Greetings,  I am not sure if Bruce Grimes' letter is a reply to my
previous one or the one from Douglas Henderson.  The point I was
trying to make is pretty much the same that Bruce makes.  These were
professionals doing an excellent job and they did not release sloppy
untested rolls.  How else can these recordings, that are approaching
100 years old, still sound great today.

My point was that there are minor variations in instruments that can
benefit from minor adjustments to the roll data.  Later rolls generally
seem better in terms or working on a greater variety of pianos that may
have slight adjustment problems.  I think that the experience of the
editors led them away from certain coding techniques that were more
troublesome.  For example, the coding of early British Duo-Art rolls
that work well on very few pianos was abandoned in favor of more
forgiving techniques.

It is impossible to expect that the rolls could be coded to work
perfectly on the range of pianos in which the Duo-Art was installed.
I would guess that important rolls were given more attention and tested
on a variety of pianos but others, such as the latest popular music,
may have been released with less testing.

I definitely can make improvements with just a slight change of the
pedal duration in certain rolls.  I rarely mess with the classical
music and I have the dampers adjusted so as to optimize that music.
Some of the popular music is just mushed over by overpedaling and it
can be improved.  Maybe the pops were optimized for smaller pianos.
I just know that they can be a bit much on my big upright.

It is also unrealistic to expect that an upright is going to repeat
the same as a grand piano.  The grand action design is just better at
repetition.  Some very subtle effects were edited into Duo-Art rolls
with adjustment of note duty cycle.  In some of these cases, my upright
misses the intended effect. It is quite rare but annoying when it
happens.

For example, the rapid repeats that happen at low playing level in the
"American in Paris" rolls sound too subdued and are on the verge of
missing.  This does not happen on grands.  This is a remarkable roll by
Frank Milne working near the end of the era.  I doubt he had the time
to test it on a great variety of pianos.  I don't think we can expect
perfection from every rolls produced.  We should be pleased that so
many of them are excellent as they are and welcome the opportunity to
improve the playing of others when we can.

Best regards,
Spencer Chase
Laytonville, Calif.
http://www.spencerserolls.com/