Here are some more thoughts I’m having as I compose my piece. I am starting to explore different medium- to long-time-scale gestures – that is, elements of the piece that arise from the combination of many different processed sound samples.
The important gestural elements seem to be:
Although the temporal and spectral distortion effects are properties of the frequency quantization effect I’m using (as I have discussed), they contribute to an emergent gestural character when applied with different parameters to repetitions of the same input audio samples. This combines quite well with the way I am spatializing the piece (which I discussed last time). The “loudspeaker orchestra” type spatialization allows the listener to localize each sound to a particular speaker. With these gestures distributed between multiple speakers, it gives the impression of a chorus of speakers with different characteristics.
I should note, of course, that my spatial diffusion is not technically in line with the loudspeaker orchestras of, say the BEAST of Birmingham. Such performances usually involve stereo audio playback, and the dynamic effects stem from orchestrating the volumes of the different speakers.
I have been pretty happy with some simple, clean electric guitar recordings. I broke these into about six phrases and processed them with various window lengths and frequency-domain bit depths.
At the moment, I’m looking for audio with more percussive low frequency content, as well as more spectral distortion as opposed to temporal. The guitar audio has yielded very melodic mid-range content, but I’d like to move to a more atonal and bass-heavy section.
I’m considering Powers of Two as a name for this composition. Seeing as powers of two arise in the MDCT window lengths, the bit depth (bits are of course powers of two), etc., it seems fitting. I’m still thinking about it, though.