The effort of Terry Smythe to start a scanning project of the Aeolian
Pipe Organ 116-note is commendable, but the devil is in the details.

To answer his inquiry regarding spacing, the two rows are spaced
exactly 0.25 inches apart, center to center, vertically.  Each row has
holes spaced at 6 holes per inch, which means the staggered holes total
12 per inch.  On the left side, the lower row is the first hole which
corresponds to note "CC" on manual I, proceeding to the right, the next
hole is on the upper row and it is also "CC", but is for manual II.

As Mike Walters stated, the operator of these rolls stays quite busy as
various printed instructions pass over the tracker bar.  The operator
must set stops, operate the swell pedals, operate the crescendo pedal
and make changes to the tempo and "coupler" which is explained below.

Within the printed instructions are commands that make proper playback
of scanned roll an almost impossible task.  There are numerous tempo
changes on the roll.  Some are completely new settings of the tempo
lever, say from "Tempo 35" to "Tempo 50".  Others are "rit.", or "accel."
or even the "bird's eye" pause.  These instructions can't make it
through the scan process without some kind of manual editing.

To make the task even more daunting, the "coupler" system has three
positions.  In "Normal", the upper tracker holes play manual II, and
the lower tracker holes play manual I.  In "Reverse", the upper holes
are now played on Manual I, and the lower holes are played on Manual
II.  In the third coupler position, "Unison", all tracker holes play
on both manuals.

In the recent discussion about recording 88-note player piano
performances, the overwhelming sentiment is that a really good roll
needs interpretation, or expression for the recorded performance to
have true merit.  The lifeless drone of expressionless recording is
just a token of the lack luster performance one will get when playing
a MIDI file of the Aeolian 116-note roll which has not included the
necessary changes in tempo, stop settings, expression shades, and most
importantly, coupler assignments.  Without that information included
in the MIDI file, the playback can at times be unrecognizable.  This
is because the "Melody" may be on the wrong stops, and may be too soft
to be heard.

This scheme of "couplers" even made it into the more advanced, fully
automatic Aeolian Duo-Art pipe organ rolls.  The earliest versions of
those rolls flip-floped the tracker hole assignments using "Normal" and
"Reverse" coding in the automatic coding of those rolls.  The practice
was short-lived as the editors realized it is much easier just to
perforate the rolls without the required switching.  The earlier rolls
(116n) could have been made that way too, but for some reason they were
not.  The coupler system is totally unnecessary if the rolls are
perforated with the switching incorporated into the perforations.

The bottom line is that any roll scan must also incorporate the actions
of the printed roll instructions.  Without that information, a MIDI
playback will not only sound bad, it will not accurately preserve the
data on the roll.

I have a large library of these rolls and will at sometime read them
into MIDI files using a pneumatic roll reader.  The recording process
will incorporate the necessary tempo changes and coupler changes.
It will not be an easy task.  With my roll reader, I have recorded my
entire library of the 176 note fully automatic Duo-Art organ rolls with
near perfect success.  The 116 library is on my "to do" list.

Bob Taylor