark brown shellac,
that piano has likely not had its pouches restored-- which means, that
piano has only been repaired. Those same valves come alive when their
original pouches are replaced and properly sealed, and they are given
new ball bleeds.
Unless the valve blocks are sanded clean--totally removing the old
shellac sealer first, it is usually impossible to find the glue line of
all the pouch blocks and split them cleanly--except for a few that have
cracked themselves and show the seam at their sides. There are usually
a few of these.
Without doing this crucial step, you end up with a real mess! Dave
shows a block divided, but interestingly, the shellac is still around
the pieces. All I can say, after having done thousands of them in my
career is, "How did he do dat?" Also, re-sealing old pouches with more
rubber cement is the worst thing that can happen to an Ampico. (I
don't recall if this was advised, however. I do know it is often done).
Cleaning ball bleeds (as he suggested) is worse than leaving them
alone, since a ball bleed leaks as a result of internal corrosive
pitting-- discovered by microscopic examination. The original ball
bleed may start sticking, but it won't leak as badly as long as it
isn't "cleaned." It is the packing of rubber particles, dirt, and lint
that still provide a seal-- of sorts. (Replacing the expression
curtains, by the way, makes the singularly greatest difference in
expression. After that, then, the expression valve pouches all around,
and ultimately, the stack valve pouches).
If player pianos are anything, they are valves. To the degree that the
valves are restored and operating correctly is the degree that the
piano will perform as originally envisioned. These, however, are fatal
exceptions that Dave describes no tests for, and then documented in his
article on building a new set of block valves of his own design, he
admits that he couldn't get the original ones to perform well.
I personally believe that the Ampico B valve is the finest valve ever
made. I have been dazzled by the results, once everything is back
together and rechecked and tested.
Hickman Diary: "Have about decided to buy stock in American Piano Co. on
strength of my single valve." (They replaced all the valves in the
perforating room with Hickman valves, too, because they were so far
> Some performances and editing jobs were better than others, too.
> B coding on B piano didn't necessarily insure aesthetic perfection.
Can't argue with that. I will even go Dave one further and say that
there were even rolls cut after the 1922 standardization that the model
B didn't play as well-- called the Hupfeld Roll Series. However, they
were largely un-standardized, and only about a dozen or so. And there
were others as well, but very few. (So what have we proved?)
> Question: Why did Ampico re-code many rolls originally having "A"
> coding into "B" format if their roll library was already fully
> compatible like it was? The most reasonable explanation would be
> that the "B" format was considered necessary for best performance on
> the new piano, so editors were put to work to upgrade performances
> having good future sales potential.
That's easy. By far most of those rolls were not basically re-edited,
but merely added to, enhanced to take advantage of the amplifier hole
and the sub-intensity capability. Other than that, no changes were
required in the vast majority.
Again, this is not to say that they did not re-edit some rolls. I hope
I will not be taken out of context about this, because they certainly
did do some basic re-editing-- but realistically, very little. Here is
what Angelico Valerio said about Dave's question about compatibility
and the new rolls:
The "enhancement" of adding the two new holes to old rolls began in
1926, according to Valerio.
Barden: (pg. 223) "If you knew you were editing for the new machine,
starting in '26, did you find new techniques (coding tricks) that were
Valerio: "No, but you had greater range." (In other words, they didn't
recode because they didn't have to! They just enhanced the range of
those rolls for the benefit of the B.
The idea was to build a piano that would play the old rolls as well or
better, and then, if a customer didn't have the old cut, well, the new
cut was the same, but included the extra expression holes. So the store
might resell that same tune to someone who traded up for the new
Ampico, if they liked the difference enough. SAME expression! Greater
range. Read the book!
Barden: "Peter Brown said that you fought with Stoddard a little bit
over having the "B" rolls work on the "A" piano and vice versa."
Hickman: "Yes, I didn't get anywhere with him! Well, he couldn't get
into it because that was the policy set by the president and the vice
..... But we finally doped it out so that we didn't lose too much."
Barden: "But it's noticeable, and there are some sections where that's
Hickman: "No, I think we did a pretty good job on that. It looked like
a pretty tough problem when it was first patented." (The Ampico
Reproducing Piano, edited by Richard J. Howe)
What Nelson Barden was reacting to is exactly the same thing so many
rebuilders and owners of this piano react to -- a degradation of
performance, under that criteria which Hickman so carefully tested for
and proved in the original pianos. And what could that be? I propose
that the reason is "Pouches."
Rubber cement gets hard and takes a dish-like set, slowing the pouch
and weakening the sub-intensity. And the expression curtain, which is
also rubber coated cloth is today a bit stiffer than it used to be-
sealed or not. While it's still fine for a pneumatic, since a movement
of just a thousandth of an inch across a wide grid is all it takes to
make a large change in intensity, replacing it is a necessity to get
that original kind of resolution! Simple though they may be, new they
ain't! Everything in that piano is as old as everything else. Replace
it! That's what they call a restoration, Dave.
> Question: Why was the "B" roll format introduced two years before
> the B piano was being shipped to dealers? Wouldn't it be to make
> sure that dealer's shelves were well stocked with the latest issues
> of "B" rolls by the time the new piano was introduced?
Actually, the B roll format was introduced long before that! If you'll
just read Dr. Hickman's own diary, Dave, you'll see that he was working
on the roll coding before 1925, because the first fully operational
model Bs were demonstrated in two models then! However, he had some
more things he had to "dope out," even after the president and vice
president wanted to build it as shown to them that year.
Remember too, It was company policy that dictated complete compati-
bility with the A rolls (not the A piano!) That was a resolution
demanded by the president and vice president, as well as Mr. Stoddard.
And they were a stone wall on the issue!
I think it should be obvious now that, from Dave's own book, Rebuilding
The Ampico Model B, that until we decide to restore the piano complete-
ly without excepting anything (especially pouches), and choosing
leather for them without pinholes and fast enough to play the test roll
satisfactorily (.008 to .009 thickness is maximum), always replacing
the old ball bleeds, then sealing that leather with something other
than rubber cement (Ampico's Achilles heel), we will never be able to
knowledgeably, adequately or honestly evaluate the Ampico model B.