Rob Goodale seems to think that when I suggest there are irreconcilable
delays in playing acoustic (real) pianos at a distance using a MIDI
signal I am talking about transmission time, or processing time.  I'm
not.

Let's abandon MIDI for a moment.  A pianist plays a note at A as softly
as possible.  This requires the action of pressing the key down to take
maybe half a second.  The hammer brushes the strings; hey presto ! we
have a minimum-strength note.

For Piano 2 500 miles away to produce this same sound at the same
time, it must have been instructed not only when the key started to
be pressed down, in real time, it must "know" how fast the key is
descending if its hammer is to touch the strings at the exact same time
as at A (assuming virtually instantaneous transmission).

All this would be very difficult to do using any system of key
detection.  With MIDI, no information is transmitted until the end of
the movement producing the note at A.  An acoustic piano at B then has
to spend another half second producing the note ! Synchronisation is
impossible.

However, I believe the way out of this is to drive both pianos using
MIDI, so that the pianist has to get used to soft notes coming out half
a second after he/she's played them, but loud notes almost immediately.
I suppose since most music transmitted by MIDI is pretty raucous,
nobody's stopped to think about it !

Dan Wilson