daemon@bartok.Eng.Sun.COM (12/29/89)
Music-Research Digest Thu, 28 Dec 89 Volume 4 : Issue 79
Today's Topics:
Copyright question
DMCS vs. Concertware for Mac
PERSONAL COMPOSER
Regarding Connectionist Composition
technical MIDI text recommendation
Xenakis article
*** Send contributions to Music-Research@uk.ac.oxford.prg
*** Send administrative requests to Music-Research-Request
*** Overseas users should reverse UK addresses and give gateway if necessary
*** e.g. Music-Research@prg.oxford.ac.uk
*** or Music-Research%prg.oxford.ac.uk@nsfnet-relay.ac.uk
----------------------------------------------------------------------
Date: 23 Dec 89 19:36:50 GMT
From: Steven Grimm <koreth%panarthea.ebay.sun.com%grapevine%newstop%sun-barr@gov.nasa.arc.ames>
Subject: Copyright question
To: music-research@uk.ac.oxford.prg
What do the copyright laws have to say about transmitting small excerpts from
a piece of music? I am pretty sure it's legal to quote a few paragraphs from
a book/magazine for review or discussion purposes; is it legal to digitize a
few seconds of music, for the same purpose?
The reason I ask is that recently, I've been talking to some people on the
net about a couple of musicians I like a lot. I don't have a musical back-
ground, though, so I don't know the terminology to describe the music. What
I'd like to do is play a CD into a Sparcstation here, uuencode the sound,
and mail it off to the other parties (assuming they have similar equipment
for playing back the sound.) If it's legal to do that, I foresee many such
excerpts showing up on the net as well, since digitized sound compresses
well enough to stick into an article without overloading the net.
---
" !" - Marcel Marceau
Steven Grimm Moderator, comp.{sources,binaries}.atari.st
koreth@ebay.sun.com ...!sun!ebay!koreth
------------------------------
Date: 26 Dec 89 23:13:36 GMT
From: Allyn Weaks <allyn%milton%caesar.cs.montana.edu%uakari.primate.wisc.edu%zaphod.mps.ohio-state.edu@edu.>
Subject: DMCS vs. Concertware for Mac
To: music-research@uk.ac.oxford.prg
(Note that followups are directed to comp.music...)
There have been a couple of requests for low end Mac music programs on a couple
of different groups, and not too many replies anywhere, so I finally stopped
procrastinating. I have both DMCS 2.5, and Concertware 5.0, and I much prefer
Concertware. It's a little more expensive, probably around $120 discount,
though I haven't looked lately. DMCS is about $85? discount. What I'd really
like is CW with a few of the DMCS features (or maybe versi vicey).
The two primary reasons I prefer Concertware are: it's not measure-based (see
below for a few of the implications) and it reads/writes MIDI files, so I can
enter stuff into CW, and if I really want a decent hardcopy score, I can dump
it into Finale (not really as straightforward as that, but that's entirely
Finale's fault.)
Some of the features and trade offs between the two:
Both have:
32nd notes to dotted whole (why does everyone leave out double wholes? :-(
crescendo/decrecendo in score and for playback
Mac voice editor
let you drag notes to change the pitch
chromatic transpose by n half-steps
'lots' of notes per chord in a single voice (all the same duration)
support Sonata laser font output
a few buglets here and there :-)
DMCS 2.5:
better control over score printing - you can tell it how many measures/line
plays back with articulations (stacatto, legato, slur)
can handle up to 40 staffs for printing (plays back 4 (Mac) or 16 (MIDI))
you can select any rectangular section, any number of contiguous voices, and
and make changes (such as articulation) all at once.
diatonic transpose
somewhat better Mac sounds, maybe (opinions vary)
only two voices/per staff, which makes some keyboard parts impossible to enter
also, you can only chose playback by staff, so you have to play both
voices or none.
does NOT read/write MIDI files (they have some screwy 'interchange' format
that no one else on the Mac except (I think) Opcode uses). No excuse for
this - MIDI format was finalized long, long before 2.5 was released...
for input, can only do step time entry (and not a good system for that - you
have to tell it how long a time chunk for a 32nd note, then do everything
proportionally; takes forever to get those whole notes in...)
mediochre lyrics: to line them up, you have to watch little location
numbers change; have to change font, style syllable by syllable
measure based - you can only do many operations (change clef, key signature,
tempo, etc) at the beginning of a measure. Can't do ritards, fermatta,
accelerandos. Also makes some editing much more difficult - it will let
you put too many or few notes in measure, than you have to painfully fix
things one measure at a time 'merging' and 'splitting' measures by hand.
Concertware 5.0
reads/writes MIDI files, both format 0 and format 1
real time MIDI entry, two voices at a time if you assign a keyboard split
flexible quantizations
nice efficient step-time entry (hit a number key to set the duration, then
keyboard keys for the pitches. Duration stays in effect until changed.)
lets you assign MIDI macros
you can choose different note heads (diamonds/crosses) for percussion, etc.
Comes with it's own laser font so you don't have to buy Sanata laser font
chord sumbols (and can define your own)
symbol library
excellent lyric handling: can import/export the lyrics from/to regular
clipboard text, all of the lyrics line up and you can move the line
with the ruler, can change font, style of an entire lyric line at once
one of only 2 programs on the Mac that isn't measure based (the other is
Professional Composer) so mass editing is much improved. You can also
force ritards, accelerandos, etc. by putting in tempo changes anywhere
(impossible in DMCS). When you insert, it inserts into the entire score,
and rebars automatically. Makes it very nice for things that constantly
change meter - you can enter everything at once, then go back and insert
the time sigs; DMSC (and Finale, and most others) you have to enter a
measure, insert time sig, enter a measure, enter time sig, and Murphy
help you if you don't notice a mistake until you're 'finished'...
will beam a selection to a beat, so you don't have to do it by hand (but
doesn't yet make intelligent decisions about which way to put the flags)
only eight voices, but you can put any or all of them on one staff
mediochre printing - no control over spacing
some not-good printing bugs - part extraction with tacit measures screws
up royally, and it occaisionally gets confused under other circumstances.
I've been able to work around all of them though, and I've never lost
anything, so I'm willing to chalk it up to teething problems in the new
(much extended) version. Great Wave, are you out there?
articulations only effect printing, so playback is even more stilted than DMCS
-----
Allyn Weaks
allyn@milton.acs.washington.edu sweaks@phast.phys.washington.edu
{backbone}!uw-beaver!milton!allyn sweaks@phast.bitnet
Writing comes easy. All you have to do is stare at a blank piece of paper
until your forehead bleeds. -- Douglas Adams
------------------------------
Date: 22 Dec 89 19:31:42 GMT
From: Bryan Sutula <sutula%hplvli%hplisa%hpfcso@com.hp.hplabs>
Subject: PERSONAL COMPOSER
To: music-research@uk.ac.oxford.prg
/ hplvli:comp.music / operator@rivm.UUCP (operator van dienst) / 9:27 am Dec 20, 1989 /
>
> Is there anybody out there who can give me the exact address of Jim Miller ?
>
> ----------
Closest I can come is:
Personal Composer
2448 76th Ave SE
Mercer Island, WA 98040
BS
------------------------------
Date: Thu, 28 Dec 89 13:16:43 EST
From: laske@edu.bu.cs
Subject: Regarding Connectionist Composition
To: music-research@com.sun.eng.bartok
Otto Laske
c/o Newcomp
926 Greendale Avenue
Needham, MA 02192
Letter to CMJ, with copy to Peter M. Todd
December 89
Below are some thoughts on Peter M. Todd's paper
entitled "A Connectionist Approach to Algorithmic Composition"
in CMJ 13.4:27-43.
First of all, thanks for printing the article. It shows very clearly
that scientific enterprises, at least in music, always have an implicit
esthetics associated with them; connectionism is no exception.
While I thought "the computer" (von Neumann style) had finally freed
us composers from existing musics, making it possible to concentrate
on possible musics never head before, the connectionist model of
composition--melodies or polyphony, or whatever--happily returns us
to model-based composition. By that term, I mean composition based
on remembered musics of the past, in whatever form. This is in contrast
to rule-based composition [Charles Ames would add constraint-based
composition which I count under rule-based], where no prior model
is used, except a model of a process perhaps, but certainly not a
model of pre-composed music. I also thought that Xenakis had freed
us from being primarily concerned with surface structures--music
as it develops in time--and had learned [re-learned] developing
deep structures which, essentially, could be played forward as well
as backwards, which for me is a criterion for a well-composed piece
of music. But no, since in connectionist models time is felt to be
of the essence, we are happily returned to surface structures--
themes and variations! So, the connectionist model cannot distinguish
between deep and surface structure, except if one wants to consider
the underlying model [the "existing" music] as a deep structure,
which it is only in a relative sense. The fact that one can produce
derivatives of musical models that go beyond the model perceptually
[i.e., are not recognized as variants of a model], is no proof to
the contrary, only a proof of the perceptual limitations of listeners,
but not a proof of the audacity or originality of the model-based
composer.
So, connectionist models of composition seem to come attached
with an esthetics that is rather suited to pedagogy and musicology
in the orthodox sense than being a sign of progress in compositional
thinking and composition theory. While it is true that model-based
composition abounds even in "computer music", nobody will convince
me that composing by interpolation and extrapolation is more than
a rather primitive notion of composition, showing a lack of notions
of composition theory.
Clearly, I prefer the freedom of composing with rules regarding
deep structure to using models of bygone ages. However, I am open
to demonstrations that show that one can do inventive composition
using networks without being bound by models of existing music.
Otto Laske
------------------------------
Date: 23 Dec 89 16:47:45 GMT
From: Paul Lansky <paul%phoenix@edu.princeton>
Subject: technical MIDI text recommendation
To: music-research@uk.ac.oxford.prg
I am teaching a computer music course next term in which I will
have almost entirely very highly computer-literate people, (many
computer science and engineering students). I would like to find
a good technical MIDI text. Most of the books I've looked at
would be a waste of time for this crew. They'll be writing
their own MIDI programs using Carnegie Mellon Toolkit and NeXT
musickit. I need something short, sweet and to the technical
point. Thanks
Paul Lansky
Music Department
Princeton University
Princeton N.J.
------------------------------
Date: 21 Dec 89 20:16:57 GMT
From: Gene De Lisa <gene%ntvax%iex%convex%texsun%newstop%sun-barr@gov.nasa.arc.ames>
Subject: Xenakis article
To: music-research@uk.ac.oxford.prg
In article <37406@apple.Apple.COM> nsj@Apple.COM (Neal Johnson) writes:
>Try to find an article by Xenakis on computer programs for
>generating stochastic based music. This is rather old because
>the program examples are in FORTRAN 2! I have a copy of it
>if you have a hard time finding it. I think there is a book
>by Xenakis which has the article.
Xenakis, Formalized Music (1969 IU Press translation from the french original)
--
Gene De Lisa
gene@dept.csci.unt.edu
You really don't need a clever quote here.
------------------------------
End of Music-Research Digest