thanks to Andrea Valle and the organizing team SMC Torino happened in an online format at the end of June. The proceedings contain an article about the single sample feedback class Fb1.
Matters 5 was included in the online edition of xCoAx 2020. There’s an accompanying description and a rather symbolic “making of” video – the piece in full length can be accessed in binaural format.
Several people active in the SC community were involved in both events: Alberto De Campo, Ron Kuivila, Alberto Novello, Pierre Alexandre Tremblay, sorry if I missed others.
Right, thanks for the reminder, the Fb1 paper is now available here:
the related follow-up on Fb1_ODE from 2021:
Concerning performance, due to their implementation, synthdefs can have a lot of ugens, but that doesn’t necessarily mean a high CPU usage. I didn’t make comparisons with other environments, enabling convenient single-sample feedback in SC was the main goal. See the section in the paper(s) for optimization techniques.
Didn’t see in description your process for composing for so many speakers. Did you finish the piece in a smaller setup then adjusted on the spot or you had a space where it was possible to experiment with the full setup?
I always start with synthesis/processing in stereo to keep things simple at first (well, kind of, a labyrinth already … :). Then I look how to expand the synthesis to more speakers, occasionally I also distribute stereo with spat algos, but this was not the case here. Eventually, I checked with 8 speakers and then binaural.