-- forwarded message, please reply to sender and MMD --

 [ Ref. 030927 MMDigest.  Karl Petersen and I shared some thoughts
 [ with Dr. Peetz; he asked that this message in reply be published
 [ in MMDigest.  -- Robbie

Dear Karl, dear Robbie,  thank you very much for your interesting
remarks.  In the next days I will answer in a more detailed way.  Now
I want just to comment on the question of the correct point of dynamic
measurements.

Obviously, the hammers final velocity in the moment when it hits the
string defines exactly the loudness of a tone.  Certainly there have
been some 20th century pianists playing the piano experimentally by
beating the hammer directly to the string (or even playing the strings
by the fingers like a guitar).  But most of the pianists prefer the
classical way to play the piano -- by the keyboard.

At least the world's best pianists are able to control the loudness of
a tone perfectly, not directly by the hammer velocity but by the key
movement.  If the time-dependent keyway-function of an artist playing a
tone can be measured and reproduced exactly, the dynamics of this tone
will be exactly reproduced as well.

However, it is hard to believe that in the early 20th century an exact
measurement of the time dependency of the key position was possible.
In addition, it is clear that the Welte-Mignon T100 reproduction system
is not able to reproduce a given time-dependent function of a key
position.

But, fortunately, the construction of the piano mechanics allows a
considerable simplification.  Basically this is a lever system, by
which the hammer is driven until the letoff point is reached, where
the hammer begins to fly freely towards the string.  On the one hand,
the velocity of the hammer at the letoff point is exactly given by
multiplying the momentary key velocity at the equivalent key position
with the corresponding lever ratio.  On the other hand, the hammer
velocity at the letoff point determines exactly the final velocity
after the free flight phase when the hammer hits the string.  Therefore
the _momentary key velocity_ at the hammer letoff position determines
exactly the hammers final velocity and hence the dynamic (loudness) of
a tone.

Fortunately I can prove that in the time around 1900 it was technically
possible to measure the key's momentary velocity by a Hg-C [mercury-carbon]
sensor combined with an inked disk writer system, even if the relation
between line width and key position is very non-linear.  I cannot prove
that this was the way followed by Karl Bockisch and Edwin Welte, but it
seems so to be.

After the hammer has left the lever system and flies freely towards the
string, the pianist cannot change the final velocity any more.  What
happens after the letoff moment is of no importance.  (It has to be
noted that the hammer letoff point of the key is still above the first
felt contact: the felts influence and its variations with humidity,
temperature, and so on play no role.)  This fact was already known in
the early 20th century and, e.g., reported in the book by Ludwig Riemann:

  "Das Wesen des Klavierklangs und seine Beziehungen zum Anschlag"
  [The nature of the pianos sound and its relationship to the touch],
  Verlag Breitkopf & Haertel, Leipzig 1911, p. 238.

A little remark: The knowledge of this book is a must for everybody
interested in dynamic recording of early pianos.  It describes the
pneumatic recording technology of Hupfeld's Phonola system which worked
already in 1906, when Ludwig Riemann assisted on Edvard Grieg's
recordings of own compositions for Phonola.

Best regards,

Ludwig Peetz
Pirmasens, Germany