With all due respect to Dr. Peetz, there are some statements in
his posting about the Welte-Mignon recording system that I want
to comment upon.

I do not believe that any springs were used in the recording system
carbon rod device.  Adding a spring to the attachment point of the
carbon rod to the piano key would defeat the whole idea of recording
the instantaneous rate of the key's velocity, by adding hysteresis
to the system.

I believe the tiny carbon rods were firmly attached to the keys.
Having a spring in the loop, and then considering the surface tension
of the mercury, would destroy any accurate measurement of velocity.
The carbon rod must be precisely guided into the mercury, and firmly
attaching it to the key would provide this accuracy.

In his item #5, Dr. Peetz gives exactly the reason why springs were
detrimental to the recording: the key and carbon rod must follow most
precisely the movement of the key.  A spring would hamper this by
giving additional uncontrolled varying movement to the carbon rod;
a guided spring would add mass to the key and no doubt anger some
temperamental pianist.  The carbon rod itself adds insignificant weight
to the key, undetectable when playing.  Obviously, Welte had to have
precise positioning of each carbon rod versus the mercury level, but
this was well within their capability.

I do believe that what was used for determining the dynamics of the
recording was the rate of the increase of the slope of the inked track
width on the paper.  The faster rise of the width of the line was the
indication of the rate of the key stroke, and thus the intensity with
which it was struck.  On the roll fragment that Dick Simonton showed
me, a most careful examination of the lines did show a difference in
the rise time of the individual lines as a key was struck.

I also do not have, and never had, any of the carbon rods.  Dick showed
me his little box with a few of the rods in it.

Dr. Peetz brings up another interesting point that was never discussed.
What was going on in that mercury trough when the artist was blasting
away with some dynamic Liszt piece?  Was it splashing all over the
place, or perhaps creating waves, as he mentions, and thus influencing
the dynamics of the recording?  Or could Edwin Welte have foreseen this
and used separators of some kind between each carbon rod: baffles,
individual tubes?  Who knows, but now this does make sense.

The "conductive line instant playback" capability --

A recent conversation with Karl Petersen brought some new viewpoints
to mind that his valued consideration has produced, all concerning that
intriguing argument on whether the Welte T-100 roll could instantly be
played back in a special roll reader, or whether it had to be punched
first.

Karl is of the opinion that what Welte did was to delicately read the
resistance of each line on the roll and then, with a most lightweight
and tiny magnetic valve, vary the bleed hole size to control the
dynamic playback.

Consider an aluminum tapered needle that is constantly varying the
bleed size, and thus the vacuum level in the pneumatic on a constantly
varying level -- a most interesting and possible way that they did
indeed directly read the roll.

What certainly does come to mind is just how they measured resistance
with a constantly varying conductive line in 1903.

Karl also mentioned that, in his opinion, Welte did not use any oil
floating on the surface of the mercury, but that they did drain and
filter the mercury from time to time to remove the oxide that floats
on the surface.  What does now come to question, is that if an oil
was used, did it impregnate the carbon rods and cause variation in
the resistance readings over time?  Or did they have something of
high surface tension that would not enter the carbon surface?  An
interesting question.

Karl just may be correct: they did not use any oil floating on the
surface, but filtered the mercury from time to time.

Whatever the arguments produce in the end, the Welte T-100 system
did indeed record and reproduce individual note dynamics.  Listening
to a 9' Steinway played by a perfectly working Welte Vorsetzer is the
most inspiring and dynamic reproduction of piano music on paper rolls
that I personally have ever experienced.  The Duo-Art and Ampico
systems are wonderful and can be very realistic, but the Welte is
still the king, in my mind.

They did use carbon rods permanently fixed to each key.  They did
record the varying resistance vs. time as key velocity.

Welte could have directly played back the master roll, if they used
a varying bleed driven by a tiny voice coil type of magnetic valve,
as Karl proposes.  At least I now think they could have done it this
way.  Perhaps this one will never be understood.

What surely would be not only historically of value to all of us,
but would settle some questions that are now almost lost to history,
is to have the T-100 Welte system reproduced, and to again see just
how this was done, with the variations mentioned, just to settle the
arguments once and for all, to revive the Welte recording technology
as historically accurate as humanly possible.

The great pity, at least to me, is that no one has seen fit to write
the definitive book on all the major reproducing piano systems.

The data is scattered and getting more scarce by the decade.  The
Vestal Press book on the Ampico system is out of print and nothing
was ever written about the Duo-Art system in book form.  Scattered
articles in various hard-to-get journals are all we have.  Collect it,
verify the facts, expand upon it and put it in book form!  I would
certainly contribute funding to such an effort.

Couple this loss with the additional almost total lack of information
on the Welte Philharmonic residence organs and their players.

Data seems to be lost.  As I am more than passionately interested in
the music these rolls contained, the lack of a Welte shipping list is
particularly painful.  We know the Wurlitzer, Skinner, Estey and Moller
organs and who bought them for their homes.  The Rollin Smith book on
the Aeolian residence organs and their music is certainly well documented
and is a stunning piece of research; but where is the same information
on the Welte organs?  Who has the definitive shipping list of
Philharmonic organs?

All great fun and a good puzzle for the technology historians of
mechanical music reproduction.  We don't have the answers yet.

Jim Crank