I felt I had to comment on Robert Buckingham's recent suggestion
about building a roll copying machine [00025 MMD]:
> It seems to me that the best way to do it (without involving
> computers) would be to employ an electro-pneumatic set up...".
This does initially seem like a simple and reliable scheme, with
the contacts directly operating solenoids that eventually cause a
corresponding hole to be punched in a copy. I am aware of several
attempts at copying rolls using this scheme.
Unfortunately, this method gives unsatisfactory results. The holes
in the copy are substantially elongated with respect to the hole in the
original. You need a way to shorten the "on" signal so that one punch
in the original gives one punch in the output.
That is where the computer comes in. When the contact closures are used
as input to a computer, the computer can monitor all note durations and
truncate the "on" duration to the proper length. The modified data
stream is then used to control the perforator.
The direct method is unsatisfactory in another way, in that a pneumatic
reader detects a lot of pinholes, tears, and buckled paper as real
holes, which are duly copied on the output roll. You really need to
separate the reading operation from the punching operation, and provide
a way to view and edit the roll data before it is used to create a
Computers to the rescue again! By making the reading program monitor
the contacts and create an intermediate disk file of the roll data,
you can get the required separation of reading and punching. Separate
programs can then read the roll data file and operate on it. For
example, one program can display and edit the data and create a new
version of the roll file with corrections. Finally, another program
can read the corrected file and control the punch to create accurate
[ When he established Play-Rite Music Rolls, John Malone quickly
[ installed a specially-built computer to truncate the signals
[ and regenerate chain patterns. As for reader errors (pinholes,
[ etc.), mother Jeanne Malone patches any errors in the recuts.
[ Read the description of Richard's perforator at the MMD Pictures
[ site, http://mmd.foxtail.com/Pictures/ -- Robbie