Slashdot Mirror
Mathematically Pattern-Free Music
gary.flake writes "'Scott Rickard set out to do what no musician has ever tried — to make the world's ugliest piece of music [video]. At TEDxMIA, he discusses the math and science behind creating a piece of music devoid of any pattern.' He used mathematics of Évariste Galois (who was born 200 years ago) to create pattern-free sonar pings which he mapped to notes on a piano, and then played them using the non-rhythm of a Golomb Ruler. Now, why didn't I think of that..."
That's nothing- rap musicians have been doing this for decades.
"That's the way to do it" - Punch
Now this sheds a new light on the term "math rock".
Christopher Pecoraro - Irventu.com
I think the performer made a mistake at 9:13
......"set out to do what no musician has ever tried — to make the world's ugliest piece of music"...... Already done... http://en.wikipedia.org/wiki/(You're)_Having_My_Baby
you won't attract the worm. Another piece of ugly music, Aphex Twin's Ventolin
2. Add Vogon poetry as lyrics. 3. Profit
The consensus was that "Friday" already held that title.
they kinda did it before this guy (at least from a rhythmic perspective), as a protest against the Criminal Justice and Public Order Act 1994 here in the UK. http://en.wikipedia.org/wiki/Anti_EP
Old hat. To discover the life of a musician who made randomization a career, see John Cage.
When the foot seeks the place of the head, the line is crossed. Know your place. Keep your place. Be a shoe.
Well, I use the mathematics of Frank Plumpton Ramsey and Bartel Leendert van der Waerden (who were born about 100 years ago) to call bullshit on this claim: There is no sequence of anything (including musical notes) which is pattern free.
cf.
http://en.wikipedia.org/wiki/Van_der_Waerden%27s_theorem
http://en.wikipedia.org/wiki/Ramsey%27s_theorem
Krzysztof Penderecki's Threnody to the Victims of Hiroshima is the most horrible "music" I've ever heard. (Intentionally so - Penderecki made it as dissonant and a-tonal and possible)
Don't believe me? Listen to it here
To make laws that man cannot, and will not obey, serves to bring all law into contempt.
--E.C. Stanton
There were a few overlapping notes from pedal suspension that created chords. Although they tried to make ugly pattern-free music, they just ended up making modern music.
Occasionally living proof of the Ballmer peak.
Random != no pattern
You might create a tune with no pattern but chances are there will be a pattern of some kind in there.
Interestingly enough, when you listen to such a piece, you hear a pattern of alternating low octaves notes and high octaves notes. Even though there is no mathematical repetition, I couldn't help anticipating the next big octave jump in a direction or the other, and the music had some kind of pattern to me. How weird that we can find meaning and feel sentiment in the most mathematically pattern-less piece of music?
Opening three notes of the Godfather Theme ("Speak Softly Love") at 8:53.
I really didn't think it was all that horrible. I was not ready to jump up and start dancing to it but I feel lot of the shit that gets played on the radio is worse. Due to the arrhythmic and random nature of the piece I bet it was actually a bitch to play on the piano.
This kills off any kabbalist's notion of the importance of numbers as such. Now music have no pattern, too.
For a good time, cat $file > /dev/dsp. My favourite so far is the PS file of Shannon's information theory paper.
Escher was the first MC and Giger invented the HR department.
http://www.youtube.com/watch?v=dOMIl3WjXc0
Something must be wrong with me, because I loved this piece immensely and would really like more. Hearing it again and thus repeating it seems to destroy the beauty of it.
The music starts at 7:40 if you prefer to just listen to the plinking of unmemorable notes...
This isn't non-mathematical random "music", it is mathematical "music" that is patternless. The first isn't music at all, while the second doesn't qualify by most definitions but could by logical extension (which mathematicians do so love.) I personally, as a musician, don't think either one would be music.
"None can love freedom heartily, but good men; the rest love not freedom, but license." --John Milton
Actually, as was explained in detail in the video, random is easy. Completely devoid of repetition is vastly more difficult. This was not simply random, this was mathematically non-repetitive. Using random numbers outside of the audible range would not necessarily preclude repetition, and using random frequencies is atonal sound, not tonal non-repetitive "music" as was the intention of the piece.
Completely random is trivial. Mathematically-sound aperiodic and repetition-free is a completely different kettle o' fish.
Note that the composition used the 88-tone chromatic scale of the standard piano keyboard. Without that constraint, you could make a much longer atonal composition, of course, but the point of the exercise was to use discrete mathematics and music to create a tonal composition completely devoid of repetition.
It would help if there were some definitions for "random" and "pattern-free" in this context. I find it annoying that he several times says that random music is not pattern-free.
It is true that their definitions are not equivalent, but it seems that he is implying that you cannot generate "pattern-free" music using randomly played notes, and that -depending of the definition of "pattern-free" of course- seems very, very unlikely.
Still, I can appreciate the effort to maximize information entropy, and the divulgation of discrete math.
Random != no pattern
You might create a tune with no pattern but chances are there will be a pattern of some kind in there.
Exactly. This is why sports fans think that there's such a thing as form. Human beings are very bad at judging randomness - we actually bias towards alternating patterns, which is decidedly non-random.
Strange women lying in ponds distributing swords is no basis for a system of government.
The chosen constraints were to compose a mathematically non-repetitive piece within the constraints of the 88-tone chromatic scale of a piano. Sure, trombones or fretless stringed instruments could easily go worse thanks to their continuous, i.e. non-discrete, capabilities, but that's a whole 'nother story.
That music isn't ugly. It does happen to be optimally dissonant, but ugly and dissonance are not the same thing. Related - but not same thing.
If you want some truly ugly music I recommend you get to YouTube and check out The Residents. They work hard to bring you the ugly.
Here is an example. It is the Residents covering the Rolling Stones Satisfaction. FAR more ugly than this mathematical oddity. You'll note that it is fairly repetitive and still PLENTY ugly.
Weaselmancer
rediculous.
Primes have no patterns, so why not just map sounds/beats to prime numbers?
It's so cute when the kids think they've discovered something nobody's tried before (eyeballs roll up in head). Welcome to the 20th century!
Bastard Noise: The Analysis of Self-Destruction is the worst voluntary atonal arrhythmic non-patterned music ever.
If you don't consider suicide during a listening session, you are deaf!
Jehovah be praised, Oracle was not selected
Nothing wrong with you at all.
Music evolved as a tool for learning. Rhythmic behaviour around the campfire teaching others how to hunt and all that.
Anything which our brain perceives as innovative in comparison to what we know is considered a new concept, and learning new concepts gives us pleasure (knowing more concepts is a survival trait).
So you get pleasure not from the repetition of patterns in Beethoven's Fifth, but from the interplay and differences. The pattern is set up initially, and then it's how the subsequent patterns are *different* from the initial setup that causes us to learn the much more complex rules which specify the variations. It's the learning that we perceive as pleasurable.
Patternless music is completely different from anything you've encountered. It's not surprising that it gives some people pleasure - the brain is wired to notice and process uniqueness and encode it via differences, and give pleasure in doing so.
I liked the music myself (a lot), and noticed some similarity to the works of Ginastera; specifically, "Concerto for piano and orchestra" which I also like.
I thought it was interesting when Tool designed their lyrics for Lateralus to the Fibonacci Sequence. While I don't think this had any weight of the harmony of the song and why it's such a powerful song, I think it added a level of complexity that I greatly enjoy in their songs. http://en.wikipedia.org/wiki/Lateralus_(song)
John Cage's music employed chance, not randomness. I posted about him back in 2007 (search for my username, my post is near the top.)
Xenakis would be a better example of a composer who used randomness in a truly stochastic sense. However, he used it in a very deliberate and purposeful way, to shape only some elements of a composition, not the entire work. In contrast, Cage used chance as a way of abdicating control, although (like Xenakis' use of randomess) he employed it for only some elements of a work.
If it weren't for deadlines, nothing would be late.
So if I understand this correctly, since 3 and 88 are relatively prime, then every number in the closed field F88 is a multiple of 3, and if you keep multiplying by 3, you'll eventually hit each number.
But that's a pattern, isn't it?
I won't say I enjoyed it. But it was better then Yawny. It was shorter.
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
The math can't, the representation of it can.
There is plenty of "pattern-free" music that is not randomly generated.
Going back to at least the late 50s, there has been "free jazz" that does not have the kind of patterns that one associates with Western music, yet it is anything but random.
As someone who has been working with electronic and computer music since the mid '70s, I am always amused when mathematicians or computer scientists try to use particular algorithms to create music, but leave out the imagination. Though I suppose you could include it in a loose categorization as "music", it lacks the most fundamental element of any art.
Hell, anyone with a copy of MaxDSP could make what this guy made without forgetting the imaginative special sauce that makes it mean something.
You are welcome on my lawn.
A random number generator can generate patterns.
The old hypothetical monkeys-at-typewriters eventually banging out a Shakespeare play describes this. Essentially the monkeys are just a bunch of random character generators. Even if they don't write Shakespeare, they'll eventually stumble across some sort of pattern purely by random chance.
Even though the pattern is not intentional, a pattern can be formed.
"I disagree with you" does not equal "flamebait."
So...this guy is trying to recreate white noise, then? I'm pretty sure there's an app for that...
"I love animals! Some are cute, others are tasty, what's not to like?" - Betsy Schroeder, Jeopardy contestant
Have gnu, will travel.
A random number generator can generate patterns.
The old hypothetical monkeys-at-typewriters eventually banging out a Shakespeare play describes this. Essentially the monkeys are just a bunch of random character generators. Even if they don't write Shakespeare, they'll eventually stumble across some sort of pattern purely by random chance.
Even though the pattern is not intentional, a pattern can be formed.
I guarantee you that the noise this guy shat out has more identifiable patterns in it than what you'd get from any RNG not used by Sony.
Saying something has no "pattern" is absurd - a pattern is not just something that repeats. A pattern is any recognizable characteristic of a thing.
1 2 3 4 5 6 7 8 9 10 11 12 has no repetition but it is obviously a patten.
sfhsl;ga;sgb has no obvious sequence or repetition, but it is obviously just me spamming on the keyboard.
Anything that is deemed "mathematically random" nowadays is, in fact, so very finely tuned to appear random that it is the exact opposite of random.
The pattern of this guy's work is, in fact "Apparently "random" noise with no obvious repetition or sequences".
cat? Real men use dd.
dd if=/dev/random of=/dev/dsp
PlusFive Slashdot reader for Android. Can post comments.
It's got to sound better than Richard MacDuff's tax return.
http://soundandcolor.com/ I have been applying RNGs to create random values for midi notes, based on shifting probabilities.
It is, after all, still better than most of what is sold as "music" these days.
Now get off my lawn...
Damn_registrars has no butt-hole. Damn_registrars has no use for a butt-hole.
...and put on some Rush. :-)
Help save the critically endangered Blue Iguana
Anything that is deemed "mathematically random" nowadays is, in fact, so very finely tuned to appear random that it is the exact opposite of random.
Well, that's the opposit of true. "mathematically random" data has patterns. This music doesn't, meaning it's not random, and would sound different from random noise, though "apparantly random" is apt.
Socialism: a lie told by totalitarians and believed by fools.
I think "emotionless" would have been a better description of the goal of the exercise, since "ugly" can invoke emotion, even if unpleasant, in any kind of music.
I found the music interesting and somewhat emotional, and thought it would make a nice sparse soundtrack piece. Sounds similar to serial music, related to Schoenberg who he mentioned. I'm only slightly familiar with this music, but it reminds me of some pieces by Babbitt. Here's a sample: http://www.npr.org/2006/05/10/5396502/a-difficult-composer-milton-babbitt-at-90. Definitely not ugly. If you're going to compare this "perfect sonar ping" to music at all, it seems like it might be worth doing a more serious comparison to previous composers who worked towards a similar goal.
Personally, I prefer V/VM, the track "Hard drive crash" in particular.
To the under 25s: This is what kids will be listening to when you're middle-aged. Enjoy. :)
http://alternatives.rzero.com/
I shudder at the thought of what happens when the noisegrind community discovers this research.
I guarantee you that the noise this guy shat out has more identifiable patterns in it than what you'd get from any RNG not used by Sony.
OK. Point them out to us. I'll listen again.
A pattern is any recognizable characteristic of a thing.
If that's how you're going to define "pattern" then everything in the universe is a pattern. Such definitions are meaningless because there isn't any useful information to glean from them.
"I disagree with you" does not equal "flamebait."
Whenever some hardcore relativists speaks about all music being equally good, I just get them to load a random exe file into a sample editor and play it back. The result is usually the most horrendous piece of crap you can imagine, that squeals and grates like a ZX spectrum on acid. I'm serious, it's not pure white noise, it really sounds horrible.
Why OpalCalc is the best Windows calc
Why dd? Neither is a block device.
Rethinking email
Meh! Any fourth grade beginning strings class does this. Mind you, the intent may be different, but the results are pretty much the same.
Some mornings it's hardly worth chewing through the restraints to get out of bed.
Or no notes, but i guess Cage's 4'33" did that already
It's only repetition-free if you can hear the intervals accurately, so that a jump from (say) a low A to an F-sharp five octaves up really sounds completely different to you from a jump from a low A to an E. I can't hear long jumps that accurately. By picking notes out of the 88-key keyboard, they get music in which the note-to-note interval jumps are much larger than they are in a traditional tune or theme. Those jumps are so large--and so divorced from any total center--that I, at least, don't hear them as musical intervals at all, but as dramatic contrasts of "high" and "low."
Well, the pattern of "highs" and "lows," divorced from interval, is, in fact one of the salient things we hear in music. There was even a dictionary of music themes once in which you looked up (and could find) themes from symphonies, etc. simply by looking up the pattern of ascending and descending notes. I forget how it was encoded, but you could look up, say START-DOWN-DOWN-UP-UP-UP-UP-DOWN-DOWN-DOWN-UP-UP and it would tell you "The Star-Spangled Banner." People who can't actually read music can use sheet music as a memory aid for remembering notes, just by reading whether the successive notes ascend or descend, without being able to tell a quarter from a half note, or sense the actual intervals. The pattern of ascent and descent conveys much less of the music than the actual intervals, or the rhythm, but it nevertheless is part of the music.
Unfortunately for the goal of producing repetition-free music, there are only so many up-and-down patterns, and in the musical clip I felt I was hearing repetitions of short up-down sequences. Similarly, NOTE, big-multioctave-jump-up, NOTE, somewhat-jump-up, NOTE sounds similar to NOTE, big-multioctave-jump-up, NOTE, somewhat-jump-up, NOTE even if the intervals aren't identical and the note durations aren't identical.
"How to Do Nothing," kids activities, back in print!
do a degree it has been done. If you search around a bit, you can find articles detailing a few bands/musicians who produced a few pieces that if you look at the sound wave pattern, or feed the output of an amp through an oscilloscope, you get pictures or words. The most detailed i recall is a song that the sound wave visualization looks like the face of the artist if i recall. (all this of course is done by fooling with the piece in post production for the most part) Searching for articles about things hidden in music/ on albums is the trick.
I've decided to Diversify my Holdings. I've divided my cash between my left and right pockets, instead of all in one.
Umm, NO. A random number generator can spit out a pattern just as easily as a non-pattern. Ex: 12345 and 57349 will appear just as common from a random number generator.
In a way I'd agree. The point of the exercise was really just to demonstrate pattern-free music. I think this could be used in a more artistic fashion, perhaps running loops of the piece played in different "axis" for instance rotate the grid 90 degrees. One could add patterns of different sizes to counterpoint as well. I'm surprised they used a person and not a Disclavier, a real piano that can be MIDI controlled, to write something much more complex.
But again, that's creating a pattern by using the patternless super pattern. So that wasn't the point of the exercise.
No, you are correct. There is a pattern. A better title would have been "...Repetition-Free Music". Any given set of notes within the "song" has no similar set of notes at another point in the piece. If you were to instead map pi's digits to music, there would be small pieces that repeat. For example, there are several occurrences of the pattern 141593 in the first million digits, and several more occurrences of 252604, and 363715, and so on. All of those would sound similar if mapped to notes. In a Costas array (which was used to make this music), there are no such repeated sub-arrays.
There is a term used by some of the improvisers, the "endless melody" or the "open line". Eric Dolphy and Claude Debussey called it the "unending line".
It is the unfolding melody, not repeating, that goes from here to there without any repeating sections. I'm tired at the moment, so all I can think of is some of the work of Jimmy Giuffre or Paul Bley, but there have been many others. In the classical tradition, LaMont Young and Arvo Part.
As I'm sure you know, human beings when presented with random data will work overtime to find patterns, even imagining them. It's what our brains are made to do. It's one explanation for religious belief and superstition. So the question becomes, if your viewer/listener is going to impose a pattern anyway, why not give him some to work with?
I remember early on being a theoritician, thinking that there was more value in this kind of exercise that this article describes than was really there. I've become more leery of theoretics the older I get, which became a problem since my area of professional expertise was in literary theory and critical theory. Now I care more about something that touches the part that is sometimes called "the soul". Instead of art which instructs, I seek art that moves one, or exhorts, or comforts, or simply celebrates. That doesn't mean it shouldn't challenge, but never at the expense of a human connection.
OK, it's my bedtime now. I've become fully incoherent. Sorry.
You are welcome on my lawn.
Actually it's "useful" in the way mathematics stuff is always beautifully useless. You see, if you wanted to do echolocation with a piano (or any other 88-note instrument), this would be the piece that gave you maximum information on the target.
http://en.wikipedia.org/wiki/Animal_echolocation#Acoustic_features
Believe it or not, if you played this music often you would (after a loong time) become able to hear the differences in the room the music was played in, just by the sound.
Bats use this to accomplish something that seems implausibly difficult, some species use it to dive through moving branches composing the upper level of a forest, in the dark (and they're blind or near-blind anyway), filled with environmental sounds and general noises, at ~ 180 km/h. When stationary they can use the tones to see through walls, and tell from the outside if anything in a room or cave is moving or not, including the rhythm of it's movement.
Try Karlheinz Stockhausen or György Ligeti on for size.
Slashdot's first reaction to VMware
Actually, it's not only music a mathematician could have written, it's also an interpretation only a mathematician could reach. In contrast to what this guy calls perfect lack of repetition the music doesn't lack all repetition. It merely lacks a very distinct property. As I understand it, the special thing about that piece is that not any two subsequent notes differ in the same amount of frequency. Analogous the rhythm. So first of all are trivial and obvious patterns of repetition left: e.g. you will find many times two notes where the second is higher than the first. Or you will find passages, where the tones are de- or ascending, maybe even tied with a similarly shaped rhythm (think of relatively fast or slow sequences). And then there are also properties of sound, which cannot be expressed in formal languages as in modern musical theory. In fact, I would argue that there kinds of music in the world, whose point cannot be expressed by classical formal musical theory. The takeaway or the crucial perspective should be here, that your listening to music is always an act of construction. You will read sense or pattern or intention or whatever you want into music, depending on your socialization, your mood or your capabilities. People will tend to simplify more complex phenomena by means of more general concepts (high following high instead of the actual and precise note pitch and so on). So this interference of mathematics is not so provocative after all. If you can't find a point in this music - fine. But that does not depend on a single and not even most prominent mathematical property.
the face was richard james (aphex twin). somewhat sadly, it's just an off-the-shelf software that does the encoding. kind of wanky, really.
"They were pure niggers." – Noam Chomsky
Just my $Currency*0.02
No, REAL men simply tap the bits on their eardrums with a magnetized needle. With frickin' butterflies.
Escher was the first MC and Giger invented the HR department.
Remember those thousand monkeys at their typewriters? Them writing Shakespeare is an example of patterns appearing in randomness.
The music composed here is perfectly predictable (no random numbers were used), but completely pattern-free in that any relationship between groups of notes only appeared once.
Much of the beauty of music is in the way it is played. The piano player in the TED video was so kind to take the music seriously and make the best of it, i.e. suggesting structure by subtly varying the duration of notes relative to how they are written in the score. The piece would be much uglier when converted into a midi file and played automatically on an electric piano.
Well, different branches of mathematics have different notions of randomness. Your post makes sense assuming a statistician's concept of randomness. Your parent post almost makes sense assuming an information theorist's concept of random. (I say almost because actually proving a sequence to be Kolmogorov random is an incomputable problem.)
Spectrograph is a word that will help, as that's a picture of the sound made by running the raw audio (usually WAV) through a n-band FFT and representing the data as an image. I remember using some software back in the late 90s that was designed to do the opposite, and creating a few short samples from interesting parts of the Mandelbrot set. I don't remember what software I used, and while the computer I used still works the hard drive has been formatted so many times I doubt any trace of it still exists. But google turns up asperes, which looks to do what you want. AudioPaint looks very similar to what I once used, as it uses the color of the pixels to tweek the left/right panning, allowing for more than just a 2D sound.
What I learned back then was that pretty pictures tended to make very horrible sounds if allowed to play for any long amount of time (more than a second) but that they made really interesting drum samples and sound effects. Uninteresting pictures could sound very beautiful in both the short and longer time ranges.
I agree it's not really accurate to call it "The World's Ugliest Piece of Music." That would imply that the beauty of music is solely a function of how repetitive it is, which is obviously false because a single note repeated at regular intervals (or maybe a square wave held indefinitely on a single pitch) would certainly not qualify as "The World's Most Beautiful Piece of Music."
However, I don't mind this as a bit of marketing. There's really no denying that this piece is intended to be listened to after a brief explanation of the mathematics behind it, and "Mathematically Ugliest Music" is a more intriguing hook than the more descriptive "equally-tempered 88-tone row without repeated pitch or interval classes, played so that no two notes are rhythmically separated by the same rhythmic distance, on the premise that repetition is necessary for beauty in music."
While this piece does minimize a positive aspect of music, it does nothing to maximize negative aspects. Dissonance counterpoint comes to mind as a better example of actually trying to write unpleasing music. Basically, it takes the rules of counterpoint theorists have used to describe the music of Palestrina or Bach, then slavishly follows the opposite of those rules.
Now we all know how he creates his music!!
It is not a pattern beforehand or during generation because you cannot predict the next element.