Ignoring the tone of previous responses, there is still a point at
issue here about the basic aims of rebuilding.  It's misleading to
interpret how machines 'ought' to work, because in reality we should
consider how they were meant to work by their designers.

For the Duo-Art, modern thoughts on the logarithmic aspects of sound
pressure tend to get superimposed on the original idea and confuse
matters.  Rolls were coded to sound right on editing pianos, and we
should try to reproduce that setting to make the rolls sound good now.
Any redesign or reinterpretation of the machine, no matter how much of
a genuine improvement, will harm the reproduction of original rolls.

Considering as we were the relative levels of theme and accompaniment,
I was trying to convey the idea that, given the Duo-Art practice of
coding theme dynamic levels equal or higher than accompaniment, the
least possible accent will be when a note is accented with equal theme
and accompaniment levels set.

Clearly this least possible accent needs to be relatively small to
permit subtle playing, and is the reason the Duo-Art manual calls for
'one degree' of difference, whatever that may mean in practice.  Making
the levels hugely different will erode subtlety, but will obviously
give greater maximum accenting.  I don't think it's of any consequence
how many 'f's you have in the crash if the piano-mezzoforte range is
unmusical.

I rather doubt that there's really much to choose between the variants
of Duo-Art expression box design.  I suspect that the London operation
never understood things as well as the original designers, and in one
or two aspects I suspect the design slipped in later years.  However,
as Paddy Handscombe has suggested, the effect of setting up the
expression springs, whether equal or different, probably comes to much
the same thing.

Sadly, not enough Duo-Arts work well enough to show the details we've
been considering.  Given the other dissertation on Duo-Art soft pedal
coding, maybe it's lucky Aeolian did as they did because there are one
or two instruments around where the soft pedal gives the only audible
change in dynamics!

Julian Dyer

 [ "Expression via the hammer lift" is remarkably effective, especially
 [ as used in European pianos and orchestrions.  I hope we can explore
 [ this topic someday in MMD.  -- Robbie