[ Ref. Player Pianos vs. Live Performances

I would like to add a further comment to Julian Dyer's insightful
observations regarding row spacing, or incremental paper advance of
piano rolls [180506 MMDigest].  The timing accuracy of music on piano
rolls can be improved in two ways.  One, as Julian pointed out, is by
placing paper advance increments closer together.

The second method is by moving the paper faster.  For example, a roll
recorded (or processed to play) at Tempo 100, corresponding to 10 feet
per minute, will need only half the time to advance from one increment
to the next as compared to an equivalent performance recorded or
arranged to play at Tempo 50.  In other words, resolution can be
improved by running the paper faster.

This is, of course, not possible in systems such as Welte-Mignon
that employ a fixed paper speed for all performances.  However, I have
observed that the Ampico editors often used a higher Tempo setting,
corresponding to increased paper speed, for performances requiring
a high degree of accuracy during rapid note sequences.  For example,
Ampico 71521, Mendelssohn's "Spinning Song" performed by Rachmaninov,
calls for Tempo 110.  The "Flight of the Bumblebee" calls for even
faster paper speed, at Tempo 115.

The Ampico people never used greater paper speed than needed, because
it would have been a wasteful added expense.  So it seems pretty
clear that the higher Tempo settings were considered necessary for
a reasonably accurate reproduction of the music of a fast and complex
character being performed in such examples as these.

Dave Saul
El Cajon, California