music and diagrams

June 16, 2003 4:22 PM

Composing with GeoMaestro To play with time, first get rid of time.
posted by andrew cooke (26 comments total)

very cool, my man!
posted by mcsweetie at 4:36 PM on June 16, 2003

oops, should have added: via the squeak mailing list (stephane rolladin is doing a squeak port). he also mentions this annotated composition
posted by andrew cooke at 4:41 PM on June 16, 2003

Looks like fun. Anyone know anything similar for OS X?
posted by tss at 6:13 PM on June 16, 2003

I just sit in front of my electronic keyboard and plunk. This stuff gave me a brain cramp just looking at it.

Ten to one the guy's a buddhist or something-that world view would lead to this musical philosophy. Someone smarter than I am will have to tell me if it makes sense or not.
posted by konolia at 7:01 PM on June 16, 2003

Absurd.

One still has to listen to the resulting output in a linear time path. Pretending to "do away with time" is just silly.

All it is really is an unmetered sequencing of the data in a way that is not simply linear. So it does it in circles or starbursts or spirals. Whoop-ee. Using a different symbolic representation does little to change the art form.

The fact that it is still represented, as their own examples prove, by a standard linear midi transcription shows this is nothing special.

Now, what I would consider special would be if this worked in reverse. Take an existing piece, sequenced in traditional linear format, and take it backwards to create a symbolic data set for it. That might be mildly interesting.

But this is basically a souped up, fancy-geometry player piano spindle.
posted by Ynoxas at 7:37 PM on June 16, 2003

tss: Opcode used to make a program called MAX -- now apparently distributed by Cycling '74 -- which worked on the same general principle: instead of composing music directly, you worked by plugging together little programmatic engines that responded to each other or to external events. Very cool to watch, very fun to play with, and almost utterly useless for producing anything even faintly resembling music. Cycling '74 seems to have an OSX version. I'd be curious to hear what you think of it, if you try it out; it's been years and years since I played with it.
posted by ook at 7:46 PM on June 16, 2003

I haven't downloaded it, but from the description, yes it does seem that it is wrong to say that "time isn't represented;" however, I would say that using a different symbolic representation of time does, in fact change the process of composition. It's my impression (as someone who's written a few tunes) that music generally happens through experimentation; when the gameboard changes, so do the ways you play with it.
posted by transient at 8:09 PM on June 16, 2003

I can't make heads or tails of this...I'm a cyber-rube...but I still want to weigh in on this.

Enchanted with Cage, I wrote a randomized piece in the early seventies. I checked a book out of the library with lists of random numbers and proceeded from there. (There were a couple of authorial choices embedded in the compositional process, but it was essentially random.)

You have to wonder, though, about what compositions like these have to do with music as it has been historically.

This evening I put in three hours at the piano in a nearby restaurant. What it was really about was connecting in a number of different ways (this would take a book) with my fellow humans.

I'm just re-posing the question: what is music?

I think that what happens on the back porch embodies what music is really about...so....could this kind of music be in a different - hyperacademic -realm entirely?
posted by kozad at 9:14 PM on June 16, 2003

I agree, kozad and Ynoxas.

What's described here may well be art, but it's not music. The author assumes that because high art music comprises highly abstract sequences represented in an arcane notation with a regular aural realisation, that the reverse is true.

If I understand correctly, the author has got the whole score->performance process arse-about-face, and is viewing the notation as the primary artifact generated by the artist, with the aural experience that it encodes as a mere and arbitrary by-product.

In fact, it's evident from the illustrations that the author is entranced by graphical notation, hence the careful choice of visually pleasing examples. There's little or no reference to listeners, audiences, or any aesthetic underpinning other than visual ones.

Stockhausen? I haven't conducted any but I once trod in some.
posted by i_am_joe's_spleen at 9:48 PM on June 16, 2003

PS: AfroCelt Soundsystem's "Noodle" program was a really, really cute sequencer with a very nice GUI, and very fun to play with, and I immediately thought of it when ook described MAX. Unfortunately you can't download it from their site any more, but there's a copy on their second album CD.
posted by i_am_joe's_spleen at 9:51 PM on June 16, 2003

This is the only truly viable composition tool. I used to use a computer to aid the process, but I found that I ended up doing the same thing as this - composing on screen, rather than getting the music right first, then transcribing it. Composition becomes an obsession with process rather than sound.

Since I switched to a more flexible medium for transcription, I've been composing a lot better. Right up to the point where someone else has to read it.

Stockhausen? I haven't conducted any but I once trod in some.
I'm coming round to Stockhausen more recently, partly because I've realised that he has developed, and there's more to him than his more notorious, horrendously dated 'avant-garde' stuff. In pieces like Stimmung (a personal favourite) he reveals himself as adept at manipulating timbres musically, in a context unusual enough to be slightly disconcerting.

This page has a good range of sound files as a starting point, including bits of Freitag aus Licht, the lates in his sequence of operas for each day of the week.
posted by monkey closet at 1:24 AM on June 17, 2003

You're a hairy-knuckled throwback, monkey closet, but I like you.

OK, I may listen to some Stockhausen. The aphorisms were the best thing about Beecham anyway.
posted by i_am_joe's_spleen at 2:15 AM on June 17, 2003

You're a hairy-knuckled throwback, monkey closet, but I like you.

Sounds like you know me, too.
posted by monkey closet at 2:42 AM on June 17, 2003

Sidenote for those in or visiting the Chicago area, apropos of "artistic visual beauty of score vs. artisitc aural beauty of composition" above: the Museum of Contemporary Art has a fascinating John Cage piece displayed right now. While I admit to being frustratingly unable to recall the precise title, it's essentially a map of the Chicago area with seventy-some-odd intersections marked out and connected via multi-colored lines, along with Cage's own instructions for what part of the composition can be heard there. Originally designed by Cage in 1978 but apparently not recorded then, it's been recently reconstructed by two UofC musicians.
posted by JollyWanker at 8:56 AM on June 17, 2003

Duh. Cage's A Dip in the Lake - Ten Quick Steps, Sixty-one Waltzes and Fifty-six Marches for Chicago and Vicinity, including a not-so-great scan of the map detailing the intersections.
posted by JollyWanker at 8:59 AM on June 17, 2003

What's described here may well be art, but it's not music.

Let's not get into that "this/not-this" tangent. joe's_spleen, your objections to this, as stated, apply not only to early Stockhausen, but to the entire genre of serialism that dominated Western musical composition for a quarter of a century. Personally, I can't stand to listen to any of that stuff, but that doesn't give me the right to say whether or not it's music.

While it's true that this notational system can't entirely escape time, because at some point the diagram needs to be "rendered" into our time dimension, I think the point is more that you could have a "composition" with the same appearance/relationship between objects, that takes on different shadings (and thus completely different sounding music within the time dimension) according to how you apply the "time" option. Yeah, it tends to produce stuff that sounds like serialism, but there will also (inevitably) be unexpected gems of sound here and there. Then you can look into what visual patterns reinforce the production of "gems."

While I'm not interested enough in this to download it, I definitely agree that Western music notation has problems representing time and rhythmic relationships. Anything that offers a different perspective on that is valuable.
posted by soyjoy at 9:45 AM on June 17, 2003

i don't see how you can declare that this is, or isn't, music without first being quite careful about what your defintion of music is, and what "this" is referring to. is the data encoded by the mp3s not music (whatever th eprocess to egnerate them)? if the process excludes the results from being music, what if they (the sounds) are later treated as found objects by another composer? do they remain not music then?

alternatively, if the generated scores are played by "real musicians" doesn't that get rid of the piano-player accusations? piano-player rolls are notations of known pieces (of music), after all.

on a different track - i think it's interesting to use computers in experiments with notation like this because you have a certain degree of faith in them following the rules. when i hear some someone playing from a squiggly score by stockausen or kagel or whatever i never have that much faith in them really following the notation - maybe they're just doing something along the lines of whatever they heard on a recording which, in turn, was the original composer doing what he wanted when he made the original squiggles, but not necesarily what he managed to express...
posted by andrew cooke at 10:27 AM on June 17, 2003

Not at all, soyjoy. Some serialists at least genuinely had what things sounded like in their minds. I cannot believe that for example Schoenberg did not know what his music would sound like when played. Serialists specified pitch, rhythm, and if it was ugly, it was reflecting the ugliness of an industrialising world full of ugly noises of its own. You seem to be confusing serialism, or atonal music in general, with all avante-garde music (or, sticking to my guns, "music").

The reason I (very provocatively, I admit) said "it's not music" was because in reducing the sound and the time to side-effects I think you throw away the essential elements.

Maybe this isn't a good analogy, but creating music that's so driven by the notation system seems to me like creating paintings according to which paints smell best together.

(Yes, I know standard notation is limited and presupposes that you are working within the western musical tradition, but at least those limits grew out of formalising a shared understanding of what consituted music and weren't chosen a priori.)

andrew cooke, I think what's bothering me most is the apparent belief that the encoding is as important or more important than what is encoded. The bits in your mp3 files aren't music; the music is the sounds they encode.
posted by i_am_joe's_spleen at 1:04 PM on June 17, 2003

Serialists specified pitch, rhythm, and if it was ugly, it was reflecting the ugliness of an industrialising world full of ugly noises of its own. You seem to be confusing serialism, or atonal music in general, with all avante-garde music (or, sticking to my guns, "music").

joe's_spleen, with all due respect, I don't think I'm confusing serialism with anything. Rather, I believe you're confusing serialism with futurism, at least in that "industrialising world full of ugly noises" stuff. The serialists were not trying to evoke anything other than the supremacy of the human brain in cleverly grouping things in order, and the whole "total serialism" movement of the late 50s / early 60s had everything to do with applying mathematical formulas and symmetries to pitch, loudness, duration, etc. and little-to- nothing to do with how the music (yes, "music") sounded. So you'll need to either expand your definition of music or declare that serialists didn't make music, one or the other.

As a side note, Schoenberg did invent serialism, but only at the very end of his career. While he was composing simply atonal 12-tone music I'm sure he did give a great deal of consideration to how things sounded. But the nature of serialism is very explicitly to elevate the process over the resulting sound, thus anything he did in that vein would not qualify - apparently - as "music" within your parameters.

But let's forget the is-it-or-isn't-it nonsense: The question really is whether GeoMaestro has a genuine musical purpose. And the key is feedback. As I said, you could use this system to generate an initial composition to be revised (once heard) just as you could with conventional notation.

And there's something to be said for generating unexpected patterns. If I sit down at the piano to "make something up," my fingers are going to automatically fall into familiar patterns immediately and try to go in directions (melodic and chordal) that are already well-trodden. If I close my eyes and hit the keyboard in an awkward way, I may come up with something I wouldn't have, something I can now work from. To me, there's very little difference between the eye-closing-awkward-slam and the computer-generation-of-visually-based-patterns, other than the level of technology involved.
posted by soyjoy at 1:47 PM on June 17, 2003

Hmm. I'm going to pass on the serialism debate.

I agree with you about the feedback.

I wonder how long it would take to obtain a sense of what GeoMaestro shapes generate what sounds.

see, even though i'm highly skeptical, i am downloading a copy to play with.
posted by i_am_joe's_spleen at 2:13 PM on June 17, 2003

I think Messiaen made a good crack at eradicating time with his "non-retrogradable rhythms", which are basically palindromes. Stravinsky used palindromes, but (in the few examples I have investigated) not in as pensive and stretched out a way as Messiaen. At first, I thought, "Same forward and backward. So what!", but listening to it, you do get a sense of timelesness out of it.

But the nature of serialism is very explicitly to elevate the process over the resulting sound

I must disagree. The sound and structure of that sound embodies the process.
posted by mblandi at 2:19 PM on June 17, 2003

Personally, I can't stand to listen to any of that stuff, but that doesn't give me the right to say whether or not it's music.

No, what gives you the right to say whether or not something is music is the fact that you have functional ears.
posted by kindall at 4:32 PM on June 17, 2003

kindall, I think (I hope) you mean that functional ears give me the right to say something "is good music," not "is music." There's a big difference: You and I both have functional ears (I presume) and we can listen to the same sound source and I can say "that is good music" and you can say "that is bad music" and we can both be right because good and bad, when applied to music, are values tied to the person evaluating something. But we can't both be right if I say "that is music" and you say "that is not music." That's a question of ontology - either it is or it isn't - and it's a question that, again, I think is better left to extremely stoned college students to sort out.

The sound and structure of that sound embodies the process.

mblandi, while that's true, it's also true of anything, including GeoMaestro. The fact that the sound's structure "embodies the process" doesn't contradict the notion of process being more important than result (sound). I think it's fair to say the serialists were largely unconcerned with how their music sounded to the average listener, but very concerned with how it would stack up to close analysis of the use of hexachords, retrograde inversion and other fetishes devices. The feedback element was largely missing, because how a piece sounded once it was played was not as important as how well-put-together it was.

Unsurprisingly, the listening public largely rejected serialism as unlistenable, because of a fatal miscalculation made by Schoenberg, in looking for something other than tonality to give music structure: People in general don't want music to be organized in a way that pleases the intellect at the expense of the ears. They want it to be organized in a way their ears can pick up. And to the ear, serialism sounds the same as aleatory music, so why bother with the math if your result is indistinguishable from randomness? Serialism had potential only as a tool, not (IMO) as an overarching discipline.

Take Messaien, for example - he used serialist and other restrictive procedures but never gave the process over to them, insisting on music with a certain sounding quality - for Messaien, the feedback factor was very high. Personally, I think his "Trois Petites Liturgies de la Presence Divine" is one of the greatest achievements of Western music, and I know he used plenty of devices such as non-retrogradable rhythms in it, but I have no idea where or how. It just sounds fantastic.

Of course, it helped that Messaien had a whole nother over-arching organizing principle working for him - God. But that's a story for another day...
posted by soyjoy at 8:35 PM on June 17, 2003

soyjoy, I really appreciate your comments, which are what I wish all MeFi comments were like.

You say: "Serialism had potential only as a tool, not (IMO) as an overarching discipline."

And I think that's the thing I would fear with GeoMaestro, which appears to prioritise the compositional process over the result.

I was made to practise serial composition both in high school and university, and the experience has left me embittered. :-)
posted by i_am_joe's_spleen at 10:12 PM on June 17, 2003

My point is that these composers were interested in creating an end result using the process as a tool. There is explicit evidence of control of this result through the process in the work of the great 12-tone and serialist composers. The movement was not explicitly about elevating process above its end sound result.

Now, turning to a "listening public" to determine whether these works as sound are valuable can be like picking a set of jurors. Lawrence Welk fans are barred! Whether the sounds are really likeable to tonal music fans is not relevant to the question of whether these works are valuable. I have heard many excellent deterministic works that require no investigation of process to be enjoyed.

My jury is still out on the equivalence of sound quality in determinism and anti-determinism. There are many things unobtainable in determinism that can be achieved with anti-determinism. The same for vice-versa. Again, Joe on the street or even many learned listeners might say "ahh, it's all shit!", but how helpful is that?
posted by mblandi at 11:48 AM on June 18, 2003

Whether the sounds are really likeable to tonal music fans is not relevant to the question of whether these works are valuable.

True, as far as it goes. (Who, after all, is not a "tonal music fan"?) But the question isn't about the works' value, it's about where their center of gravity is, on process or result. I think the fact that those who could understand the background of these pieces embraced them, while those who were trying to listen to them simply as music overwhelmingly rejected them, speaks pretty clearly about where that center of gravity was. And again, I wasn't saying the movement was "explicitly about elevating process above its end sound result" - I think many of the serialists bought the line that this new way of organizing tones would result in listener appreciation, once the public educated themselves and learned to listen more closely. But they were wrong, insofar as the music didn't sound good enough at the outset to make most people want to change their way of listening, and serialism as a movement died.

(Then again, I'm a tonality heretic: I'm not even convinced that the crux of sonata form - i.e. the restatement of the secondary theme in the tonic key from the dominant - has any aural impact on most listeners, either consciously or unconsciously.)

Again, Joe on the street or even many learned listeners might say "ahh, it's all shit!", but how helpful is that?

It's helpful to me as a composer to know that for all the result Ive come up with, I could've just spent the time throwing the i-ching instead of sweating over complex mathematical operations, because next time, I won't waste my time on the latter, if I've got any sense.
posted by soyjoy at 1:20 PM on June 18, 2003