[sdiy] (OT?) FEEDBACK as a topic in electronics, music, science, etc?? BOOKS?

[Scott Nordlund]

> According to my (limited) understanding and research into video feedback and video synthesis, the actual feedback loop operates at lightspeed, and it is the limited rate of change and persistence-of-glowiness of the phosphor coating on the inside of the CRT that slows the evolving patterns down to a visually perceptible speed. Lcd/tft monitors operate in a different manner, and give different effects. Tube based cameras (Ie ones that use vidicons, trinicons etc) provide different results to ccd sensors.
> Ive been building a large analogue video modular for the last couple of years (time flies when youre having fun!) and the different types of feedback texture are fascinating to watch and play with. Direct feedback paths (I e out of the output mixer and patched back in to a spare input) give completely different effects from feedback patches involving cameras and monitors. Combining the two approaches is even more fun. Im currently working on a way of micing up the picture tube to capture the buzzing tones from the LOPT transformer on the top of the picture tube (which audibly changes, depending on the picture being displayed) and feed them back into the system. It's a lot of fun. Further research directions will be involving video delay networks and filters patched into the feedback paths. Im guessing that even milli or micro second delays will be enough to produce useable/interesting results due to the frequencies involved.
> Cheers
> rich

Well, there's some latency because each pixel is scanned only once per frame. And tube detectors have that weird sort of nonlinear halation thing, I suspect that had a lot to do with the specific look of the original Doctor Who opening. I've noticed that improperly mixing two non-synchonized video signals can result in all sorts of interesting "rolling" effects as the synchronization signals fight against each other. It's not exactly electronically kosher, but I wonder if you could mix feedback signals from independent sources just using a pot, for an extra layer of unpredictability. I've sort of intended to get into video stuff for years (the low level, hands-on, intuitive flavor), but never really found the time or resources.  I've seen interesting results on youtube from circuitbending a digital frame buffer. I think Gijs Gieskes did it. I'd like to see what that does with feedback...