Pete's Duo-Art is never going to play right as long as he believes that
he shall do it by ear and a dB meter. Pete, there is a right and a wrong
way to regulate these things, and if you have decided that you are not
going to leave the accordion values by the book because the piano
already plays too loud, then you are beginning on a flawed premise.

If you will simply trust me about this and do it right, you will
eventually have a perfectly playing Duo-Art. If you want to wing it,
then you're on your own. Nobody could help you. Don't worry, if we need
to voice the hammers, then we will. But that isn't where you should
begin! The tinniest hammers in the world would not produce the complaint
you have, altogether. Besides, you have already described the piano's
sound as "wonderful." That doesn't sound like hard hammers to me, so I
suspect you are not wearing a hearing aid.

When you described your response chart as non-linear, actually having
the reverse response of the hysteresis curve, you fully explained to me
why your piano seems to play too loud, Pete. It's because you have
already "gray-scaled" it in an attempt to make it quieter, but your
first two accordion steps travel too much, in an attempt to overcome the
incorrect felt spill (which has never been changed). When your spill
valve is mostly closed, its oversize hole is out of the circuit, and the
extreme slope of the curve between 0 and 4 jumps up your other values as
they add in to the total at the higher intensities. However, your pump
is only capable of so much total power, so your midrange is too loud and
your upper range is not that much louder.

Now one reason some people have trouble at the low intensities is
because the prescribed travel of the first accordion (1/16") isn't
enough for them to even measure the difference on a vacuum gauge. They
may have one of those nearly worthless Marshalltown gauges, or they may
have a linkage problem with the box. The difference should only be,
like, 1/4" of water pressure or a little more, perhaps up to _almost_
1/2" at the very most, and that is often hard to see. But too often,
that is set up higher so it can be seen on a cheap gauge. When that is
done, it acts as a multiplier to the louder intensities (loudness-wise).
It sounds like this is what you are describing.

Another problem you can encounter is spill valve interference with the
top of the box when all the accordions are pulled down. So once you
adjust the accordions to their correct travel, you will clamp them by
hand to be sure they are all bedding with their adjustment screws. If
they do not, then you have a travel problem which must be corrected
before you can go further (If so, I'll help you with that. It's easy).
So if you feel the top board of the accordion stack tipping a little
when the accordion stack is clamped, you will know that the travel is
too much.

I also want to stress that you cannot clamp the accordions by hand or
fingers and get an accurate travel adjustment! So measure from the hook,
never the leaves, and let the vacuum do the pulling. (That's why it's so
much easier to do this on the bench.)

The reason we are starting at the expression box is because the box is
the "heart" of the instrument. You should use the note tester on the
test roll and set the piano to Duo-Art, instead of Duo-Art Off. After
you have the first accompaniment int. #0 adjusted as perfectly as you
can using the accompaniment arpeggio on the test roll (and sure that the
Theme is not interfering), Play the full compass of the piano notes a
number of times, first with the sustain pedal off, and then with it on.
The piano should play each note softly with the pedal on, and should
miss a few with the pedal off. In other words, the damper weight should
put the notes at the threshold of player actuation. Chalk the notes
which are too loud, and chalk the ones that don't play. Once all your
notes play well, all the way up and down the keyboard at both of these
pedal settings, fine. We can progress to the next step.

You also may have too little travel in some of your expression valves
(the 8 valve box), your Theme may be bleeding over into your Accompani-
ment through the theme cut-out flap valves, your stack valves may leak
too much, your pump may lose power during high demands, etc. All these
possibilities can cause you to compensate so that the overall effect is
just "too loud." Actually, your piano is not loud enough at "crash"
intensity, and way-too loud in its middle range, although just fine at
the lowest intensities (as they always are). Don't let the lowest
intensities fool you. That is your set point.

I will guarantee you that if you will follow along here, your piano will
finally play correctly, and nobody who loves real, live music is going
to say it plays too loud. Stop and think: If new reproducers all played
too loud,  how could they have ever sold them?

Also, in regard to the comments about the Ampico loudness with the
Aberdeen Spark Chronograph, the comment by Dr. Hickman was quoted by him
from the spur of the moment. We cannot be sure why that phenomenon
happened, but I didn't find it in his diary, so possibly he discovered
something different later, yet remembered his first reaction. It very
likely could have been overcompensation by the pianist due to the fact
that behind the black curtains lay a vast expanse of factory which he
was obligatorily playing to, absorbing the piano's sound. Or the piano
may have had soft hammers. The artist may have been accustomed to a
heavy action, or some recording technician may have accidentally
maladjusted something. Duo-Art or Welte never mentioned experiencing
such a thing. I have always had my doubts about that one.

Granted, Hickman sat down and listened to the playback without the
artist, but on a different piano! So it was simply a scaling problem.
The rate of change of the loudness curve was unrealistic, for whatever
reason.

I question the "presence theory" because we can take the same rolls, put
them on concert grands, and play them with an orchestra. If the piano
has been set up properly, it is so beautiful it sends chills down your
back. You don't get more presence than 100 musicians and a conductor. So
if in fact THAT is what someone called the degenerate 20's version of
"Musak," then I really don't need to hear the live performance. I will
be completely satisfied with their "Musak." (Granted, the concert pianos
run faster to take full advantage of the wider dynamics of a large
piano, but that doesn't change the roll coding).

By the way, there appeared an artist recently (I don't know who it was,
the show was in New Zealand about 1993) who made reproducing rolls. He
claimed that between the Victor recordings versus the rolls, that he
much preferred the rolls as an exact duplicate of his own playing to the
Victor records. Maybe someone has seen this program and can elucidate.

Craig Brougher