Music For Shuffle

Sketch #20

I keep banging on about attention-adaptive music, so I thought I’d better try and make some. I sketched out a little tune with a squelchy FM drone, a piano, guitar and drums, then chopped up the parts in a few different ways – as a whole track, as a set of phrases, and as individual notes.

Then I added a soupçon of QC/Pd to play around with everything. The tabs at top right rearrange the music in real-time, letting you vary how repetitive the patterns in it are. Hopefully this helps affect the levels of attention it demands of you. That’s the idea, anyway.

Hit Linear mode, and the parts get sorted into a standard, sequential arrangement. Verse, chorus, rinse and repeat. There’s a melody, harmony underneath it and so on, and the timeline looks pretty normal:

Modular mode breaks up the main phrases, and puts them in branches. Fragments of the melody and harmony drift in and out in tempo, and the timeline folds out to a grid to show that there are several directions the music could go in at any moment:

Free mode disconnects everything: phrases, notes, chords, tempo, meter – the lot. Everything just floats around in a delicate, Eno-ish way. The timeline breaks apart completely, and the playhead just bobs around at random, playing notes rather than phrases:

The tabs don’t switch modes straight away when you click them – I’ve made the player keep track of volume and BPM, so they only switch state ‘in tempo’ during moments of low volume. There are loads of these cueing-style knobs and switches in music production software – what happens when we use them in a consumption context? What would happen if the system could infer how much attention it was being given?

This sketch mainly came about from remembering Jonty chatting about listening to the stems/tracks ripped from Radiohead’s Polyfauna app. There’s loads of it on YouTube – presumably plenty of people want to ‘experience’ Polyfauna in a low-attention way, without having to do so much work messing around with the app. It’d be nice if the music itself had that capability.

Anyway. A tiny step. Some chewy bits here though, I think.