Okay Craig, I've read both your emails.  Thanks for including me in the
discussions.

The beveled recording rollers, with their ingeniously simple inking
system, sounds like a great idea.  The graphite electrodes dipping into
the mercury, deeper if the key were struck harder, would give a
proportional current, which would make each note's recording magnet
drive its beveled roller harder against the paper, flattening its edge
and giving a wider line.

Just measuring the initial width of a marking would give the needed
info on the key velocity.  The rollers would have to be of soft rubber,
which would get hard as the years passed, but for a one-off factory
system this would be okay.  This master roll might be two or three feet
wide.

As for playback, it would certainly be a good idea to use magnetic
amplifiers, or saturable reactors (same thing).  These used to control
stage lighting, where a small rheostat could handle a bank of 100-watt
bulbs through a transformer-like device.

I think Welte might have been willing to get instant playback through
photocells, each cell reading one note track on the master roll and
driving one vacuum tube whose output then modulated a magnetic
amplifier, one of all these per note.

It sounds like a _lot_ of hardware, but then I once owned a Hammond
Novachord, a 73-note keyboard with two *large* tubes per note, and it
all fit in a small square piano case.  By the way, with a few simple
precautions, arrays of 100s of tubes can be made reliable enough to get
through a day of piano recording and playback without one burning out.

Magnetic amplifiers are a bit slow to respond, however.  So maybe each
note track would have two photocells -- one to read the line width a
little early and set the current to its magnetic amp, and a second one,
delayed, to actually fire the note by closing the power AC circuit to
the hammer solenoid.

To duplicate all this hardware for 88 notes would have been well within
the range of a German industrialist, not known for thinking small, and
a lot less elaborate than some physics laboratory setups I've read
about, not to mention the first electronic computers.  Again, these are
one simple circuit multiplied many times.

The playback piano would have used large, AC solenoids, rather bulky
and noisy, but perhaps in the form of a Vorsetzer.

So yes, I'll retract my statement.  If Welte had wanted to, badly
enough -- where "badly" translates to money and space and a constant
maintenance hassle -- he could have built a recording piano system with
nearly instant playback, complete with expression.  And that's
individual expression for each note!

The one thing missing is any easy way to edit recordings: to fix wrong
notes or alter expression levels, although a patient "artist" with
white-out and ink could doctor the recording in a few places.  Or the
pianist could re-record the bad portion, and Welte could splice the
replacement paper into the master roll.  Not instantly, but within a
few minutes.

Craig, I'll copy this to the MMD as well.

Mike Knudsen