Just in case there be anyone left who is still interested in the
Duo-Art, may I request that if they read my postings they note my exact
use of my native language, so that statements that I have never made
are not attributed to me.

Let me be clear.  Since long before my contributions I have never been
in any doubt that Duo-Art regulator springs were the normal type with a
linear rate.  However, this does not predicate, and I have never stated
and do not believe, that the rate of the box is determined by the
springs alone, or was intended to be.

Rebuilding player pianos since the late 1960s, I have taken measurements
frequently on the test-bench as well as on complete instruments.  For
accuracy my preferred vacuum gauge is a 50" water gauge (U-tube), with
a calibrated Venturi orifice plate to measure mass flows.

The plots I have invariably obtained for Duo-Art expression boxes in
and out of their pianos have been exponential curves and never crooked
lines.  My aim has always been to achieve glitch-free curves and correct
interlacing without any form of fudging.

Anyone who has played about with reproducing pianos for a reasonable
amount of time knows just how important it is to obtain the maximum
possible subtlety and thus greatest possible number of levels at low
dynamics for them elicit all the nuances in the rolls and to sound
convincing.  I do mean convincing to professional pianists and
musicians, and they do find a correctly regulated and voiced Duo-Art
playing the best-edited rolls musically unexceptionable.  Unfortunately
there are so many Duo-Arts played regularly in public which sound so
poor that most musicians believe they cannot reproduce pianists'
technique with any accuracy and dismiss them as having no place in
the history and study of pianism.

Votey and his fellow engineers must have known the importance of
low-dynamic subtlety and surely adopted the exponential curve gladly.
There is certainly no evidence from the plethora of Aeolian patents
that they made any attempt to obviate it despite their clear grasp of
various means of obtaining any desired characteristics.

Speaking of subtlety, I can assure readers that the 1914 US Steinway O,
which has enormous stack pneumatics, does indeed play pianissimo, pedal
on, without missing at 2-1/2" H2O measured at the stack.  Fortunately
it has large diameter pouches so the valves will lift and seat at this
pressure, but, as I said, to obtain this performance it required the
most sensitive pouch leather and careful regulation (Steinway 49 gm
touch weight throughout for the keys to bottom).

And by the way, English and US roll performances are quite
distinguishable on correctly regulated Duo-Arts.  They differ greatly
in editing approach, and Busoni noted that his US recordings sounded
different on the British editing pianos, and vice versa.  To complicate
matters, there are several versions of some recordings, varying in
quality, and rolls perforated in the UK from US recordings were
sometimes re-edited, some several times, and vice versa.

Patrick Handscombe
Wivenhoe, Essex, UK