Guess what? I’m sick again. I get the feeling that my body is betraying me sometimes.

In between fits of coughing and bouts of medicine-induced delirium (this is an exaggeration I assure you) I’m doing what I can to retain any semblance of humanity. I dragged myself out for a meeting this morning, and made a remarkably tasty batch of masoor lentils for dinner. The smell would certainly have been warming my heart throughout the day as it bubbled, if only my nose worked.

On the way home from the meeting, I had one of those brilliant moments where the structure of the piece I’m working on just spread itself out on the mental table and said “here, this is how it is!” When I got home I successfully scribbled it all out before something knocked my brain for a loop, so I’ve now got a road map. The piece, entitled Langue, and scored for flute and video, is about how things interconnect to create the potential for meaning. Kind of like a context-free grammar, but with a lot less math and more flute.

The artistic problem for me is how to meaningfully relate the flute and the video. So much “multimedia” art feels feckless and lost to me, like two different pieces duct-taped together. What does the video offer the flute — what does the flute offer the video? Can there be a meaningful interplay to the two, a real coming together? This is what I hope to address.

Technically it’s going to be different. Unlike Dramatis Personae, where I used Jitter to process mostly live video, Langue uses as source material a large number of photographs, which I am still amassing. There’s probably going to be some pre-rendered video as well as some light filters, which ought to save some precious CPU cycles (real-time video rendering does bad things to a computer).

Input is going to be mostly via audio. I started with a fragment of the Jitter tutorial code and ended up with something that can read in peak amplitude, assign a number to it, and have something happen depending on that number. This is pretty neat. Here’s an example, excerpted from the rough draft:

When this passage is played, there will be three loud pitches nestled among the quieter ones. These three will be picked up an assigned numbers; let’s say that on a scale of 0 to 5, with 5 being the loudest, these three translate to 4, 5, and 4. I set it up so that something happens when a 5 is picked up, something different happens with a 4. So as the flute plays, three different events are triggered on the video screen. These might not be big things, but things that build ever so slightly on the event event. Of course there’s no way I can guarantee a 4,5, and a 4. On the day, it might be 4,5,5, or 4,4,5, or a 4,3,3 or so forth. But I don’t want to guarantee it. Musicians are not machines, and writing for them like they are takes all the fun out of it (for me and for them).

Given this, setting up some sort of structure and cohesion becomes a lot more like landscape design than architecture. Set up the paths, tend the plants, make sure that a left turn is taken about here, and voila! That’s what I’m striving for. Over the course of the piece, the video gradually responds differently to the flute, and eventually the flute starts to respond to the video. And then what? Only my scribble of paper knows.