Ive spent the past 1 year sampling shorter bits from my gear in chains, thats all worked out well.
I am having trouble creating a longer sound in this case its a jp8 string sound from the jdxi.
My mind tells me that if the loop is sampled right it can be relatively small in length lets say a bar but i can loop it and you wount be able to tell its looping and itll create a longer continually looping sound.
When the sound plays when it gets back to the beginning it sound like it has more amplitude. ive tried the amp section and changeing the attack, release and hold but not much joy there.
Ive was looking at a youtube page where the guy copied tail to front and then cross faded but is there a simpler way using the OT itself.
I am sure its just my inexperience at either creating the loop or inexperience post sampling using the sample.
ive been using the ot to send the midi note to the jdxi and then add the trig to start the sampling
If you want to capture samples that loop smoothly, you need to set the loop on the “sustain part” or the sound. Usually sounds have some kind of initial part that should be left out from the loop.
Is this what you meant?
look for the small squares in the OT sample editor when setting for smooth loops, they indicate the zero-crossing points of the samples. You can see these in the sample editor when you zoom in down enough.
I thought a recent chat along these lines may have shed some light on this, similar to reply above - explore the possibility to define a looping section after the (once only) attack phase (of the sample) - then create a release phase (of the sustain part) using the AMP release
Ive found the zero crossing further in at the beginning as you described after the attack and ive moved the end point a little too, in the trim section so its at its minimum once set does the trim stay like that? or should i delete those sections? i assume now i go out of the editor and use the amp part. ive noticed now ive trimmed it down, that it plays for the length of the loop then resets for the remainder. my pattern is 4 bars ive set loop on, ive clearly missed a setting i think
No need to trim as long as you’re not short on RAM. Make a destructive trim only after you’ve found the perfect loop points and have already played with it for some time.
its very close i have the constant sustain part, but i have a small click. moving the release and attack help a little hold is on full. is it only down to those settings now?
I wonder if i am sampling wrong. will it be better if i manually start sampling and allow a tail to form. that will deal with the end and leave a tail. can i tell ot to fade in I am sure i saw that somewhere, I read it but forgot where. if i do that in theory I should not get the clicks. that seems like deadlock. if i move in to get to the sustain part i have the wall of sound going in when the advice is to fade in and out to rid clicks.
Yeah, of course the polarities must match too. If you combine a raising end to a falling start, it will not sound correct.
AFAIK, getting the loop points to match and line up fixes clicking most of the time. A perfect end result might require several attempts from different points / cycles of the waves. This is also why I’d advise against destructive sample trimming until you’re sure you got something that’s working.
your exactly right, I haven’t eliminated it entirely but its very close. I am going to check the rise and fall too, that’s a great piece of advice. It has taught me to that the length of the sample doesn’t have to be too long just long enough for that sustain part, because even with length it could still click.
oh god yes, very good point! Stereo samples can be very tricky to get looping properly. If there’s not much stereo trickery going on with the sound, you’re always better off sampling mono, or converting to mono (discarding one side of the stereo information).
I stuck with this til about 1900, then as I could still here the albeit small click at the beginning tried mono and got it even less to click but still evident when just the one sound was playing, I tried a new idea. Its either my crappy editing or not being able to find the loop points in the sustain part, I tried a lot believe me. I hooked up my rc505 set that to record 1 bar and stop with a mono channel. I pressed the key on the JDXI before recording so I was already in the sustain part. I had not set fade in or out. I got a seamless loop.
Today I will shove that loop into the OT and see if its seamless as is in that.If it is then I will try doing the OT loop manually and see what happens
Interestingly the RC505 made a great sounding loop where no clicks when played back on the RC505. When I brought the loop over to the OT back was the click. This told me that the OT is just as good as the RC505 and it was my technique that was the issue.
So what i did then was set the JDXI to play the note I wanted sampled but I set the AP to on and set the hold to on so a continuous sound was being played, this gets back to i think what Tsutek said about going beyond the attack part of the sound.
I used the trig recorder sampled at 4 bars and hey presto a sample that played back perfectly in loop mode.
I know there are lots of threads on this going back through the years but on the whole its not all about what you do in post, trying to get zero crossings or looking for rise and fall at the beginning and end.
The success here was to sample correctly in the first place ie not to trigger the synth with a midi trig at the start and record at the start.
I have an rc505 as well and I think it
always applies a tiny fade at loop start and end points, so you don’t hear the click. these might be applied dynamically, so when you copy over the loop you will hear the click
and Papernoise that is exactly what happened, so back to square one. unless i record that playing rc505 loop into the OT. i will try that, it seems every turn i make i am back to square one. I did try that on the OT with the fade but it didnt help. the string waveform is just very difficult to loop.