I’ve been reading Christopher Scoates’s book on Brian Eno’s work as visual artist, Brian Eno: Visual Music and, inspired by a section describing some of Eno’s process in making his ambient works, thought it might be interesting to work with a similar process in making an abstract sound collage. So I did: the result is called “Styrax.”
To construct this, I “seeded” it by using the first ten digits of pi to cull the first ten concrete nouns from the published version of Eno’s 1992 talk on perfumes (reproduced in Scoates). I chose the 3rd concrete noun, then skipped 1 and chose the next, then skipped 4 and chose the next, and so on, until I had a list of ten nouns. I then used those nouns, rather in the manner of Oblique Strategies prompts, to search for videos on YouTube that might provide usable sound samples. (The list of nouns: collection, aroma, wax, leaves, years, names, styrax, root, unrecognizability, experiments.) Then, having found videos with usable sounds and extracting the audio from them, I found the most suitable segments of each one. The sounds ranged from highly organic (field recordings and the like) to entirely electronic: with one exception, no “music” conventionally produced was used. Then I messed around with each sound: slowing it down, speeding it up, stretching it, adding reverb, etc. I then added a period of silence to the end of each sound sample as a proportion of the sample’s length, according to the source’s particular digit of pi.
To organize it into a piece of music, I again used the first ten digits of pi to determine entrance: I delayed the entrance of the second element by a multiple of 3, the second by a further multiple of 1, and so on. I then looped each sequence with no intervening space: since each segment was of different length, at any one time there would always be a new interaction among each of the ten sounds.
Then I cheated and threw all of that out. Well, not quite, but…I muted several sections of some of the segments, because they felt too familiar…or I cut out bits of the silence to avoid having several seconds of actual silence (sometimes, the silent sections of all ten parts aligned), and did a little more sound processing (mostly in terms of adjusting levels and EQ, although I altered one sound each time it occurred because its very distinctiveness meant that its return ended up feeling too familiar otherwise).
A bit more EQ and level-adjustment, and the addition of a ghost, and I was done.