is there a way to pause mono processing?

stefan · 7 Aug 2006

Are you sure? Selectors seem to work just fine here.

luisi · 7 Aug 2006

ok I try again, <grin>

thanks

dreamkeeper · 7 Aug 2006

Oddson posted an output switch a while ago. It unlinks the outputs if there's no signal present for a period of time and actually does reduce CPU. I was curious how it'd work in conjunction with some FX, so I chained some effects and changed the threshold from 0 to 1e-5 or so.

Well, the switch didn't reduce the CPU load. Only switching the bypass on the individual FX did that, dunno why that is. The threshold I chose was clearly high enough to make the trigger work.

So I just linked the trigger output to the individual bypass switches, et voilà! And this is what you can try if it won't work otherwise: bypass individual sections, triggered globally.

Nevertheless, there might be something fishy. I seem to recall, that it (the original approach) worked fine with one FX but not with several in series. With bigger mono sections, it seems to depend on where the bypass is being made. Funny thing is, it can consume CPU even if the actual connections to the output device are removed, just tried - hmmm...

werner

BTW: here's the modified OSM. (oops, just realized I saved at a point when I had the output connections removed - doesn't make a difference though)

luisi · 8 Aug 2006

many thanks dreamkeeper, I see you faced the same issue,

BTW even the poly section seems to use still a bit of CPU when no voices are used, can you confirm that?

now I check your file

dreamkeeper · 8 Aug 2006

Not quite sure, but it seems reasonable to me that there will be a bit of processing overhead involved with any stream components in the schematic. What does make me wonder however, is that this happens even with disconnected modules - both poly and mono. 😕

Tom7777 · 8 Aug 2006

If your interested heres the original topic (im still waiting for a 0-1 value output module) 😉

0-1 value from stream signal?

I don't know if it's still the same (havent had the chance to check out dreamkeeper's version yet) but the original output switch if used at the very end of a synth after the effects cuts a bit of audio if theres a delay ect so its best to put it before the effects. 🙂 At least thats what happened when i tested using a long reverb and delay hehe 😀 I guess you could adjust the time though, if we can get a stream to 0-1 readout module though we could make a perfect version of this 😀

dreamkeeper · 8 Aug 2006

Tom7777 wrote: ... the original output switch if used at the very end of a synth after the effects cuts a bit of audio if theres a delay ect so its best to put it before the effects.

Not sure what the problem is, I didn't see this behaviour. Leif's original switch triggers 2 seconds after the signal has gone to 0 and stayed there (nifty trick with that array and GMax btw!), so it shouldn't cut out anything, long delay/reverb or not.

The only mod I made to the original trigger module was to check against an adjustable threshold, rather than 0. This was made mainly for test purposes, I didn't want to wait for ages until the FX output reached 0. 😉

Also I don't see a cut off with long sustained notes, like you reported in the other thread. But maybe I misinterpreted that?

werner

dreamkeeper · 8 Aug 2006

Tom7777 wrote: ...but the original output switch if used at the very end of a synth after the effects cuts a bit of audio if theres a delay ect so its best to put it before the effects.

Ah, wait a moment... did you tap the trigger input before the FX? This would explain it, clearly the trigger has to be fed with the very end of the signal chain, otherwise all tails longer than the trigger delay will cut out.

Tom7777 · 8 Aug 2006

Ah, wait a moment... did you tap the trigger input before the FX? This would explain it, clearly the trigger has to be fed with the very end of the signal chain, otherwise all tails longer than the trigger delay will cut out.

Can't remember, i think i just put it after a long custom delay i made, as i remember it was just a timer event and not an actual 0-1 volume correct conversion so i think my delay was just set much longer than the timer or something. If not that reason it probably that i did what you just said 🙂