Hey so Ive always had an easy time multiplying a clock phasor but wanted to divide it. Today I had a flash of insight and made a counter using all signal-rate objects using the same algorithm as the feedback float-and-adder counter used pretty often in PD. This divide patch is super accurate with a high frequency input. Although it expects a phasor at the input I have fed other things in and gotten interesting results.

Anyways it's all documented in case you want to dissect it. Enjoy

http://www.pdpatchrepo.info/hurleur/rate~.zip

that's cool.

i had a quick go in vanilla.

can see how this would be pretty useful for phasor~ based sequencing too.

i got ALMOST everything ok using just vanilla objects, but one issue is that using [rpole~ 1] as an accumulator for the counter, i couldn't figure out a way to make it easily reset when the divided wave also wraps. So, this patch is gonna blow up if you leave it running too long without doing a manual reset with the [clear( command.

i also tried with some more 'interesting' waveforms like phasor~ with FM, and that seemed to work nicely too.

anyway, great idea. the signal-based counter i really like, and can see how it could come in really handy in a lot of projects.

http://www.pdpatchrepo.info/hurleur/phasorDivider.pd

thanks, I dont mess with vanilla too much, but doesn't it still have expr~ and therefore access to a modulo function?

Noticing some bugs using this right now, mostly mis-counts. Let me know if you (plural) find anything.

EDIT: i think I just done understand what your rpole filter is doing.

man, i still have no idea what the story is with expr~. it's LGPL license last i heard, and can't find too much info if that's ok with releasing embedded in apps, etc..

here's the original thread with the [rpole~ 1] idea, thanks maelstorm:

http://puredata.hurleur.com/sujet-6194-sample-played-automated-varying-speed-detect-completion