hello,
ive got some trouble with list handling, maybe somone can give me a hand.
i like to group an unnown number of floats in a list. (these floats are the result of different divisors, used all for a fixed value. like 100/16, 100/32...)
what i want to do is to put them randomly into the list as long as the sum of all items in the list equals the size of the fixed value. this seems to be pretty difficult...since i use random i cant know what the next float will be.
how to handle that? any experience with that issue ?
thanks, johannes
Hallo, YOhannes hat gesagt: // YOhannes wrote:
ive got some trouble with list handling, maybe somone can give me a hand.
i like to group an unnown number of floats in a list. (these floats are the result of different divisors, used all for a fixed value. like 100/16, 100/32...)
what i want to do is to put them randomly into the list as long as the sum of all items in the list equals the size of the fixed value. this seems to be pretty difficult...since i use random i cant know what the next float will be.
I don't know if I understand correctly what you mean, but maybe [list-equalize] from the [list]-abs collection can help. It scales all elements in a list so that their sum is 1.
If you add a new element into the list this will change the value of old elements in the list of course.
Ciao
Frank
Frank Barknecht schrieb:
Hallo, YOhannes hat gesagt: // YOhannes wrote:
ive got some trouble with list handling, maybe somone can give me a hand.
i like to group an unnown number of floats in a list. (these floats are the result of different divisors, used all for a fixed value. like 100/16, 100/32...)
what i want to do is to put them randomly into the list as long as the sum of all items in the list equals the size of the fixed value. this seems to be pretty difficult...since i use random i cant know what the next float will be.
I don't know if I understand correctly what you mean, but maybe [list-equalize] from the [list]-abs collection can help. It scales all elements in a list so that their sum is 1.
If you add a new element into the list this will change the value of old elements in the list of course.
Ciao
maybe i should explain a bit more... the list is thougt as a sequence in (ms)values. i use the length of a master sequence (ms) and divide it by these values ( 2, 4, 8, 16, 32). now my idea was to use random and gate to trigger the different results of the division into a new slave sequence, the list. so the length of it depends on the random values...
thanks.j
On Wed, 4 Mar 2009, YOhannes wrote:
what i want to do is to put them randomly into the list as long as the sum of all items in the list equals the size of the fixed value. this seems to be pretty difficult...since i use random i cant know what the next float will be. how to handle that? any experience with that issue ?
Make a random recursive partition of the fixed value, then shuffle it.
For example:
start with the list 16 (one element)
for each number in the list, throw dice to figure out whether you will replace it by two equal numbers. any new numbers shall have the same procedure reapplied to them (recursion), else I wouldn't be explaining it by saying "for each" when it's about a one-element list. this could give you, for example: 4 1 1 0.5 0.5 1 2 1 1 0.5 0.5 1 2
a fair shuffle could give any sequence with the same number of 0.5,1,2,4
elements in it, let's say 0.5 4 1 1 0.5 2 1 0.5 2 1 0.5 1 1.
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal, Québec
thanks guys,
i made a simple patch based on mathieus idea. (see attachment) i like to apply more control on listlength, without specifying it, like i would have to with [list-equalize].
now, depending on the devisor of 2, with my master-sequence of lets say 2000 (ms) i get lists with very small and very big values.
for example: 15.625 1.95312 3.90625 1.95312 500 250 1000 31.25 62.5 7.8125 125
any ideas how to get smoother/smaller differences between the list-items?
thanks. j
#N canvas 437 216 195 346 10; #X obj -72 146 urn; #X obj -72 96 iter; #X obj -72 24 inlet; #X obj -72 173 pack 0 0; #X obj -72 293 outlet; #X msg 12 100 clear; #X obj -72 120 t b a; #X obj -72 61 t l l b; #X obj -72 259 zl group; #X obj 10 126 list length; #X obj -72 238 pipe; #X obj -72 216 unpack 0 0; #X obj -72 194 zl rev; #X connect 0 0 3 0; #X connect 1 0 6 0; #X connect 2 0 7 0; #X connect 3 0 12 0; #X connect 5 0 0 0; #X connect 6 0 0 0; #X connect 6 1 3 1; #X connect 7 0 1 0; #X connect 7 1 9 0; #X connect 7 2 5 0; #X connect 8 0 4 0; #X connect 9 0 0 1; #X connect 9 0 8 1; #X connect 10 0 8 0; #X connect 11 0 10 0; #X connect 11 1 10 1; #X connect 12 0 11 0;
#N canvas 91 22 294 416 10; #X obj 35 0 loadbang; #X obj 111 27 bng 15 250 50 0 empty empty empty 17 7 0 10 -262144 -1 -1; #X obj 111 53 random 2; #X obj 179 243 t f f; #X obj 111 192 f; #X obj 111 279 list-extend; #X obj 77 307 list; #X text -31 137 master(ms); #X obj 77 337 shakeitt; #X obj 77 359 print that; #X obj -22 1 inlet; #X obj 111 1 inlet; #X obj 153 337 list-accum; #X floatatom 153 360 5 0 0 0 - - -; #X obj 111 74 route 0 1; #X obj 179 192 f; #X msg 35 139 2000; #X obj 111 96 t b b b; #X obj 179 214 / 2; #X msg -22 27 set $1; #X text 14 359 slave(ms); #X connect 0 0 16 0; #X connect 1 0 2 0; #X connect 2 0 14 0; #X connect 3 0 5 0; #X connect 3 1 4 1; #X connect 3 1 15 1; #X connect 4 0 5 0; #X connect 5 0 6 1; #X connect 6 0 8 0; #X connect 6 0 12 0; #X connect 8 0 9 0; #X connect 10 0 19 0; #X connect 11 0 1 0; #X connect 12 0 13 0; #X connect 14 0 17 0; #X connect 14 1 15 0; #X connect 15 0 18 0; #X connect 16 0 15 1; #X connect 16 0 4 1; #X connect 17 0 16 0; #X connect 17 0 6 0; #X connect 17 1 5 1; #X connect 17 2 4 0; #X connect 18 0 3 0; #X connect 19 0 16 0;
On Thu, 5 Mar 2009, YOhannes wrote:
i like to apply more control on listlength, without specifying it, like i would have to with [list-equalize]. now, depending on the devisor of 2, with my master-sequence of lets say 2000 (ms) i get lists with very small and very big values. for example: 15.625 1.95312 3.90625 1.95312 500 250 1000 31.25 62.5 7.8125 125 any ideas how to get smoother/smaller differences between the list-items?
You can also split note durations using a 2:1 ratio (2/3 + 1/3) and/or using a 3:1 ratio (3/4 + 1/4). In "normal" music this tends to happen non-recursively, that is, you would rarely ever apply such a ratio on an interval is already 2/3 or 1/3 or 3/4 of something else. But if you are ok to randomise intervals by a plain shuffle, you aren't making "normal" music anyway, so, why worry.
There is music in 9/8 time-signature, and there is also music in 3/4 time-signature that has triplets inside it, but recent history has shown that people are a lot more tolerant to bottom-up rhythm construction than top-down: that is, if you take a 4/4 rhythm and split it twice in thirds, it's harder to get the mind into the beat than switch the beat to 3/4 and then split it in thirds, or just switch the whole beat to 9/8.
You can also find music in 5/4, 7/8, 11/8, 13/8, 15/8, and many combinations of those in the same tune, but you will rarely have anything be split in fifths.
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal, Québec
--- On Thu, 3/5/09, Mathieu Bouchard matju@artengine.ca wrote:
From: Mathieu Bouchard matju@artengine.ca Subject: Re: [PD] list issue To: "YOhannes" THIS_IS_POP@web.de Cc: pd-list@iem.at Date: Thursday, March 5, 2009, 5:39 PM On Thu, 5 Mar 2009, YOhannes wrote:
i like to apply more control on listlength, without specifying it, like i would have to with
[list-equalize].
now, depending on the devisor of 2, with my
master-sequence
of lets say 2000 (ms) i get lists with very small and
very big values.
for example: 15.625 1.95312 3.90625 1.95312 500 250 1000 31.25 62.5
7.8125 125
any ideas how to get smoother/smaller differences
between the list-items?
You can also split note durations using a 2:1 ratio (2/3 + 1/3) and/or using a 3:1 ratio (3/4 + 1/4). In "normal" music this tends to happen non-recursively, that is, you would rarely ever apply such a ratio on an interval is already 2/3 or 1/3 or 3/4 of something else. But if you are ok to randomise intervals by a plain shuffle, you aren't making "normal" music anyway, so, why worry.
There is music in 9/8 time-signature, and there is also music in 3/4 time-signature that has triplets inside it, but recent history has shown that people are a lot more tolerant to bottom-up rhythm construction than top-down: that is, if you take a 4/4 rhythm and split it twice in thirds, it's harder to get the mind into the beat than switch the beat to 3/4 and then split it in thirds, or just switch the whole beat to 9/8.
You can also find music in 5/4, 7/8, 11/8, 13/8, 15/8, and many combinations of those in the same tune, but you will rarely have anything be split in fifths.
Unless you happen to be listening to Carter, Cowell, Ferneyhough, Johnston, Nancarrow, or anyone who has ever happened to use a quintuplet (Chopin, Elvin Jones, maybe Al Pacino in "Heat")
-Jonathan
_____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal, Québec_______________________________________________ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On Thu, 5 Mar 2009, Jonathan Wilkes wrote:
Unless you happen to be listening to Carter, Cowell, Ferneyhough, Johnston, Nancarrow, or anyone who has ever happened to use a quintuplet (Chopin, Elvin Jones, maybe Al Pacino in "Heat")
I didn't say that quintuplets don't happen!
I mean that quintuplets are relatively rare, and they rarely are used that much in one piece. You could take a complete piece of 5/4 and write it 4/4 and it would have quintuplets all over, and so you could take a piece that is mostly made of quintuplets and disguise it as a 5/4 without quintuplets, but if it wasn't all made of quintuplets, then the result should have some kind of anti-quintuplets, that is, multiplying the note durations by 1.25 times instead of 0.8 times. But I've never seen this in one notational form or another. Independently of how it's written, I've never heard it either.
So, there is this asymmetry whereby you can find plenty of patterns of accretion of durations in multiples of 5, but not much splitting of durations in multiples of 5.
I'd be glad to get references to specific pieces that contain a lot of quintuplets. Of the five composers you named, I only heard Nancarrow, I only heard of Nancarrow, and then, I don't recall any quintuplets in it, but I didn't see the score and perhaps I couldn't grasp the rhythm of it just by ear (?).
It reminds me, I once applied in music composition at UQÀM, but when they sent me a letter telling me they wanted to test my piano skills with sheet music, I didn't reply, because I don't have any piano skills and can't read sheet music in realtime. The only instrument I was really ever proficient with was ScreamTracker, and boy did I abuse that thing. I made pieces in 21/8 time and whatnot. I found that notation much easier to deal with than classical notation... except for making quintuplets, that is.
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal, Québec
--- On Fri, 3/6/09, Mathieu Bouchard matju@artengine.ca wrote:
From: Mathieu Bouchard matju@artengine.ca Subject: Re: [PD] list issue To: "Jonathan Wilkes" jancsika@yahoo.com Cc: "YOhannes" THIS_IS_POP@web.de, pd-list@iem.at Date: Friday, March 6, 2009, 1:29 AM On Thu, 5 Mar 2009, Jonathan Wilkes wrote:
Unless you happen to be listening to Carter, Cowell,
Ferneyhough, Johnston, Nancarrow, or anyone who has ever happened to use a quintuplet (Chopin, Elvin Jones, maybe Al Pacino in "Heat")
I didn't say that quintuplets don't happen!
I mean that quintuplets are relatively rare, and they rarely are used that much in one piece. You could take a complete piece of 5/4 and write it 4/4 and it would have quintuplets all over, and so you could take a piece that is mostly made of quintuplets and disguise it as a 5/4 without quintuplets, but if it wasn't all made of quintuplets, then the result should have some kind of anti-quintuplets, that is, multiplying the note durations by 1.25 times instead of 0.8 times. But I've never seen this in one notational form or another. Independently of how it's written, I've never heard it either.
You don't see that because composers generally don't hide quintuplets by adjusting the time signature. Put anything intended for human musicians into 5/4, for example, and then you must subdivide the measure-- 3+2, 2+3, etc.-- because that's the first thing a player will want to know/feel when reading it. So now each measure has some assymetrical subdivision. Then any "anti-quintuplets" would have to be shown as tuplets (e.g., 4 in the space of 5), which, even if there are fewer of them, would be unecessarily complex because common time is, well, common, symmetrical, and easier to subdivide.
With quintuplets, there isn't (or at least ideally shouldn't be) an implicit hierarchy of beats within the quintuplet. I think that's why they are favored in late-Romantic piano music, like Scriabin's Prelude No. 1, Op. 15, to get a kind of floating, improvisatory quality in a melody against the underlying pulse. The "anti-quintuplets" here would be everything that's notated in the standard fashion.
So, there is this asymmetry whereby you can find plenty of patterns of accretion of durations in multiples of 5, but not much splitting of durations in multiples of 5.
You see plenty of examples of what Kyle Gann, in his Nancarrow biography, calls divisive rhythm-- dividing a measure up into groups of 5, 7, etc.-- in Schoenberg and many others. There was evidently a big divide between supporters of this procedure, and the additive procedure of Stravinsky et al of building asymmetrical rhythmic groups from a common small subdivision like the 16th-note. But for divisive rhythm, subdividing three or four levels deep quickly gets complicated and hard to read/conceptualize/perform/etc. Many of Stockhausen's early scores (ab)use this degree of complexity. He and others from the 50s evidently didn't find the need to stick strictly to one type of asymmetrical division, however-- there are triplets, nested in quintuplets, nested in whatever. But who knows, maybe there's some integral serialist "pioneer" out there who decided to break with convention and write a piece only using nested quintuplets.
Henry Cowell divised a system of notation for things like "fifth-notes" in his book "New Musical Resources" from 1930. I think he actually used them in Quartet Romantic, though I haven't seen the score in a while.
I'd be glad to get references to specific pieces that contain a lot of quintuplets. Of the five composers you named, I only heard Nancarrow, I only heard of Nancarrow, and then, I don't recall any quintuplets in it, but I didn't see the score and perhaps I couldn't grasp the rhythm of it just by ear (?).
What is the nature of the rhythm you're trying to grasp by ear? If it's rhythm within a common pulse that divides strictly into subdivisions of fives (instead of twos), I don't think I know examples of any music that do that. But it's intriguing to think about some kind of dance music being written with this constraint.
What I have heard are a) quintuplets to notate an independent tempo in a multi-tempo context: Elliott Carter's "A Mirror in Which to Dwell" b) quintuplets as obsessively/excessively written-out rubato: Ben Johnston's "Sonata for Microtonal Piano" and c) tempo ratios like 4/5/6 in Nancarrow's Study No. 49.
There's also the music of Brian Ferneyhough, which I don't know as well as the others. But it's basically divisive rhythm in which long pulses are divided up into any number of large and small nested tuplets. In this case the large tuplets have the effect of changing the tempo within the larger pulse. I'm guessing if you took the patch that's part of this thread and periodically randomized the type of beat divisions, you'd end up with similar-sounding rhythms, with the restriction that any nested tuplet would divide into the same number of parts as it's parent.
It reminds me, I once applied in music composition at UQÀM, but when they sent me a letter telling me they wanted to test my piano skills with sheet music, I didn't reply, because I don't have any piano skills and can't read sheet music in realtime. The only instrument I was really ever proficient with was ScreamTracker, and boy did I abuse that thing. I made pieces in 21/8 time and whatnot. I found that notation much easier to deal with than classical notation... except for making quintuplets, that is.
So did you have to make your quintuplets in another program and then import them?
_____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal, Québec
On Thu, Mar 5, 2009 at 7:29 PM, Mathieu Bouchard matju@artengine.ca wrote:
On Thu, 5 Mar 2009, Jonathan Wilkes wrote:
Unless you happen to be listening to Carter, Cowell, Ferneyhough,
Johnston, Nancarrow, or anyone who has ever happened to use a quintuplet (Chopin, Elvin Jones, maybe Al Pacino in "Heat")
This reminds me, I was just reading The Negative Dialectics of Poodle Play by Ben Watson, and he quotes an interview with Frank Zappa about his favorite records:
"'Can I Come Over Tonight' – The Velours. Any musicologist that can find that record and listen to the bass singer ... he's singing quintuplets and septulets. And considering where it came from and when it was made (it was on the East Coast Onyx label) it was amazing."
Judge for yourself: http://www.youtube.com/watch?v=d-eO1TIgT8A http://www.youtube.com/watch?v=6Tn9TONsftM
--Stefan
here is my humble interpretation of mathieus idea.
Am 05.03.2009 um 12:21 schrieb YOhannes:
thanks guys,
i made a simple patch based on mathieus idea. (see attachment) i like to apply more control on listlength, without specifying it, like i would have to with [list-equalize].
now, depending on the devisor of 2, with my master-sequence of lets say 2000 (ms) i get lists with very small and very big values.
for example: 15.625 1.95312 3.90625 1.95312 500 250 1000 31.25 62.5 7.8125 125
any ideas how to get smoother/smaller differences between the list- items?
thanks. j
#N canvas 437 216 195 346 10; #X obj -72 146 urn; #X obj -72 96 iter; #X obj -72 24 inlet; #X obj -72 173 pack 0 0; #X obj -72 293 outlet; #X msg 12 100 clear; #X obj -72 120 t b a; #X obj -72 61 t l l b; #X obj -72 259 zl group; #X obj 10 126 list length; #X obj -72 238 pipe; #X obj -72 216 unpack 0 0; #X obj -72 194 zl rev; #X connect 0 0 3 0; #X connect 1 0 6 0; #X connect 2 0 7 0; #X connect 3 0 12 0; #X connect 5 0 0 0; #X connect 6 0 0 0; #X connect 6 1 3 1; #X connect 7 0 1 0; #X connect 7 1 9 0; #X connect 7 2 5 0; #X connect 8 0 4 0; #X connect 9 0 0 1; #X connect 9 0 8 1; #X connect 10 0 8 0; #X connect 11 0 10 0; #X connect 11 1 10 1; #X connect 12 0 11 0; #N canvas 91 22 294 416 10; #X obj 35 0 loadbang; #X obj 111 27 bng 15 250 50 0 empty empty empty 17 7 0 10 -262144 -1 -1; #X obj 111 53 random 2; #X obj 179 243 t f f; #X obj 111 192 f; #X obj 111 279 list-extend; #X obj 77 307 list; #X text -31 137 master(ms); #X obj 77 337 shakeitt; #X obj 77 359 print that; #X obj -22 1 inlet; #X obj 111 1 inlet; #X obj 153 337 list-accum; #X floatatom 153 360 5 0 0 0 - - -; #X obj 111 74 route 0 1; #X obj 179 192 f; #X msg 35 139 2000; #X obj 111 96 t b b b; #X obj 179 214 / 2; #X msg -22 27 set $1; #X text 14 359 slave(ms); #X connect 0 0 16 0; #X connect 1 0 2 0; #X connect 2 0 14 0; #X connect 3 0 5 0; #X connect 3 1 4 1; #X connect 3 1 15 1; #X connect 4 0 5 0; #X connect 5 0 6 1; #X connect 6 0 8 0; #X connect 6 0 12 0; #X connect 8 0 9 0; #X connect 10 0 19 0; #X connect 11 0 1 0; #X connect 12 0 13 0; #X connect 14 0 17 0; #X connect 14 1 15 0; #X connect 15 0 18 0; #X connect 16 0 15 1; #X connect 16 0 4 1; #X connect 17 0 16 0; #X connect 17 0 6 0; #X connect 17 1 5 1; #X connect 17 2 4 0; #X connect 18 0 3 0; #X connect 19 0 16 0; _______________________________________________ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-
Frank Barknecht -
Ichabod -
Jonathan Wilkes -
Mathieu Bouchard -
Max -
YOhannes