bradr@bartok.Eng.Sun.COM (Brad Rubenstein) (06/01/90)
Music-Research Digest Sun, 27 May 90 Volume 5 : Issue 50
Today's Topics:
Fruitful research areas
Fruitful research areas - summary (3 msgs)
Marsden, Cognitive Musicologists, the Power of the Net and other topics (2 msgs)
*** 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
*** Back issues, index, etc.: send "help" in a message to archive-server
*** @uk.ac.oxford.prg (in the UK) or @bartok.sun.com (elsewhere)
----------------------------------------------------------------------
Date: 23 May 90 20:50:54 GMT
From: Mark Gresham <mgresham%artsnet@edu.gatech>
Subject: Fruitful research areas
To: music-research@prg
Message-ID: <839@artsnet.UUCP>
In article <3390@psivax.UUCP> torkil@psivax.UUCP (Torkil Hammer) writes:
>In article <1990May8.065258.2249@metro.ucc.su.OZ.AU> monro_g@maths.su.oz.au () writes:
>#
># (2) What are some fruitful topics for graduate students?
>#
>Try "Counterpoints in J.S.Bach's music: Are they recognizable by AI?"
>
Or try "I-Ching related distribution patterns in
John Cage's music: Are they recognizable by AI?"
If nothing else, you should get an uncontrollable, emotion-laden
argument going amongst your review-panel. :-)
Oh, yes, and everyone knows the most important part of a graduate
thesis or any other academic paper:
A two-part title with a colon in the middle. :-)
(Not a flame to T.H., just an observation of academia.)
Cheers,
--Mark
========================================
Mark Gresham ARTSNET Norcross, GA, USA
E-mail: ...gatech!artsnet!mgresham
or: artsnet!mgresham@gatech.edu
========================================
------------------------------
Date: 25 May 90 09:00:38 GMT
From: news%metro%munnari.oz.au%samsung@com.think
Subject: Fruitful research areas - summary
To: music-research@prg
Message-ID: <1990May25.090038.20344@metro.ucc.su.OZ.AU>
I asked a little while ago for references to good research in the area of
comp.music, and also for fruitful topics for graduate students.
All the replies to the net were collected neatly in Music-Research Digest
Vol. 5 #46, so I won't repeat them here. Stephen Page also sent me a copy
of his reply. Additionally, Dean.Rubine@CS.CMU.EDU sent me a long and
helpful reply, much of which is reproduced below. There was no response
from one or two people who have told us at length about bad research and
fruitless research areas.
Thanks to everyone who responded.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>From Dean Rubine:
>I've worked in the following subfields, all of which, IMHO, have some good
>research going on (I'm basically a technical person):
>
> A. Analysis/Synthesis of instrument tones
> B. Digital Signal Processing
> C. New Instrument Interfaces
> D. Real Time MIDI performance interfaces
> D. Languages for Computer Music/Real Time Control
>
>As for references, I'll give a few, but there are lots more.
>
> CMJ=Computer Music Journal
> ICMC=Proceedings of the International Computer Music Conference
> JAES=Journal of the Audio Engineering Society
>
>Analysis/Synthesis of instrument tones:
> Moorer, CMJ 1(1)
> Chowning, JAES 21(#?) July/Aug 73
> J. Smith & X. Serra, ICMC 87
> Karplus & Strong CMJ??
> X. Rodet (CHANT program, and FOF) CMJ ?(?)
> M. Serra JAES 38(3)
>
>Digital Signal Processing
> Moorer CMJ 1(1)
> Moorer "About this Reverberation Business", Foundations of Computer
> Music, Roads & Strawn, Eds
> J. Smith ICMC 85 "...Waveguides..."
>
>New Instrument Interfaces
> All of CMJ 14(1), for example
>
>Real Time MIDI performance interfaces
> Dannenberg, "...computer accompaniment..." ICMC87, ICMC85(??)
> (maybe "Bloch and Dannenberg", don't have it handy)
> X. Chabot ??
>
>Languages for Computer Music/Real Time Control
> Mathews (The Music "N" languages) (No reference handy)
> Dannenberg et al, "Arctic...", CMJ 10(4)
> Dannenberg "Canon", CMJ???, "Fugue" ICMC89
>
[Some fruitful topics for graduate students]
>In the technical fields, computer music grad students come in two flavors,
>either electrical engineers, or computer scientists. EEs like to build low
>level hardware and/or do software signal-processing or synthesis. CS types
>like to do interactive MIDI stuff, algorithmic composition, and languages and
>operating systems. As for specific topics, pick a nice big open problem, and
>go after the pieces. As an example, real-time human/computer improvisation
>involves
>
> 1. Tracking of human instrumental input
> use a MIDI device or monophonic pitch detector
> RESEARCH AREA: polyphonic pitch tracking to MIDI
> 2. Extraction of beat information from input
> RESEARCH AREA: A computer "foot tapper" which given, e.g.
> MIDI input, determines where the downbeats are
> 3. Real-time harmonic analyis
> RESEARCH AREA: Given MIDI input (and maybe some style
> assumptions), produce chord charts as output
> 4. Real-time composition
> RESEARCH AREA: Given a "lead sheet" produce an accompaniment
> in real-time
>
> Other big problems which can be similarly broken up are computerized aids
>to transcription (e.g. input: recording; output: sheet music), score editing
>(sub-issues of music representation, user interfaces, ...) and, with all the
>fast hardware that's appearing, doing some of the new synthesis algorithms
>(e.g. physical models) in real-time with real-time human gestural control is
>also becoming feasible. I could easily go on, but you can get just as many
>ideas by reading the tables of contents to recent (or not-so-recent) CMJs and
>ICMC proceedings.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Gordon Monro
University of Sydney. Internet: monro_g@maths.su.oz.au
------------------------------
Date: 26 May 90 05:00:42 GMT
From: Eliot Handelman <eliot%phoenix@edu.princeton>
Subject: Fruitful research areas - summary
To: music-research@prg
Message-ID: <16758@phoenix.Princeton.EDU>
In article <1990May25.090038.20344@metro.ucc.su.OZ.AU> monro_g@maths.su.oz.au () writes:
;I asked a little while ago for references to good research in the area of
;comp.music, and also for fruitful topics for graduate students.
; There was no response
;from one or two people who have told us at length about bad research and
;fruitless research areas.
I assume you mean me.
A Fruitful Research Area: There is today growing concern with the need
for a new information-processing metaphor, and major corporations
are expressing interest in science-fiction author W. Gibson's concept
of "cyberspace," in which information is given form as a "consensual
hallucination" mapped out directly in the operator's mind via direct
brain stimulation. The metaphor to date has been strictly visual.
Part I of your PhD project is to describe and implement (a simulation
will do) the auditory analogue of this metaphor. Justify the metaphor
with references to L. Beethoven, R. Wagner, J. Baudrillard, E. Bloch,
H.R. Jauss, U. Eco, J. Kristeva and I. Xenakis. For part II of your
PhD project you must write a concerto for midi piano and large orchestra
with concertante piano and rock percussion. You must then find an
orchestra, copy all of the parts and conduct it yourself.
------------------------------
Date: 26 May 90 15:23:50 GMT
From: Stephen Smoliar <smoliar%venera.isi.edu%usc@edu.ucsd>
Subject: Fruitful research areas - summary
To: music-research@prg
Message-ID: <13648@venera.isi.edu>
In article <1990May25.090038.20344@metro.ucc.su.OZ.AU> monro_g@maths.su.oz.au
() writes:
> There was no response
>from one or two people who have told us at length about bad research and
>fruitless research areas.
>
MEA CULPA, Gordon! However, there is a reason for my silence, which is that I
believe that graduate students should only venture into this area with proper
guidance and supervision. I once observed a very sad experience in which a
graduate student had one advisor in computer science and another in music.
He was able to keep each very happy as long as they did not talk to each other!
I was the unfortunate agent who closed the loop of communication, and it became
quickly apparent that this student really did not have the musical competence
for the project he had proposed. My feeling is that until we have some faculty
who are qualified to supervise such research (that is, qualified as individuals
to take responsibility for ALL aspects of the research, whether they involve
computers, music, psychology, or even brain science) it is a BIG MISTAKE to
encourage graduate students to go looking for topics in the field. One of the
reasons I report at great length about "bad research and fruitless research
areas" is in the hope that others will not make similar mistakes.
=========================================================================
USPS: Stephen Smoliar
USC Information Sciences Institute
4676 Admiralty Way Suite 1001
Marina del Rey, California 90292-6695
Internet: smoliar@vaxa.isi.edu
"By long custom, social discourse in Cambridge is intended to impart and only
rarely to obtain information. People talk; it is not expected that anyone
will listen. A respectful show of attention is all that is required until
the listener takes over in his or her turn. No one has ever been known to
repeat what he or she has heard at a party or other social gathering."
John Kenneth Galbraith
A TENURED PROFESSOR
------------------------------
Date: 22 May 90 05:21:05 GMT
From: Eliot Handelman <eliot%phoenix@edu.princeton>
Subject: Marsden, Cognitive Musicologists, the Power of the Net and other topics
To: music-research@prg
Message-ID: <16623@phoenix.Princeton.EDU>
This is in response to the three or four recent metaposts about how to
moderate the research digest.
First of all, I'm sometimes a big believer in the power of the net: and
some of the discussion here easily beats out much discussion to be had
in other, more formal, circumstances. But the net is better, because
one can write when one pleases, read when one pleases, consult a reference
if one pleases, etc. There is an old jewish legend about paradise, according
to which it's seen as a large, round table, and on this table are all of
the books that you'll ever need, and sitting around it are all of the
sages and scholars you'll ever want to talk to: and I like this idea,
because it presents paradise as a place where the intellect can proceed
untrammeled by the inconveniences of ordinary life, and there's no reference
in this legend to enlightenment, or perfect knowledge. One does there
the thing one likes doing best, which is to exercise the mind, sometimes
drawing brain-blood. The net could be such a place.
But we all have misgivings about the net. There's neither money nor
credentials in it: you can't even state something like "Marsden
stated (ref: Internet 86356256) that he was often irritated
with Handelman's postings, and Handelman countered that he thought
Marsden was a dullard and a booby (ref: Internet 698789)" in a respectable
document, that is, one suited for publication. You don't put the
net down in your CV. And so we now have a contradiction in the
academic marketplace: conferences count, published papers count,
but the net doesn't count, and yet the information is the same.
This may seem like an absurd contention. But about 99% of all the
material I've read on Music/AI/Cognition has been utterly vapid, so
it's not really all that hard. Music/AI/Cognition isn't yet a
field. It's actually a name in search of a project. Far better to
air open discussion on the nature of this project than
to listen to sordid implementation details of projects that haven't
yet been implemented, and are not going to be implemented. (ref:
Laske's latest paper on Blackboard Architectures, which I'll review if
there's interest.)
The net has the power to cast conferences and closed-door seminars on
unclassified matters (such as musical semantics) into immediate
obsolescence. And I think that one of the reasons that Laske formally
withdrew, and that Bel, for all we know unsubscribed, and that Marsden
chose to post something that was utterly contentless, is that these
people, whose livelihood might partly depend on the conferences
they throw, have already come to that realization. Conferences are sort
of like blind dates, perhaps more like singles parties. The more you
talk with the participants the less of a reason there is to go to these
functions. But a stronger reason is that it's bad for someone like Laske,
who does a hell of a lot of abstract reviewing, to lose credibility
by constantly losing arguments to everyone who chooses to challenge a
view or two of his. Similarly, Bel had the opportunity to enlarge
on some comments he made, which I questioned, or supplemented, and we
might have had the kind of discussion which, no doubt, he hopes to
have at HIS conference. But Bel avoided any music-theoretical
discussion, preferring instead to assault me on a completed unrelated
issue due his faulty understanding of american english. Reason:
the "real" Bel is available in the flesh, not by public access net.
You must physically go to get the "real" discussion. But THIS is
the real discussion.
I don't want the net to become a conference place. I want it to stay
what it is, namely an anarchistic firing ground. Only I want more
professional participation. And I'll continue to flaunt whichever gizmos
I judge expedient to that end, as long as I have the time to waste.
------------------------------
Date: 24 May 90 02:53:24 GMT
From: Stephen Smoliar <smoliar%venera.isi.edu%usc@edu.ucsd>
Subject: Marsden, Cognitive Musicologists, the Power of the Net and other topics
To: music-research@prg
Message-ID: <13606@venera.isi.edu>
In article <16623@phoenix.Princeton.EDU> eliot@phoenix.Princeton.EDU (Eliot
Handelman) writes:
>
> And so we now have a contradiction in the
>academic marketplace: conferences count, published papers count,
>but the net doesn't count, and yet the information is the same.
The contradiction lies in the possibility that the net may be confronting the
academic world with potential obsolescence. The "academic marketplace," as you
call it, has lived quite comfortably for quite some time with rules of its own
making. If you wanted to be part of that world, you learned to play by the
rules. If you played by those rules, you would eventually find a comfortable
niche from which you could make sure that future generations would play by the
same rules. Your role in life was not so much one of venturing into unknown
regions as it was one of defending against anarchy.
Now we have this new computer-based medium of communication which is about as
close to anarchy as you can get. Much to the horror of many traditional
academics, it has turned out to be a rather effective channel for the exchange
of ideas. As a matter of fact, in many ways it looks back fondly on an earlier
age in which argument lay at the root of all education. The student was
encouraged to talk back, rather than nod his head obediently. (If he wanted
to learn obedience, he could become a priest!)
What makes matters worse is the desperation with which the academic world
clings to its standards. Publication continues to be the criterion which
will make or break you. Unfortunately, publication takes a lot of time these
days . . . primarily because the volume has gotten out of hand. Usually, by
the time what you have to say appears in print, you aren't thinking about it
any more. You may even have discredited it! The net provides a much more
immediate environment for thesis and antithesis. It is also an environment
where those who want to pay attention do so and those who don't can go away
without any sense of guilt or concern that they are not getting what they
paid for.
>This may seem like an absurd contention. But about 99% of all the
>material I've read on Music/AI/Cognition has been utterly vapid, so
>it's not really all that hard. Music/AI/Cognition isn't yet a
>field. It's actually a name in search of a project. Far better to
>air open discussion on the nature of this project than
>to listen to sordid implementation details of projects that haven't
>yet been implemented, and are not going to be implemented. (ref:
>Laske's latest paper on Blackboard Architectures, which I'll review if
>there's interest.)
>
There is definitely interest on my part. I have also voiced my agreement with
Eliot on the general state of the art of Music/AI/Cognition in a variety of
forums. Perhaps we are witnessing a dangerous artifact of the current operation
of Eliot's "academic marketplace." I wrote, above, about finding your niche.
I guess if you can't find one, the next best thing is to make one. It had not
occurred to me that the institutions of conferences and publications had now
become some flexible that making your own niche may take little more than lots
of persistence and noise. One thing is certain: concrete results no longer
seem to be a major factor in establishing such a niche.
>
>The net has the power to cast conferences and closed-door seminars on
>unclassified matters (such as musical semantics) into immediate
>obsolescence. And I think that one of the reasons that Laske formally
>withdrew, and that Bel, for all we know unsubscribed, and that Marsden
>chose to post something that was utterly contentless, is that these
>people, whose livelihood might partly depend on the conferences
>they throw, have already come to that realization.
More likely, the reason is simply an unwillingness to play by a new set of
rules. Once you have found your niche, you are going to be VERY reluctant
to give it up. Eliot's "academic marketplace" is also rather good about taking
care of itself. Unlike the world of natural selection, weak members are not
selected out. Rather, the number of niches can only go up. Young Turks are
always entitled to their own opinion, but they are NEVER entitled to question
an already-established niche.
> Conferences are sort
>of like blind dates, perhaps more like singles parties. The more you
>talk with the participants the less of a reason there is to go to these
>functions.
They are also like the sorts of social events documented by Galbraith in my
.signature (which has made at least one comp.ai reader VERY uncomfortable)!
>
>I don't want the net to become a conference place. I want it to stay
>what it is, namely an anarchistic firing ground. Only I want more
>professional participation. And I'll continue to flaunt whichever gizmos
>I judge expedient to that end, as long as I have the time to waste.
I think the net is the best place to be for anyone who genuinely wishes to
pursue new questions and does not feel bound by the rules of a decaying
academic society.
=========================================================================
USPS: Stephen Smoliar
USC Information Sciences Institute
4676 Admiralty Way Suite 1001
Marina del Rey, California 90292-6695
Internet: smoliar@vaxa.isi.edu
"By long custom, social discourse in Cambridge is intended to impart and only
rarely to obtain information. People talk; it is not expected that anyone
will listen. A respectful show of attention is all that is required until
the listener takes over in his or her turn. No one has ever been known to
repeat what he or she has heard at a party or other social gathering."
John Kenneth Galbraith
A TENURED PROFESSOR
------------------------------
End of Music-Research Digest
--
---Brad Rubenstein-----Sun Microsystems Inc.-----bradr@sun.com---