Hello all,

Lots of discussions going on about the Duo-Art.  I thought I'll add
some of my humble findings to the subject.

1)  All the Duo-Art pianos I have seen so far except for one Weber
upright brought here from the states recently have round valve plates
except for the pedal control valves.

2)  The pellet valve on the soft pedal which activates #2 accompaniment
pneumatic is present in all the models I have seen so far.

These are all as installed by the factory.  The Duo-Art pianos we have
in this country all seem to be made after 1927 or so and most of them
are the so called export model which run electric and with pedals.

I deduct from this that Aeolian must have recognized that the round
valves are better and they also knew about the lost motion with soft
pedal 'on'.  They would not have modified their pianos without reason
or what?

The next issue I would like to comment on is the zero adjustment using
the test roll:

The zero-adjustment section on the test roll which plays the arpeggio
with pedal on and then with pedal off is completely inadequate to my
opinion unless the piano is in a like new condition.  It is not
responsible but allows for an incorrect zero adjustment.

I state this because the arpeggio only plays a couple of notes on the
piano but does not give a valid picture of the whole piano and each
note.  One can have this test pass the play and not play but the rest
of the piano can be no play for both tests.

This is due to mechanical resistance differences between the different
notes of a piano.  It can be because of some centre pins more loose
than others, spring tension, felt friction etc. etc.

This problem can be easily verified as follows: Run the test roll tempo
section that tests each note for repetition (tempo calibration section)
but at a speed of about 30 and with Duo-Art 'ON'.  You must press the
sustain pedal while doing this.

This will simulate the zero test but with all the notes, one for one.

Notice how inaccurate the zero adjustment that you made with the
original test really is, some notes play louder than others and some of
them skip altogether.  The louder ones are usually in the mid section
as this is usually the most used and worn.(less friction)

Now try to re-adjust the zero levels using this test and adjust it so
that all the notes will play (no skips allowed here) with the pedal
manually pressed.  Needless to say, try to keep them as soft as
possible.

You have now adjusted the piano to the minimum usable zero level and
you will find that no notes will skip in the very soft passages of the
music but the original test for zero adjustment will most probably fail
the arpeggio with the pedal 'off' (notes should skip).

Also, you might find that the overall playing in the soft passages is
now to loud.  If this is the case, you will either have to tolerate it,
or bring the zero levels down again and then tolerate missing notes.

The proper way to remedy al this is to ensure that all the piano notes
play with an equal force.  This you can only accomplish with a special
lever jig system which can, with the player action removed, lift the
wippens with a certain force (with pedal pressed!).  I repeat, the
wippens, not the piano keys, as there is a difference.

The adjustments can be made by adding or removing spring tension to the
hammer-butt spring and the jack-spring.  This can only be done in
limits of course and other problems like loose center pins, regulation
etc.  must be corrected first.  Also, the leather nuts on the player
action pneumatics must not be too tight.

Once the piano has been corrected this way, the above zero test must be
repeated and one will be surprised how soft the piano can now perform
without skipping notes or the odd loud note.

It goes without saying of course that the player action and all other
components, valves etc.  must be in proper working order before doing
the tests.

One other cause of skipping notes can be a sluggish theme valve and
theme-perforations (snake bites) which are too small (QRS Duo-Art for
example).

As a matter of interest, here is some science-theory to keep in mind:

All of these problems are due to the fact that the Duo-Art, like all
original reproducers is a so called open loop system.  This is a term
which comes from control-theory and means that there is no feedback
from the output of the system back into the system to compare and
correct the real output to match the desired output.

In the case of a reproducing piano this means that the piano's output,
the  volume of the note played, is not monitored by the expression
mechanism to match it to be exactly the loudness desired by the input,
the music roll'.  In other words, the piano does not know how load or
soft or whatever it is playing for a given input.  Therefore it cannot
correct any errors in volume due to differences in friction etc.  The
result is that notes are too loud or skipped or too soft.

Such an open loop system relies completely on the consistency of all
its components and calibration thereof  to produce the correct output
for any given input.

This does not hold true for a human pianist and to some extend to
the normal foot-operated players as the ears of the pianist/operator
provide feedback from the volume and other characteristics of the notes
played back to the system via the fingers or the feet (player).

This means that skipping soft notes as experiences in a tremolo for
instance will immediately be sensed by the pianist and his fingers will
adapt the striking force so as not to skip the notes.  Also, in a
manual player, if it becomes too loud or skips notes, one immediately
pumps it differently to compensate for this.

Therefore, I can safely make the statement that no two reproducers
can play the same roll exactly the same due to differences in their
mechanical characteristics, their location and even the weather.  All
we can do is to try and equal out as many of these differences as
possible and to provide the initial calibration as accurate as
possible.

I have been told that one or two modern new electronic systems now have
appropriate sensors to measure the speed of the hammers as they travel
towards the strings and to adapt the solenoid force to produce the
correct speed (= striking force) for a given input, thereby eliminating
most of the differences in friction between notes etc.

Any comments about this out there??

Regards,

Bernt W. Damm, Restorer of Automatic Musical Instruments,
Cape Town, South Africa
Tel/Fax:  +27 21 592 5124