The questions and points raised by Bernt are interesting.  I would like
to respond with my opinions on the subject.  First, the issue of the
nomenclature when discussing the sixteen intensity levels.

Bernt, it becomes far less confusing if the word "level" always equates
to the binary result of the position of the accordions.  True, there
is an error in the way Kitner describes his gauge readings on page
nineteen of his service manual -- I suspect the last three levels in
each column should be labeled "11", "14" and "15" -- but it really
makes no difference anyway.  I feel you are placing far too much
importance on this table which really is of little value in correct
regulating.  Kitner even says so.

Page sixteen of the Aeolian service manual uses the column header
"No." (short for number) and you have chosen to supplant that term with
"level" which indeed is confusing.  In fact, the term "level" is not
used anywhere in that table.

Enough about that.  Let's all agree that level ten is where the spill
valve is first completely closed.  Level ten occurs when accordion
positions 2 and 8 are collapsed.

It has been my experience when adjusting grand Duo-Art boxes that the
spill valve adjustment is a complete no-brainer.  With the correct zero
adjustment set with the test roll and the accordions gapped as listed
in the table, simply loosen the set screw on the spill control lever
(the one with buckskin) and let it just touch the knife valve control
lever (on both sides of the box).

This will result in the spill valve closing as the accordions first
reach level ten.  If it is off slightly, no big deal.  The Duo-Art
expression box was not designed by rocket scientists.  It is crude at
best.

Don't get me wrong about my last comment, I love the Duo-Art and the
music it makes, it is really quite good.  But as a control mechanism,
the Duo-Art Box is really not something that warrants all these graphs
and measurements.  The only way to make a Duo-Art NOT sound mechanical
is to adjust it through trial-and-error by listening to it -- many,
many times.

Here is why.  As you point out, it is an open loop feedback system.
As such, the individuality of each piano much play a huge part in
regulating that piano.  We all agree that the weight of the piano
action pretty much controls the "zero" level adjustment.  Whether it
is 5 inches or 4 inches, it just doesn't make any difference as long
as the notes don't skip when playing very soft music.

But what is the vacuum at level fifteen (with the crash valve open)?
It is based on the tightness of the complete player plus the size of
the pump and has no bearing on the piano action at all.  In the Kitner
table the highest number is 40 inches of vacuum.  My own Steinway XR
goes to 85 inches when the crash is on!!  I've seen others that go to
75 and some 60 inches.  So right away you see that any table of values
is really worthless.

 [ Many Duo-Arts have no crash valve, or it may be disabled for
 [ certain tests.  This should be noted in any tables.  -- Robbie

My personal technique is to make the piano perform the test roll -- the
sections of play and don't play, and then start listening to good
music.  Slight adjustments to the accordion gaps can be used to smooth
out unevenness or even soften the top level of play if the room is too
small or the piano too harsh sounding -- whatever.  When finished with
all this, go back and readjust the crash and spill as necessary.

One final thing, the pump spill pressure is highly variable, depending
on pump size and tightness of the various control systems .  On a very
tight piano, a low spill setting will result in a huge amount of air
noise at the spill inlet.  I think compromise is the best guide here --
it is not rocket science.  Again, how crude can this be -- just
stuffing a wade of felt in a hole?  A spill setting in the range of 25
to 40 inches is OK.

I know that some rebuilders, after making initial adjustments, then
plot the pressures for that specific piano, and make the resultant plot
a smooth line by making slight adjustments to the accordion gaps.
While I have never done this, it seems to be the most sensible way to
use a chart.

Duo-Art regulating is quite subjective and setting pressures to conform
to a table that doesn't allow for all the characteristics of the
complete piano system, will result in a performance that is mechanical
in nature.  To avoid the pitfalls of making the regulation another
"open loop system", the technician must close the loop with his good
ears and knowledge of good musical performance.

Bob Taylor