Yeah - Endjinn is my work. Pleased to hear you find it bonkers @analogue01
I literally only looked at the NW granular stuff for the first time yesterday, and I don’t think it does granulation live - just files.
The super easy (and great sounding) external for that is grainstretch~
You can do stuff like grain clouds from live input, or grain pitch randomisation easily. It also does up to 16 overlapping grains (in the [p more grains] bit) which sounds awesome for thick chorus style sounds.
Really easy to use. Just bear in mind that a lot of stuff in additional patchers (eg [p grain_settings]) can be used in streaming mode - they are just not hooked up in the streaming example, so copy them into that patch.
Looking into building my own granular playback patch (again, or should that be ‘finally’?..)… and there sure aren’t very many satisfying tutorials out there. I also dislike most of the forum patches I’ve tried… except for the classic sugarsynth/granular2.5 patches.
Fav new discover: the ola.granular patch from the gen examples. Getting really nice sounds out of it, but i’m having trouble understanding the code in order to kill some artifact problems I’m finding.
fwiw getting artifact-free buffer stuff out of max is just super hard - best to roll with a well-made external (which you could use a multiple of in a specific way) or just keep it edgy. (possibly already the conclusion of this topic hehe)
I think it’s more that there’s like a lot of scenarios that need a crossfade at the right time, and neither max or gen really have any tools to help with that. which is crazy cuz I have to imagine that’s a super common use case. but yea that’s exactly why karma~ was made, but it’s high level. I definitely wish for something like an interpolated peek~ and poke~. I may try making that, actually.
and yea also softcut is great & yea I literally can’t stop brining it up.
Just used your softcut max patches everywhere on a track by the way and it was great. Is softcut suitable for granular synthesis? I seem to remember the granular norns script wasn’t made with softcut but maybe that changed since?
I mean, if you set the loop start and end points short and move them around a longer buffer, mess with the fades, then that’s granular, yea ? if there’s live input maybe you’d want to dedicate a separate recording voice. I guess most proper granular relies on having a lot of playback voices, in norns yr limited to four, but in max you can place as many instances as u want. It’s kinda CPU intensive though when you start adding more voices, I seem to remember karma~ being better there. the petra library above looks like it has more options for the windowing/crossfades/envelope which I assume is useful. it’s fixed in softcut~ right now but I think there are some more curves under the hood. and petra is probably much more CPU friendly with hella polyphony.
ya I think I’m mainly speaking for like one-voice chunky kind of granular (which I enjoy !) for the polyphonic/very small grains stuff petra library is probably gonna serve ya better
To be clear, softcut loops can be arbitrarily short with subsample accuracy. But they lack many typical parameters for granular synthesis and have orher core features that are expensive and irrelevant
You could put softcut in a poly, disable recording, handle all timing and rate changes in max. But timing would be loose (message queue) and you can do that with other stuff anyway
Weird that max doesn’t have something built in like supercollider BufGrain. Granular engine external seems straightforward
Petra package looks cool. But, already I wish some of the params (pitch) could be more extreme.
Yes. This. I was thinking this early, pondering if I shoudl learn some supercollider just for BufGrain and TGrains. But there’s only so much time in the day…
I’m actually surprised they didn’t make munger~ a standard object one or two versions ago.