no i have to check shitons of addons i messed all those years! looks like there is some really neat suff in there
yeah you are right its prob descent enough for the repaint.
im also testing something a bit different atm, (a custom slicer thing module sort of)
but that does quite a few wavforms array/path redraws every 5ms i was afraid of perf;
and i was quite impressed how usine handle it very well finally,
some wip , here it pr loads files, but i think theorically such circular buffer rec would allow kinda same on live reced stuff
https://www.youtube.com/watch?v=LbQ5fgNOkV4
will check all those addons yeah will prob trig ideas
Statistics: Posted by 23fx23 — 06 Jan 2024, 20:24
]]>
for shift i think i see what you mean yeah, cpu heavy tho
for the display part, you'll need to recalculate 700 points out of the whole buffer which is ok
did you have a go with the user module "looper Flite V3" ?
has multiple playhead but only for live sampling (no file loading)
Cheers
Olivar
Statistics: Posted by oli_lab — 06 Jan 2024, 18:02
]]>
one other simple application exemple that was my goal here is bouncing a wav file of what we currently hear, with same exact timming, so no need to mess with refinding real start later on.
maybe if man can control custom record pos that would be even better allowing various setups
for shift i think i see what you mean yeah, cpu heavy tho hehe if have to shift potentially heavy arrays each bloc, faced that challenge in a synced scope ive build, i faked a scroll using a different method:
think some trick could have similar result while lightweight , internally considering a 'relative' to write position offseted read position instead (like 0 to minus 100%) rather than an absolute
something like read pos = (write_pos + ringsize_size - wanted_delay) modulo ring_size , that also let man request anything always past to write pos but with a light load
but yeah, thks for confirming my thought^^ upgrade could be nice indeed ;p
Statistics: Posted by 23fx23 — 06 Jan 2024, 16:07
]]>
In your case, we probably need a sampler with multiple playheads (that read the same recorded sample)
On the circular function, I think we also need that the recording head stay in a fixed position (ie at the end of the sample) and that the whole sample is shifted while recording.
That will allow to use the samplers as far fetched delays (for example granular delays...) . Because a fixed record head will allow to shuffle inside the sample without meeting the "stitch" left at the recording head.
Will be a great improvement in the sampling department!
Cheers
Statistics: Posted by oli_lab — 06 Jan 2024, 13:33
]]>
what id like to achive:
sampler A play a loop file , say on bar
sampler B record at exact same sample/ppq position, within same duration, ie bar. (so for me a circular buffer ?)
so if setup is ok, that means if then i would crossfade A and B, both playing, they would be perfectly time aligned, (mean phase cancelation if one is polarity inverted)
for now i always get unsyc results, curious what im doing wrong..
any idea how man would patch such thing, if some usiner know by any chance?
i managed to have a perfect cancel, but doing it in a pretty *wild* way,
(writing/getting to arrays each bloc, but i need 4 bars, so that makes quite huges 300k+ values arrays lol .. so feel that's def not descent way)
Statistics: Posted by 23fx23 — 03 Jan 2024, 22:57
]]>