I think this is a sampling problem and not an octatrack problem… zero crossing sometimes is just the best crossing, and not necessarily zero… If you wanted, you should go with the shortening of the samples, and try that LFO trick…
And I think having a pad, or a bed of white noise… or a drone on the tonic of the song your in… or room tone… all of these are a great way to smooth over some irregularities… and it doesnt have to be loud… you can have it faint, and it will still do the job.
Oh, I agree. But it is an Octa problem in the sense that the live aspect of the sampling sets additional constraints on the situation.
For example, on the MPC to get around this problem when slicing breaks, I usually just bump up the BPM a bit after slicing. Can’t really do that here.
Manual page 54 following. FIN & FOUT can have tiny values but they apply right after start and after end of RLEN i guess. Those are values also hold in 4 parts. So you can have different values according to machine-track / rec-trigs design you choose. Means if you would make use of FadeOut in recording of lets say 16 steps and want the fadeout to be part of the 16steps, you would need to choose RLEN 15steps + the according fadeout shorter then up to 1.
Another hint is that drastically lowering the HOLD param and instead increasing the REL might help.
For a sampling track I usually dedicate an fx slot to the parametric EQ, keep it in PEQ mode, set Q to maximum, boost the gain all the way and sweep the frequency to find the clickiest setting (much like you would in a DAW to hunt down troublesome frequencies). Once I find the clicky frequency I turn the gain all the way down. I repeat this for the second EQ as well and offset the frequencies a bit. Often I end up with the frequency between 108 - 111. This obviously changes the tone of your sample, and is not a perfect solution, but with right source material, it can be usable.
Was able to partially address the clicks while minimizing gaps by setting flex length to slice, then setting the length parameter to 2. This effectively plays back two contiguous slices back to back with no click in between.
Then I set the amp attack to 10, and the hold to 4 (I have my linear locks every 8th note). This results in a minor gap, but with a groove and a melody or pad playing, it’s not super noticeable.
Another thing to play with , amp page 2, turn sync settings OFF.
This gives you much finer control over the amp settings, as by default the attack hold and release are tempo synched. (Which is dumb in my opinion)
I feel like sync is actually helping me in this case, but I’ll try to see if I can get it more precise with it disabled.
This would require different attack/hold/release parameters for different BPMs no?
Possibly.
I prefer it off. Erm. YMMV or whatever
OT is a clicky recorder.
Fade in an Fade out sucks.
I also think Amp is (unfortunately) the best/simplest way to play sliced recordings directly if you want to avoid clicks.
Amp set to 1 can solve attack problems easily. Set Hold below desired lenth, Release by hear.
Then you can hear a little gap/cut. Can’t work on a continuous sound obviously.
For granular stuff, I tried many things, including lfo designer, imho it is better to forgive or forget if you want experimental live sampling. I’d be glad to read something working “clickylessly”.
Add hihats on clicky steps is simple. 
I even tried a dedicated track for resampling with an envelope, resampling it. Wasn’t without clicks.
I prefer Atk 1 to FIN, and FOUT adds recording time, which totally sucks for loops.
Check Elektronauts, FIN / FOUT sucks imho, eventually with minimum values and Pickups, which have many flaws too. Imho FOUT is useful only with Pickups overdub.
Yeah Fin / Out doesn’t help me because it doesn’t do it per slice.
FIN can help at the beginning of the recording, and you can place all slices at the beginning of the recording (different lengths, tuning purpose).
Unfortunately FOUT is added, it changes recording length, except with Pickups.
The goal here is to live sample something from the inputs and then apply a slice grid to it, so most of the slices would not be at the beginning
Ah yes sorry, I got confused with a topic about granular stuff, direct recording.
So you shouldn’t have clics with good AMP settings. Just very little gaps but it is necessary if you want to modulate Slices.
Hold values are steps. It has to be shorter than Slice duration. Beware of long Attack, it can increase Hold time. Adjust release by ear.
I prefer not aligning markers to zero crossings. It changes slice length, and imply clics at the end with regular linear locks.
If you add pitch up, Slice duration is shortened.
PIPO loop can help in this case.
With trigs every 8th, Hold should be just below 2 (step values). 1.975 or something.
Try with Atk 1, Rel 1.
Don’t use that for that purpose.
You get unequal slice lengths, shifted start/end
Ah, good point about PIPO loop. That alone could help a ton, what a cool feature.
I tested with 2 tracks alternating random slices just after recording. Works really well. Overlap, so no gap, and no clicks.
Test recording a classic radio on the fly.
So @Isoparmesan what do you think about using 2 tracks with alternated slices ? And eventually skip the Neighbor track to keep the sale number of tracks…
With only 1 track, the best results I had was with ATK 1 and a release longer than normal playback using PIPO. Wasn’t perfect of course.
You can eventually post here some of your recordings so I can give a try with your audio…
I already have two tracks, but they’re dedicated to separate inputs, so it wouldn’t really work. I think I’m going to be able to do what I want well enough with the ENV and Flex settings I landed on.
my project is based on EZBot’s techno performance template, and I’ve already removed one set of neighbor tracks, I need at least one per thru track for the FX scenes I have set up.
An audio example and settings would be interesting.
I will try and get around to recording one.
