Impossible piano piece visualized and (mostly) performed
March 23, 2008 10:47 AM Subscribe
John Mark Harris provides
a interactive
graphical score synchronized to his realization of the
architect-composer Iannis
Xenakis's Evryali, a piano piece that
is intentionally
impossible to play as written. Harris's notes on the piece are behind the non-obvious "on Evryali" button on the score page. Things start getting really interesting around page 22.
This is amazing on so many levels. An incredible piece, an incredible rendering, a very interesting score and a brilliant interface that allows the listener to follow along and view and hear the patterns that are being sounded. It's excellent stuff. Best of the best of the web.
Thanks dfan.
posted by salishsea at 11:37 AM on March 23, 2008
Thanks dfan.
posted by salishsea at 11:37 AM on March 23, 2008
Thanks, this was a workout, and I'm glad I read the article before going through the score. In my ignorance, I was hoping the dots would light up as they were played, so I could better see which branches were taken and not taken (in a karaoke-classical manner) but I guess that's a bit much to ask. (I could totally play page 18 though--I was all over that one in the listening.)
Very cool. Humans are pretty talented creatures.
posted by roombythelake at 12:40 PM on March 23, 2008
Very cool. Humans are pretty talented creatures.
posted by roombythelake at 12:40 PM on March 23, 2008
I don't know what I like more, the hand-graphed, scanned image, or seeing the notes in this format, or hearing the piece played as I watched the graph. Simply amazing. Great find!
posted by Oxydude at 12:43 PM on March 23, 2008
posted by Oxydude at 12:43 PM on March 23, 2008
Hi, dfan. Your first post is good and you should feel good.
posted by sveskemus at 12:43 PM on March 23, 2008
posted by sveskemus at 12:43 PM on March 23, 2008
I feel like I'm asking for the Cliff's Notes for Finnegan's Wake, but is there anywhere a synthesizer-played version of this piece that doesn't leave anything out? I'm curious to hear what it would sound like compared to the anatomically possible version.
posted by L. Fitzgerald Sjoberg at 12:47 PM on March 23, 2008
posted by L. Fitzgerald Sjoberg at 12:47 PM on March 23, 2008
Yeah page 18, I was totally like, I could really rock this part...
posted by From Bklyn at 1:09 PM on March 23, 2008
posted by From Bklyn at 1:09 PM on March 23, 2008
Bah! Child's play. If he really wants a challenge, he should try taking a crack at John Stump's "Faerie's Aire and Death Waltz". Remember: Like a dirigible. Like a dirigible.
posted by Rhaomi at 1:14 PM on March 23, 2008 [4 favorites]
posted by Rhaomi at 1:14 PM on March 23, 2008 [4 favorites]
is there anywhere a synthesizer-played version of this piece that doesn't leave anything out?
this guy says he's realized every note of it with computer
posted by pyramid termite at 2:51 PM on March 23, 2008
this guy says he's realized every note of it with computer
posted by pyramid termite at 2:51 PM on March 23, 2008
Uh, yeah. Not wanting to be up myself too much, but if he was aiming for 'impossible to play', he could have done a lot, a lot much better.
posted by chrismear at 3:35 PM on March 23, 2008
posted by chrismear at 3:35 PM on March 23, 2008
f he was aiming for 'impossible to play', he could have done a lot, a lot much better.
he tried but the cats got tired
posted by pyramid termite at 3:56 PM on March 23, 2008
he tried but the cats got tired
posted by pyramid termite at 3:56 PM on March 23, 2008
Yeah, here's my idea for the impossible-to-play piano piece:
Hit all the keys at once.
Gotcha! Am I a genius yet?
posted by synaesthetichaze at 4:07 PM on March 23, 2008
Hit all the keys at once.
Gotcha! Am I a genius yet?
posted by synaesthetichaze at 4:07 PM on March 23, 2008
That's an amazing work and amazing performance. Thanks for posting this.
And yet... there are bits of that score that made me think of a Dungeons and Dragons map.
posted by ardgedee at 4:09 PM on March 23, 2008
And yet... there are bits of that score that made me think of a Dungeons and Dragons map.
posted by ardgedee at 4:09 PM on March 23, 2008
Yeah, here's my idea for the impossible-to-play piano piece:
Hit all the keys at once.
Gotcha! Am I a genius yet?
posted by synaesthetichaze
Sorry, they solved that one a lo-o-ong time ago. You use a custom-cut board. Wham!
I am not making this up! This is art we're talking here!
posted by tspae at 5:26 PM on March 23, 2008
Hit all the keys at once.
Gotcha! Am I a genius yet?
posted by synaesthetichaze
Sorry, they solved that one a lo-o-ong time ago. You use a custom-cut board. Wham!
I am not making this up! This is art we're talking here!
posted by tspae at 5:26 PM on March 23, 2008
Something particularly interesting about this is that Xenakis seems to have invented a new graphical format for musical score display. Correct me if I'm wrong, but it appears to be a graph where X represents time, and Y represents the keys on the keyboard of a normal piano. So in theory any musical piece played on the piano, or any similar instrument, could to some extent be represented in this graphical format, although some subtleties could be lost. The relative force with which a key is to be brought down could be represented by the size of the blob, a key held over time could be shown as a line (interestingly, Evryali is almost entirely staccato), and I suppose a few more things which are beyond my (very basic) knowledge of music could be done with colors.
It would be trivial (although tedious) to convert this, or any similarly displayed piece, to a MOD file, or even to a pianola roll, although the roll would be at serious risk of tearing along the perforations. One would probably need to use a very strong paper-like plastic, rather than the usual paper. Print it at the right size, align it correctly, and use it as a guide to punch holes. This would also have the advantage of showing the keys as pressed.
With all due respect to Marc Couroux, who actually knows the field which he describes, it does seem to me that a combinatorial approach can be taken to "virtuosity" as he describes it. If we take as a given Xenakis's representational format for piano pieces, expanded a little as described above, then any given "moment" in the score is a number. (Although it might be easier to treat it as 88 separate numbers.) Alternatively, it's representable as a graphical object, one of a limited set of such: I'll call them "pianospectra". The constraints on playability of pieces by any human being at all are describable; clearly, we can dismiss a lot of the pianospectra in the set according to a few rules, eg requiring the simultaneous pressing of keys that are so far away from each other that no human being (even a seven-fingered pianist with vast and yet delicate hands) could press them at once. If we allow the pianist access to a series of rods with which to press keys that otherwise cannot be, we can expand the set a little, but that's not really the point of the exercise.
On the other hand, so to speak, we can classify a lot of the pianospectra as representing trivial, or easy, keypress combinations. Any single key, anywhere, anyhow; any two; any three, where two are no more than X keys apart; any four, where any two are no more than X keys apart; the same for up to 10, assuming a normally-equipped pianist who can wear a medium glove. However past four, we're asking the pianist to form his/her hands into configurations in order to press chords, and some of these, ergonomically speaking, may be more stressful than others. Some may actually be impossible, as per the "paralyzed finger" party trick. Between these extremes of definite yes and definite no, lie some maybes, which each require certain amounts of skill and natural talent, part of the "virtuosity" Couroux describes.
The other part of physical virtuosity, it seems to me, is in the sequenced combinations of the pianospectra themselves. Some of these are outright impossible too; there is a physiological limit to how fast the human hand can be moved from one key to the key next to it, by the fastest possible human being. This could serve as a kind of "Planck time" of virtuosity, and any sequence of pianospectra requiring faster action than that can be dismissed. There is another "Planck time" (or plonk time) as well--that's a contradiction in terms, but anyway, carrying on: since our pianist is assumed to have two hands, no more, no less, there is also a maximum distance we expect his/her hand to travel: plonk the three keys on the extreme left and the three keys on the extreme right at the same time, then move the right (or left) hand to beside the left (or right) hand and plonk all six. This maximum-distance travel itself has a physiological minimum time, which is different from the minimum time for key X to key X+1. Indeed we can describe distinct minimum physiological times for X --> X+1 all the way up to X --> X+87, and dismiss all configurations that exceed these.
Actually the smarter approach might be to classify all motions according to their approximate energy requirement, ie how tired they might make our poor suffering virtuoso, and simply dismiss all motions that are beyond any human limit.
Assuming we're still having fun, we can next look at the effect of repetition on the pianist over the course of the piece. He/she obviously starts fit, rested, fed, perhaps caffeinated, and in any case ready to play. Over the course of the piece, he/she will become more and more tired, more and more cramped in the fingers, until his/her limit of willingness or limit of capacity is reached, whichever is sooner. Having done the work in the last paragraph, and with access to the annals of exercise physiologists, we can infer with some limited accuracy for how long the fittest possible pianist might be able to play a given sequence of notes. (At which point it becomes crystal clear exactly why there are such gaps in Xenakis' Evryali.)
I assert that we (by "we", I mean you, because I for damn sure am not interested in actually doing this series of bizarre and somewhat cruel experiments in pianology) can, according to this method I describe, classify piano pieces along a scale of physical fitness requirement, in terms of both speed and endurance, and we can therefore classify virtuosity in terms of capacity to play higher-rated pieces.
None of this in any way considers aesthetic musicality. I will say that although I am impressed by Evryali as a concept, and would be impressed to watch it performed, I found it somewhere between dull and annoying to actually listen to.
No doubt we could further classify sequences of pianospectra according to their tolerability (at some assumed normal volume) to the human ear. Some such pieces, I'm sure, will induce fits in susceptible audiences; perhaps there is a series of notes which, being played, might induce fits in all but the least susceptible human being. Perhaps some of these pieces might likewise induce fits, or attacks, or even as-yet-unknown conditions in the pianist him/herself, as well as driving the audience to ever-greater spirals of intermingled despair and rapture.
Ladies and gentlemen, I give you the Yellow Coronation Dance (Aeschenkarnos, 2008, described but uncomposed). No really, I give it to you. I don't want it any more.
posted by aeschenkarnos at 6:09 PM on March 23, 2008 [1 favorite]
It would be trivial (although tedious) to convert this, or any similarly displayed piece, to a MOD file, or even to a pianola roll, although the roll would be at serious risk of tearing along the perforations. One would probably need to use a very strong paper-like plastic, rather than the usual paper. Print it at the right size, align it correctly, and use it as a guide to punch holes. This would also have the advantage of showing the keys as pressed.
With all due respect to Marc Couroux, who actually knows the field which he describes, it does seem to me that a combinatorial approach can be taken to "virtuosity" as he describes it. If we take as a given Xenakis's representational format for piano pieces, expanded a little as described above, then any given "moment" in the score is a number. (Although it might be easier to treat it as 88 separate numbers.) Alternatively, it's representable as a graphical object, one of a limited set of such: I'll call them "pianospectra". The constraints on playability of pieces by any human being at all are describable; clearly, we can dismiss a lot of the pianospectra in the set according to a few rules, eg requiring the simultaneous pressing of keys that are so far away from each other that no human being (even a seven-fingered pianist with vast and yet delicate hands) could press them at once. If we allow the pianist access to a series of rods with which to press keys that otherwise cannot be, we can expand the set a little, but that's not really the point of the exercise.
On the other hand, so to speak, we can classify a lot of the pianospectra as representing trivial, or easy, keypress combinations. Any single key, anywhere, anyhow; any two; any three, where two are no more than X keys apart; any four, where any two are no more than X keys apart; the same for up to 10, assuming a normally-equipped pianist who can wear a medium glove. However past four, we're asking the pianist to form his/her hands into configurations in order to press chords, and some of these, ergonomically speaking, may be more stressful than others. Some may actually be impossible, as per the "paralyzed finger" party trick. Between these extremes of definite yes and definite no, lie some maybes, which each require certain amounts of skill and natural talent, part of the "virtuosity" Couroux describes.
The other part of physical virtuosity, it seems to me, is in the sequenced combinations of the pianospectra themselves. Some of these are outright impossible too; there is a physiological limit to how fast the human hand can be moved from one key to the key next to it, by the fastest possible human being. This could serve as a kind of "Planck time" of virtuosity, and any sequence of pianospectra requiring faster action than that can be dismissed. There is another "Planck time" (or plonk time) as well--that's a contradiction in terms, but anyway, carrying on: since our pianist is assumed to have two hands, no more, no less, there is also a maximum distance we expect his/her hand to travel: plonk the three keys on the extreme left and the three keys on the extreme right at the same time, then move the right (or left) hand to beside the left (or right) hand and plonk all six. This maximum-distance travel itself has a physiological minimum time, which is different from the minimum time for key X to key X+1. Indeed we can describe distinct minimum physiological times for X --> X+1 all the way up to X --> X+87, and dismiss all configurations that exceed these.
Actually the smarter approach might be to classify all motions according to their approximate energy requirement, ie how tired they might make our poor suffering virtuoso, and simply dismiss all motions that are beyond any human limit.
Assuming we're still having fun, we can next look at the effect of repetition on the pianist over the course of the piece. He/she obviously starts fit, rested, fed, perhaps caffeinated, and in any case ready to play. Over the course of the piece, he/she will become more and more tired, more and more cramped in the fingers, until his/her limit of willingness or limit of capacity is reached, whichever is sooner. Having done the work in the last paragraph, and with access to the annals of exercise physiologists, we can infer with some limited accuracy for how long the fittest possible pianist might be able to play a given sequence of notes. (At which point it becomes crystal clear exactly why there are such gaps in Xenakis' Evryali.)
I assert that we (by "we", I mean you, because I for damn sure am not interested in actually doing this series of bizarre and somewhat cruel experiments in pianology) can, according to this method I describe, classify piano pieces along a scale of physical fitness requirement, in terms of both speed and endurance, and we can therefore classify virtuosity in terms of capacity to play higher-rated pieces.
None of this in any way considers aesthetic musicality. I will say that although I am impressed by Evryali as a concept, and would be impressed to watch it performed, I found it somewhere between dull and annoying to actually listen to.
No doubt we could further classify sequences of pianospectra according to their tolerability (at some assumed normal volume) to the human ear. Some such pieces, I'm sure, will induce fits in susceptible audiences; perhaps there is a series of notes which, being played, might induce fits in all but the least susceptible human being. Perhaps some of these pieces might likewise induce fits, or attacks, or even as-yet-unknown conditions in the pianist him/herself, as well as driving the audience to ever-greater spirals of intermingled despair and rapture.
Ladies and gentlemen, I give you the Yellow Coronation Dance (Aeschenkarnos, 2008, described but uncomposed). No really, I give it to you. I don't want it any more.
posted by aeschenkarnos at 6:09 PM on March 23, 2008 [1 favorite]
Best of the fuckin' web.
It's interesting (though a bit disconcerting) to turn the virtuosity of the performer into an element of the piece. The pianist is then overtly a tool of the composer, although I guess that's always been the case.
posted by mammary16 at 8:07 PM on March 23, 2008
It's interesting (though a bit disconcerting) to turn the virtuosity of the performer into an element of the piece. The pianist is then overtly a tool of the composer, although I guess that's always been the case.
posted by mammary16 at 8:07 PM on March 23, 2008
Something particularly interesting about this is that Xenakis seems to have invented a new graphical format for musical score display.
I guess I don't see the difference between this and a player piano roll. Can you elaborate?
John Mark Harris provides a interactive graphical score synchronized to his realization of the architect-composer Iannis Xenakis's Evryali, a piano piece that is intentionally impossible to play as written.
If you enjoy this kind of "machine-only" music, I definitely recommend checking out Conlon Nancarrow, who in my mind was the father of this stuff.
posted by Blazecock Pileon at 12:23 AM on March 24, 2008
I guess I don't see the difference between this and a player piano roll. Can you elaborate?
John Mark Harris provides a interactive graphical score synchronized to his realization of the architect-composer Iannis Xenakis's Evryali, a piano piece that is intentionally impossible to play as written.
If you enjoy this kind of "machine-only" music, I definitely recommend checking out Conlon Nancarrow, who in my mind was the father of this stuff.
posted by Blazecock Pileon at 12:23 AM on March 24, 2008
aschenakarnos: many tracking tools and editing tools use this format either primarily or as an alternate representation. It's the only musical inscription form I tolerate. Derived, as Blaze says, from the player piano roll.
It's good to push the boundaries of artistic expression and expectations. And it's better to do it with conscience aforethought. I like this piece, and this post.
posted by seanmpuckett at 5:24 AM on March 31, 2008
It's good to push the boundaries of artistic expression and expectations. And it's better to do it with conscience aforethought. I like this piece, and this post.
posted by seanmpuckett at 5:24 AM on March 31, 2008
« Older I'm mellltiiiiing! | Jesus Shaves Newer »
This thread has been archived and is closed to new comments
posted by LooseFilter at 10:58 AM on March 23, 2008