I stumbled upon something that seems like weird behavior… (example attached)
The phrase glitches, mostly in pattern 1 at line 104
As seen in pattern 2, this can be fixed with GFF effects. Tedious but… ok…
Change the tempo from 140bpm to 139bpm and pattern 1 suddenly works.
Is it a possible rounding error with the shuffle algorithm? Maybe the glitch is partly due to the main pattern note not being aligned with the phrase shuffle grid (I don’t think so… ). And it’s still weird that 139bpm works better than 140bpm.
My first suggestion was that could be due to the shuffle grid not aligning to the main pattern grid (kind of).
But I don’t believe that it’s intended. The error is not even reliable. If i set it to 139bpm and loop the pattern say 4 times, sometimes it will play back correctly and sometimes it will glitch. In any case, it should produce the same result anytime that it’s played back.
Best would be if no-instrument-number would perform as reliably as when using Gxx - shuffles or no shuffles. There clearly seems to be some inconsistent magic going on with the current calculations.
IDK I suppose first you would have to know what that ‘phrase shuffle’ percentage routine is internally doing to even start to decide if it is a ‘trigger/shuffle bug’ or that is just how the sound result comes out?
I did this as another little example, switch on the metronome. I mess around with phrase shuffle percentage. Try 0%, 15%, 25% and 87% etc… I’ll let other people draw conclusions.
Thanks. The note-off behavior is expected, so that’s not an issue.
Set the shuffle to 19%, play the pattern and watch the phrase play pos bar. It kinda confirms the first notion I got. Main pattern notes seem to make the shuffle algorithm jump a bit crazy, doesn’t it?
The “shuffle timer”+phrase-pos timer seem to get confused by being reset by a new note. My guess is that the shuffle-timer gets reset and in turn it confuses the main timer some way.
