Many thanks to Jody & Mike for their most helpful comments on interfacing
Concertola and MIDI.

As a newcomer to the group, I can't help but notice that similar questions
to mine seem to appear and recycle in a _lot_ of past digests.  When I
started on this thread I was unaware of much of this material and feel a bit
embarrassed at bringing up what must seem to many to be old hat.  I can't
help but think  that there would be a good market for a book on the subject
of interfacing MIDI and mechanical instruments.  Does anybody know of such a
book and  if not, whom we should  pester to write one?

On the subject of contact bounce I knew it was present, but had not made the
obvious connection to MIDI's compulsion to record every event.  I expect that
there are ways to clean the signal up, but no really simple (read
inexpensive) ways come to mind.  I'm not sure that simple integrating
resistors and capacitors would do the trick without sacrificing response time.

On the subject of the piano pallet magnets, they are of fairly high
resistance (therefore low-Q) so that diodes function well on a stand-alone
basis without need for a series resistor.  The diodes help _considerably_ in
both extending the time between contact cleanings/unstickings and
eliminating interference from taped performances.

Just for the sake of interest, although I know they are more common than the
piano version, I have never seen an organ Concertola.  Mike Ames describes
the contacts as "limber contact switch(es) on the end of which are the
typical organ crossed silver chucks of wire."  Not being an organ person I
am not sure of his terminology and if he is describing exactly what is in my
setup, or if the arrangement is something different.  I have seen organ
contacts in which a "single" contact consists of a parallel array of silver
wires alternately connected to one side of the circuit and to the other.
Activation of the contact consists of applying a shorting bar across this
brush, completing the circuit.  In the contact chest of my concertola, under
the celluloid window and in vacuum, the individual pouches directly push a
_single_ spring wire against the common bus to complete the circuit.  The
chest itself is quite compact being only 21-1/2" long by 2-1/4" square and
contains 92 vertically-mounted (by this I mean they operate with a
back-and-forth as opposed to up-and-down motion) pouches and contacts in
four horizontal rows of 23.

The Devtronix  equipment sounds very much like a good place to start,
although I shall have to give a lot of thought about the recording side of
the process!

Idle questions dept:

Would Hall-effect switches eliminate the problem of contact bounce ?

How do I contact Devtronix ?

Has anyone developed a standard for mapping notes and expression/function
coding to MIDI ?

On the subject of converting rolls to MIDI playable on standard synthetic
pianos and modern instruments, has anyone considered attaching pressure
transducers to the stack of a well-restored instrument and sampling at (or
preferably just before, if you have a crystal ball) the moment a note is
sensed as a means of obtaining the appropriate MIDI velocity ?  It would
have to be done at _exactly_ the right time, because once the valve opens
and the pneumatic starts to close things won't be the same as when it all
started !

So much for the screed of the day.

Cheers!     Jim