The subject regarding the lack of tempo compensation in Ampico rolls
is also interesting from another angle.  When I was just a little kid
I visited my grandparents once a year.  They had a player piano that I
could not get enough of.  Then when I got back home I had the rest of
the year to try to remember how it played certain tunes.  I learned to
play by ear as a result, but, to this day, I cannot keep a steady tempo.
I invariably speed up toward the end, just like the roll!

Later, when I started rebuilding reproducers, I noticed that the music
on Ampicos and Duo-Arts kept a relatively constant speed.  Maybe not
perfect, I have never checked, but for sure its much better than the
old 88-note player rolls.  So there seems to be more to this than
simple tempo compensation -- otherwise, the music on an Ampico would do
the same percentage of speed-up that the same size (30-35 foot) roll
does on the 88-note pumper player. Why, if no compensation were used,
doesn't the Ampico enjoy the same (approx.) 15% change in tempo by the
end of the roll?  We realize, of course, that it would totally negate
the authenticity of the music were it to have done that to the same
degree.

When the same rolls are played on the Ampico B with the large take-up
spool, I haven't noticed any difference.  So if there is a difference,
it must be mostly academic.  Charles Stoddard was a very discerning
gentleman.  He was picky and loved high quality and small details in
everything.  I am surprised that if someone had such an important
objection as serious as tempo buildup, that it would not have been
addressed immediately.

We must also keep in mind that if there were, in fact, tempo problems with
the Ampico, Aeolian Duo-Art would have certainly capitalized on it.

We shouldn't forget that both of these reproducing pianos appeared on the
world's great concert stages together with the world's great musicians,
performing -- live -- at same piano.  They were called "Comparison
Performances."  I was told that, at Carnegie Hall, scorecards were given
to New York music critics who then were to guess when the pianos were
being played by the musicians, and when they were being played by the
roll.  Critics usually have a  well-developed ear, and nobody at any time
ever wrote a word about escalating tempos.  The consensus was that it was
really impossible to tell who was doing the playing, I understand.

Now I suppose there could be tricky ways around a tempo buildup problem
on-stage for your average concert-goer, but I have met those who seem to
have a metronome built into their brain, and fifteen minutes later can
tell you if you are a beat too fast in a dozen measures.  You aren't going
to fool everybody.

I'm not that sensitive to tempo, so I have to say I can't really hear what
you are speaking about.  Could it be a combination of roll brake tension,
transmission gearing, and air motor governor compensation that varies
between pianos?  I have corrected tempo problems also with the right
lubricants as well.

By the way, if anybody is interested, the percentage of tempo buildup is
proportional to the increasing length of the roll times the thickness of
the paper.  That in turn is equal to the difference in cross section area
of the take-up spool and the paper when it is spooled on the take-up
spool. So A' - A" = Ap = L X thk. Where A'= cross sectional area of paper
plus take-up, A"= area of take-up spool cross section, Ap = Resultant
cross sectional area of the paper itself. L = length of the roll, and
thk + thickness.  Percentages would be really easy to figure out as the
roll paid out.

Craig Brougher