Some fragments of documentation .. Ben has some nice documentation from our experiments with salad spinning things, since then we’ve been trying to salad spinning a microphone and FM transmitter, and a wireless mic. Not sure if we managed to capture any of that.. Modifying the salad spinner is next for that.
I’ve also been thinking about how the programmes of washing machines could be captured from all those youtube videos, and their inherently musical properties extracted and used to make loud gabba or whatever. One way could be using computer vision, I tried out optical flow detection with opencv:
It seems to do a pretty good job of capturing which way the drum is turning, before losing track as it speeds up. But maybe this could be used in conjunction with machine listening. Here’s a quick pitch tracking algorithm applied to a washing machine, I took one of Nick’s supercollider examples for this:
Your should hear the original sound of the machine on the right, and the detected pitch is reproduced on the left. It seems to detect the increasing periodicity of the sound OK when it is spinning fast, probably with some tweaking this could be used to detect the speed.. Although maybe just using the loudness would be just as good if not better.. Anyway, combining computer vision to detect direction and sound to detect speed seems like it could be fruitful.
So this is promising, although I’m running out of time to improve on this before I have to return home tomorrow. I’m getting preoccupied with spinning Tidal patterns, more on that later..