[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[grant@master.symcan.dialnet.symbolics.com: Music on the Symbolics Machines]

    Date: Fri, 21 Apr 89 16:33 EDT
    From: grant <grant@master.symcan.dialnet.symbolics.com>
    Subject: Music on the Symbolics Machines
    To: slug%warbucks.ai.sri.com@VERMITHRAX.SCH.SYMBOLICS.COM

    We would like to find out if anyone out there has had any experience 
    with the Symbolics system entering, displaying and playing musical 
    scores.  The prospect would like to enter the musical notes via 
    some sort of "standard" interface (perhaps a MIDI file, if one 
    could be created, assumin there is such a thing as a file form of MIDI), 
    have the Symbolics display the full SCORE (as it might originally appear 
    on sheet music), and play the tune back, but highlighting the notes as 
    they are played, on the color screen, in order to later lay the frames 
    out on to video.

    If anyone has any suggestions, please forward information to 

    Grant Chenier
    Eclin System corp.
    Date: Fri, 28 Apr 89 14:08 EDT
    From: Len Moskowitz <Len@HEART-OF-GOLD>
    Subject: Re: Music on the Symbolics Machines
	    A few years back I attended a Symbolics demo at a conference
    and remember seeing a machine playing Bach fugues.  Anyone know what
    became of that program?
    Len Moskowitz
						       Allied-Signal Aerospace
    moskowitz@bendix.com (CSnet)                       Test Systems Division
    moskowitz%bendix.com@relay.cs.net (ARPAnet)        Mail code 4/8
    arpa!relay.cs.net!bendix.com!moskowitz (uucp)      Teterboro, NJ 07608

Sorry about the slow response, these got to me somewhat indirectly.

In early 1984, Bachmeister Bernie Greenberg (BSG) started cobbling up a
music-playing system.  Audio facilities were just being added to the
3600 at the time.  Beethoven fan Dave Plummer (DCP) helped quite a bit
with the low-level audio support.  I soon got involved in adding to the
system and the library of music (which was 90% J. S. Bach).

Pretty soon, this facility did a number of things:

(1) Played pipe-organ-like music with many stops and registrations

(2) Compiled music expressed in a Lisp-like notation (via a DEFMUSIC

(3) Played back compiled music

(4) Displayed the score of compiled music

(5) Similarly printed the score on the LGP1

(6) Composed and played concise, correct, and fairly boring two-voice

Included were a bunch of interface kludges, of course.  But the system
(Musicsys 3600) was far from product quality.

BSG took it on the road several times in support of Sales.  At the
Hannover Fair in late spring of 1984, the Symbolics Gmbh. booth even had
demonstrations presided over by an actor dressed in 18th-century clothes
and powdered wig, and with BSG struggling to hold forth in German.

This is the system that Moskowitz recalls seeing.  BSG was also adamant
about not letting the system be released; he had ideas about making it a
product later.  A music publishing system was also discussed.  So it
never was given out to customers, which is why you haven't seen it

The system still mostly works, having been minimally maintained through
the transitions to Dynamic Windows, SCT, et al.  It didn't make the
transition to the LGP2, so the printing ability has rotted.  So have
other small things.  The ability to play music is completely lost, at
this point, on Ivory machines.

Apropos of Chenier's questions, MIDI interfaces were discussed, and a
little work was done, but nothing that really worked ever came of it.
We never had anything as coordinated as a full animation of the
displayed score in sync with the music, although the idea was mentioned.

Data entry of music into some internally useful form always was a weak
point.  Originally, all the music we compiled had to be entered
painstakingly by somebody like me, transcribing a score to the Lispy
notation in a Zmacs buffer, much in the manner of monks hunched over
illuminated manuscripts.  

We speeded this up quite a bit by writing an interface that allowed
input to a pseudo-piano-keyboard mapped onto the console keyboard, which
played each note as struck, and displayed a note on a staff.  It could
be edited with RUBOUT; SPACE defined a rest.  When a (single voice)
phrase was complete, an appropariate Zwei command captured it and
translated it into the internal language in the current Zmacs buffer.
Note, however, that this only captured the pitches of the notes.  The
time values and other attributes had to be edited in by hand, although
that was the easier part of the job.

Again, a MIDI interface to a real piano-like keyboard was discussed, at
least for entering pitches, but it never got off the ground.

As you can see, people have been working on this problem before.  I'll
be happy to comment further (time permitting) on any of our experiences
with particular problems that I can remember.