Crosstalk between MI "lmnts/reverb" & "rngs/chorus"

#1

today i observed some kind of strange crosstalk between two of the MI models. i got some audio from the lmnts/reverb even if nothing was connected to its inputs, as long as the rngs/chorus had an input signal. i recreated the problem in a minimal setup and observed that this crosstalk only occurred if the chorus was placed after the reverb in execution order.

@thetechnobear: Maybe there are other crosstalk problems beside this special case?

I really like the sound of both modules and even the crosstalk has some interesting characteristics. I just wanted to report this behavior.

2.jpg1330x614 85 KB
1.jpg1330x614 80.6 KB

#2

Thanks , coincidently I discovered the same thing yesterday between brds and lmnts.
It's going to either be a memory overrun or static buffer issue .
I'd guesss it's the latter probably as a result of the fact real MI modules only exist in isolation from one another . So yes , I think it's isolated to MI modules.

I'm refactoring all the MI code for the next release , so will investigate further during that.

#3

I have also noticed that. That when for example loading 2 reverbs in the same patch it sounds very crazy.

#4

I've noticed the same thing between rngs/ensemble and rngs/reverb.

I had an input going into rngs/ensemble. I cut the connection between that and rngs/reverb, which was the final object in my patch, so the reverb had no input at all, and I could still hear the input to ensemble coming out the reverb.

Are they sharing a buffer or something?