> > > It's proprietary slub technology. > > :-)
> No it's here: > https://gitorious.org/flotsam/flotsam/trees/master/livenoisetools/syncup > It's quite simple, but still needs explanation - if you want to use it. > > Depends. If it would allow me to sync Fluxa to PD with relative ease then yes; I probably would. > You got to the core of the problem with time in fluxus - it needs to be > sorted out (from spawn-timed-task to fluxa, frisbee and beyond) there > are lots of experimental approaches and they desperately need reining > in. The thing could well be the driver of this would be adding netclock > support. > > Yes, I noticed some weirdness. At least spawn-timed-task conforms to Fluxa's perspective on time so that makes things a bit easier. I don't see the difference between "time" and "time-now" being a huge issue here, I'm more concerned with the relatively low resolution of basing such things on the fps and the lack of time-stamps in the midi clock. I think that the amount of jitter I'd get if I wanted to calculate the timing of the next few Fluxa clock pulses based on our midi clock from within every-frame would be a issue. Could you give me a explanation on syncup, assuming we could make it slave Fluxus&Fluxa to PD? The other way around isn't a real option as clock jitter would cause glitches in Rob's looped-buffer playback. Thanks. Kas.
