Monday, July 7, 2014

A Glance at Spectral Music

There is a recent school of music composition that I don't think I have mentioned on this blog. Originating in France it is called "spectralism" and involves analyzing the overtone structure of musical notes. In other words, the timbre of a musical sound is regarded as structure. I wrote that and I'm not sure what it means! Maybe I just don't have the math. I did know one composer in Montreal who worked with some of these methods so perhaps I can offer that experience as the Wikipedia article seems too full of generalizations with not enough specifics to really give a clear idea.

This composer wrote a piece for percussion. Percussion instruments have particularly irregular and complex overtones and he analyzed their structure somehow and used this to mathematically derive a series of rhythmic values. I suppose you might think of it as making the vertical horizontal. In any case, analysis of the overtones gave him the rhythmic structure of the piece. My comment to him was: "that is the syntax of the piece; what is the semantic?" In other words, is there any actual expression?

The same question comes to mind as I listen to a piece by the spectralist Tristan Murail. Tom Service at the Guardian has an article up introducing spectral composition on the occasion of a festival of this music at the Aldeburgh Festival. Right now I am listening to this piece from the article:


The same question comes to mind. What is the aesthetic expression of this music? As with my composer friend, there is a description of the structuring of the music. Here is an excerpt from the Wikipedia article:
The composition of spectral music is concerned with timbral structures, especially when decisions about timbre are mathematically informed by Fourier analysis using the computer-efficient fast Fourier transform (FFT). FFTs can be run to provide graphs that illustrate details about the timbral structure of a sound which might not be initially apparent to the ear. Also, when creating sounds with computers, FFTs can be used to transform the timbre of a sound in various ways, such as by generating hybrid timbres through a collection of processes known as cross-synthesis, or applying a room reverberation to a sound by means of convolution. If the music is to be performed by live musicians (as opposed to being played electronically via computer through speakers), then these novel effects must be translated into an extended traditional notation that can be read and executed by a human being with some additional training.
I like to be able to figure out what is going on myself in a piece, but since I don't have the math, I will just have to take all this on faith! Let's assume that this is exactly how the composer works out the structure. But the important thing is how it sounds to the listener and perhaps the kind of math used is no more relevant to the final product than the kind of pencil used or the brand of music software used to make the score. The bottom line is how it sounds.

After listening to the piece I embedded above, also found in Tom's article, I have to say I don't find much of interest in the music. At the risk of seeming flip, I have to say it sounds like someone crunched a lot of numbers and this is what it sounds like. In other words, I don't hear much of aesthetic interest.

By way of contrast, let's take the example of a symphony by Pettersson. Apart from noticing some obvious structural features, such as concentration on a certain interval or rhythmic figure, parallel perhaps to the concentration in the Murail piece on the note 'C', I really don't know anything about how Pettersson arrived at his structure either.


But here is the point: with the Pettersson symphony I hear a great deal of aesthetic expression: the music "speaks", if you will. I just don't hear that with the Murail. Tom describes the music like this:
 defining and refining a language whose sensuality and vividness feels like it can create a whole world of feeling, from the awesome power of natural phenomena to intimate, human passion and delicacy.
The only problem I have with Tom's criticism is that he doesn't realize he is a critic! He seems to regard himself as an advocate of everything he writes about: "here, listen to this fantastic music!" Where I hear sterile noodling, he hears sensuality and vividness. Or is it just that he thinks he has to say he hears sensuality and vividness? Search me.

But if anyone has a better understanding of what the spectralists are doing, please weigh in in the comments.

3 comments:

Christopher Culver said...

It’s a real pity that you came across these references to "Fourier transforms" and so forth, and that this made you form a preconception that spectral music is "mathematical". While spectralists have often used advanced technology, it’s to save work in realizing a sort of sound that they wanted to achieve from the start. Spectralism after all goes back to the late 1960s before anyone of these computers could tinker with computers. Indeed, founding father Gérard Grisey saw his work as standing very much in opposition to "mathematical" serialist and post-serialist trends of the time:

"We are musicians and our model is sound not literature, sound not mathematics, sound not theatre, visual arts, quantum physics, geology, astrology or acupuncture".

Grisey saw spectralism as an aesthetic concerned with one's perception of time. His pieces often speed up or slow down sounds to reveal a sort of inner life. Take, for example, his "Vortex Temporum" which kaleidoscopically presents a quotation from Ravel in innumerable guises (fast, slow, sine wave, square wave, with this or that formant louder than the others), or "Partiels" that assigns the same sound as a trombone timbre to various instruments of an orchestra but vastly slower.

Murail, on the other hand, is a more conventionally "France" composer very much working in the vein of Debussy, Ravel, and Messiaen ("Gondwana" quotes Messiaen's Turangalila Symphony). He is concerned with mood and atmosphere instead of Romantic pathos. Why you would expect everyone to sound like Petersson, who represented only one particular scene within classical music -- the Germanic/Nordic symphonic tradition -- is beyond me.

Bryan Townsend said...

For some inexplicable reason, Blogger is refusing to post a comment left on this post, but which I received in my email. So I will put it below, along with my response.

from Christopher Culver:
"It’s a real pity that you came across these references to "Fourier transforms" and so forth, and that this made you form a preconception that spectral music is "mathematical". While spectralists have often used advanced technology, it’s to save work in realizing a sort of sound that they wanted to achieve from the start. Spectralism after all goes back to the late 1960s before anyone of these computers could tinker with computers. Indeed, founding father Gérard Grisey saw his work as standing very much in opposition to "mathematical" serialist and post-serialist trends of the time:

"We are musicians and our model is sound not literature, sound not mathematics, sound not theatre, visual arts, quantum physics, geology, astrology or acupuncture".

Grisey saw spectralism as an aesthetic concerned with one's perception of time. His pieces often speed up or slow down sounds to reveal a sort of inner life. Take, for example, his "Vortex Temporum" which kaleidoscopically presents a quotation from Ravel in innumerable guises (fast, slow, sine wave, square wave, with this or that formant louder than the others), or "Partiels" that assigns the same sound as a trombone timbre to various instruments of an orchestra but vastly slower.

Murail, on the other hand, is a more conventionally "France" composer very much working in the vein of Debussy, Ravel, and Messiaen ("Gondwana" quotes Messiaen's Turangalila Symphony). He is concerned with mood and atmosphere instead of Romantic pathos. Why you would expect everyone to sound like Petersson, who represented only one particular scene within classical music -- the Germanic/Nordic symphonic tradition -- is beyond me."

Christopher, usually your comments are very fair and informative, but this one seems not to meet that standard. First of all I "came across" those references to Fourier transforms because they formed the core of the discussion of spectralism in the Wikipedia article. You make it sound as if I dug it up somewhere obscure. Second, as was perfectly clear in the post, my sense of how these folks work was derived from an acquaintance in Montreal who used these methods. Not a "preconception" but a simple observation.

Why I chose the counter-example of Pettersson was also make perfectly clear in the post: it was the most radically different approach to composition I could think of at the time. I certainly don't expect everyone to sound like Pettersson--in fact, I don't expect anyone to. Really, has the internet caused everyone to lose sight of how to further a discussion? If you want to attack straw men, do so elsewhere!

Rickard Dahl said...

Well, the piece was more listenable than many other extreme-modernist pieces. However, just like with most extreme-modernism, it lacks good aesthetic expression.