I call it generative asemic poetry. It is a direct text-based representation of the glyphs seen in the video (and yes, those glyphs *do* work together to conjure the sounds heard in my live coding setup).
Each stanza corresponds to a row. Each word corresponds to a glyph in the row. The syllables in the word are used to encode the actual pattern of the glyph.
The syllables used for encoding are organized physiologically: pa, fa, thu, de, yu, go, ha. The sounds start in the front of the mouth and work backwards.
Only 7 syllables are needed to encode 3-bits, because 0 is used as a separator in the bitrune parser.
With any luck, these poems should work on screen readers. Gibberish, but patterned gibberish that faithfully represents the symbols used to conjure the music.
