When my shop restored the Berry-Wood A.O.H.R. for the Sanfilippo
Collection in 1988, I had the opportunity to examine a number of early
and late rolls, dating from the early 'teens through the mid-1920s.
I was also familiar with the music arranging on Berry-Wood rolls, having
copies in my collection of several cassette tapes made by Lyle Martin
of one or more Berry-Wood orchestrions in California in the 1960s.

The early rolls that I've seen and heard were all made by the
coin-operated division of QRS, and the late ones by the Clark Orchestra
Roll Co.  That's not to say that other orchestrion rolls were never
made by Berry-Wood or another supplier, but the ones I've heard were
all derived from QRS arrangements.  They were probably converted to
Berry-Wood from Automatic A rolls rather than Automatic G rolls.  The
Berry-Wood versions retain some of the tenor countermelodies that made
it through the conversion process from 88-note to A, which were deleted
or moved up an octave in the G version to accommodate the G-roll bass
octave coupler that runs all the way up to G below middle C.

(When Tom Pletcher bought QRS and moved it to New York, Ernest Clark
moved the old coin-op roll division of QRS back to DeKalb and renamed
it "Clark Orchestra Roll Co."  He continued to use QRS masters for a
large variety of coin-operated rolls, including various types of
Berry-Wood rolls.  Thus, many early Berry-Wood rolls were made by the
coin-op division of QRS, and later rolls by Clark Orchestra Roll Co.,
all derived from QRS player piano arrangements -- the same as Automatic
A, G and H rolls, Marquette M rolls, Clark A, 4X and O rolls and
others.)

Berry-Wood pianos were sold during the same era that Peerless Elite
coin pianos and Peerless O-roll orchestrions (i.e. the Arcadian and
Wisteria) were being made.  Benjamin Austin, the staff arranger for
Peerless rolls, specialized in very full arrangements with many notes
playing at once, with mechanically precise rhythm reminiscent of
a marching band.  These arrangements were custom-made just for Peerless
pianos.  I believe the people at Berry-Wood wanted to simulate this
ornate, full arranging style but didn't want to employ a musician to
arrange their music from scratch, so they used two mechanical means to
convert the QRS arrangements into something more mechanical and robust,
to compete with Peerless.

First, the QRS marking machine that printed out the pencil masters for
Berry-Wood was tubed with octave-doubling in the bass and treble.  This
made an extra copy of each high note an octave higher, and each low
note an octave lower, on the Berry-Wood master.  Hand editing obviously
followed, for if you analyze the rolls you can see that the division
between the basic notes and the coupled ones isn't always in the same
place, but roams around with the melody line, to avoid making the music
too "thick" or "muddy," and to reduce the number of places in the music
where the melody runs out of the range of the piano stack.

This doubling was most likely done with tees in the tubing going from
the A roll reader to the Berry-Wood master roll marking machine, with
more notes teed than necessary, and then with an editor marking which
notes should be punched into the master roll and which ones should be
ignored.  That would have been easier than teeing too few notes and
having to add more notes by hand with the scale stick that is known to
have been used for orchestrating other types of rolls.

If the octave-doubled notes weren't enough, Berry-Wood then added a
pneumatic octave coupler, which coupled the treble octaves to make the
music even more "busy."  The original advertising copy for the A.O.S.
includes "coupler system giving greater range to pipe scale," and for
the A.O.W. "Contains...thirty-four wood flute pipes, thirty-four
violin pipes, coupler giving ninety-two pipe tones...."

In the A.O.H.R. that we restored, the coupler was gone but the screw
holes for it were present.  I chose not to reconstruct the coupler
because the music is "thick" enough as is, without coupling the highest
piano notes back down to play yet another octave in the pipes.  Also,
A.O.H. rolls contain no automatic registers and there is no mute device
for the piano treble, so both ranks of pipes and piano play together
all the time.

A Berry-Wood orchestrion is interesting and quite different sounding
than any other orchestrion.  However, since the basic roll arrangements
are the same as many other Automatic and Clark arrangements, transcribing
them to H or G rolls would, in my opinion, just create some rolls that
still sound like Automatic/Clark rolls with QRS arrangements, but don't
operate the registers in those orchestrions correctly.

The only known Berry-Wood A.O.W. is the one in the DeBence museum.
Several A.O.H.R. and A.O.S. orchestrions exist, and it shouldn't be too
hard to find their owners.

This is the first draft of the A.O.W. scale that I worked out by
examining one roll from the DeBence Collection:

1-5 Unknown (one of these is probably used for rewind)
6-63 58 playing notes
64 Percussion (wood block or triangle?)
65 Percussion
66 Percussion (castanets or snare drum rim?)
67 Cancel flute and violin pipes
68 Violin pipes on
69 Flute pipes on
70 Percussion
71 Xylophone on
72 Bells on (71 and 72 might be reversed)
73 Percussion (kettle drum?)
74 Percussion (tambourine?)
75 Mandolin off
76 Mandolin on
77 Cancel xylophone and bells
78 Unknown register on
79 ?
80 Percussion (snare drum rim or castanets?)
81 Sustaining pedal on
82 Cancel sustaining pedal and soft pedal
83 Soft pedal on
84 Snare drum
85 Bass drum (or bass drum and cymbal together?)
87-90 ?

The flute, violin and mandolin registers are obvious, as are the
general cancels for pipes, xylophone and bells, and pedals.  The
percussion tracks are less obvious but it would be relatively easy
to see what instrument sound the best tubed to each hole if a number
of rolls could be auditioned on the A.O.W.  I suggested the above
scale to a technician working for the DeBence museum in 1995 but
I haven't had any contact since then.

Art Reblitz