Wednesday, February 20, 2008

recursive canons

My "Canons" (formerly "Orchestral Canons" -- they don't have new a name yet) are a 4-movement piece I had originally started a very long time ago -- probably the start of my Junior year (2003), but I'm not sure. They're a ridiculously simple idea, which probably accounts for why they're very effective.

The only work I had left on them was orchestrating the final movement, and it looks from my notes like I had a good start on that. But in this case I'm actually glad I didn't finish. In reflection, with the benefit of a much clearer philosophy, I can see that I was screwing them up.


Here's what it sound(s/ed) like:

The orchestral reading sessions were (I assume still are) one of the highlights of UW-Madison's School of Music. Anyone who wants to (granted he knows what he's doing and can commit to the massive amount of work involved) can sign up to have something sight-read (and, time allowing, rehearsed for a few minutes) by one of the school's two orchestras.

Here's the second movement ("Gag"*): UW-Madison Chamber Orchestra -- Reading Session (Bear in mind that this is sight-read.)

It sounds kind of cool and pretty in that Ligeti sort of way, and I was pretty damn pleased at the time. It doesn't sound at all like me, but that's not something I really care that much about -- either it's worth writing or isn't. But the problem is you can't tell what's going on.


Here's what's going on:

Each movement has a very simple theme, or germ or kernel are better words. In the case of "Gag," the kernel is only four notes in two voices:

The shape of the kernel dictates the shape of the piece as a whole. Each note will represent a statement of the theme (or vice versa). First there will be a statement of the kernel at a given transposition and augmentation. (For non music people: transposition is pitch level (how high or low it is), and augmentation is something like duration (how long the notes are).) The second note (in the second voice) begins halfway though the first note, so the second statement will begin halfway though the first statement. The second note is a 4th lower than the first note, so the second statement will be a 4th lower than the first statement. The second note is 3/4 the length of the first note, so all of the notes in the second statement will be 3/4 the length of their counterparts in the first statement. And so on.

You could call this recursive or, at a stretch, fractal. There is such a thing as fractal music, typically rendered by synthesizers and written by computer programs, and this is certainly not that. I've pointed out before the importance of clarity to me. The audience should be able to hear exactly what's happening, and everything that's happening.

In that regard, these pieces are complete failures as I've orchestrated them.

The problem is that I used an orchestra. Blending is something an orchestra, especially the strings, does automatically. It's one of its major strengths. For these to work, every statement needs to be clearly distinct from everything else that's going on. If there are four simultaneous statements of a two-voice kernel (and there are), it should sound like four distinct things happening, not an 8-note chord.

My next step is to figure out what instruments I'll use. The trick here is that most of the movements call for three groups of three instruments, while the second movement calls for four groups of two instruments.

Ideal for 3x3 would be something like 3 woodwinds (flute, clarinet & bassoon), 3 brass (maybe trumpet, horn, tuba), and either piano or 3 strings (maybe violin, viola, cello). Piano has the advantage of not blending as easily as strings, while strings have the advantage of being able to sustain a note for ridiculously long periods of time (important for the piece).

Ideal for 2x4 is harder to pin down - probably 3 pairs of identical instruments plus piano. I keep coming back to piano because of its advantage in being able to be both a 2-voice or 3-voice (or 4-voice) group.

So, how could I make a 2x4 instrumentation out of the suggested 3x3 instrumentation (or something similar to it)? I'm thinking as I type, here. I could use woodwinds, strings and piano for the 3x3, then use (1) high winds, (2) high strings, (3) low wind and low string and (4) piano for 2x4. No, strings blend too clearly - if I use them at all, they have to stay together. I could have a player from each group of 3 sit out for the second movement, and have the piano cover 4 voices. I don't like that, though -- the second movement builds up, and having two players sit out would counter that effect. This is clearly going to take more thinking than I have room for in an already long post.


*The other movements are "Tic," "Sneeze" and, probably, "Song." At the time, I liked to name things after I was done with them. I haven't decided weather or not to keep the titles; they're cute, but don't really add anything.

2 comments:

Anonymous said...

I don't know if I understand all this, but how about more stuff with percussiveness like a mandolin or a lute? :) How about this ensemble: bouzouki, oboe, accordion, cello, flute and theremin. Tuba? No, accordions and oboes sound too similar...

Daniel said...

Accordion would actually work really well. Sadly, the ability to write for accordion is one of those skills that gets overlooked in the land of edumacation.