Robbie Rhodes mentioned some of the "given" philosophies in music roll
editing, and it was excellent. However, I would like to add an
abridgement to these lest we accept this as standard protocol. So in
regard to a few of his entries, perhaps he could clarify,
concerning(split stack) reproducers.

Item:

>  c. durations of all notes in a chord are usually made equal, if only
>     for better visual appearance.

More often than not, chords in the best rolls are not equally aligned
so that certain notes in those chords are able to be played either
softer or louder than the others. By staggering the time a note in a
chord is played, the leading note is given a higher velocity, and
immediately, the other(s) are dropped, accenting that particular note
of a chord.

>  d. "all music is quantized to perfect eighth notes or triplets, or
>     whatever--"

I don't think that is necessarily right, although Ampico corrected the
rhythm automatically on popular dance numbers by swinging the
trackerbar reader.  Still, interpretive works, even of a popular vein,
like Lee Sims, doesn't sound very mechanical or quantized, to me. I
believe it would be a mistake to start setting these things down as
gospel when the jury is still out. The only quantization required is
the resolution of the punch. Most perforators perforated at about 20
steps/in. Ampico perforated at 30 steps/in. Richard Tonnesen perforates
at 45 steps/in. But some tunes, due to the triplet restrictions,
required that the paper move faster than tempo 70-80. Those were cut at
at least 90, and I have seen some rolls cut as high as 120, which would
record somebody like Art Tatum!

    If readers were sync'd at some multiple of the original punch
stepping rate, the resolution would be improved, over a continual feed
or a rate that is close to the old punch rate. 10 to 1 has always been
accepted as good resolution for all kinds of readers, so maybe it would
work well for this, too. That isn't too hard to get, with a stepping
motor and a blanking circuit that clocks the computer off and on at the
same rate, I should think.

>  e. hand-played recordings must be adjusted to metronomic precision..."

Here again, this is only partly true, and it isn't necessary, unless
you want the tune to sound perfect. Granted, 20's jazz sounds just
great when set exactly to the measure lines, but an interpretive score,
like Gershwin, would not.

    I don't know how to program, fellows. I'm just speaking from
experience, now.  I have yet to buy my first MIDI controller, and when
I do, it will be for my "O"-Roll Orchestrion. But We'll get around to
that shortly, I hope. Meanwhile, I don't want to see the great NEW
arrangements which are just around the corner, trimmed by demands made
on them because of slow and limited perforating techniques. I think
this way: If you want a "clinker," you should be able to get a
clinker!

Craig B.

craig_brougher@msn.com