More fun with the 'sstart cord.

This is where to post your questions and tips on creating presets and using cords.

Moderators: stu, Ole

More fun with the 'sstart cord.

Postby sampleandhold » Thu Feb 19, 2004 8:48 am

I played around with this idea awhile ago, it was more of a theory than anything. But now I have found that my theory was correct... to a degree.

I am working on this choon where I sampled part of a movie's theme. some piano bit or synth crystal thing, something like that. However this sample has timing "errors". The sample itself is 16 beats long. So here I am sitting here thinking to myself "how am I going to get the timing tighter on this sample?" I could chop it up, but that might take awhile and I am tired and feeling lazy, plus I have had some nasty times trying to chop up samples of this nature. So instead I decided to see what I could do with the 'sstart cord.

I set up my cord like this: modwhl>'sstart>100% *yawn*

what I found with this chord is, not only could I do that strange timestretching bit, but by playing the note that trigger's that 16 beat sample 16 times, and setting the modwheel controller at intervals of 8 (16/127=7.98somethingsomethingsomething) I found that I could trigger the quarter notes to play each beat of the sample. So by doing this I was able to have the sample "snap to grid" in a sort of way. Now the 8th beat, the one that really seemed off now is on.

but wait, there is more....

I run eos 4.01 and as such, have no real time Q control. But with this technique I can now modulate the Q (in a step kind of way) of a long sample with no problem. At quarter note intervals, the modulation sound about as smooth as the lpf. And why stop at resequencing the sample in it's original pattern. one could, if they chose to, rearrange a sample just like a chopped up break. and what is more, you can take the sample and change it's pitch like a chopped up break as well with no real loss in timing, to a degree of course. I took my 16 beat sample and found i could go up as high as seven half steps before I started to hear the start of the following quarter note. But the starting quarter notes stayed consistant at any pitch. kind of like a chopped break.

Now for the negative side effects...

This cord's usefulness is limited by the length of the sample. The longest length one could have is a sample with 127 quarter notes, but of course this only allows a resolution of up to quarter note. there is no way to make this have a finer resolution. So the longest sample I would use with this cord would most likely be 32 beat... maybe. quarter notes would be 4, 8th's 2 and 16th's 1. 16 beat length is more useful given that you can fine tune down to 32nds, and 8 and 4 beat lengths are the most useful. A one measure break could be easily "chopped" with this cord.

I hope this may help some of you speed up your production, perhaps make "quantizing" a sample more accessible. This will by no means replace chopping up a sample, but I find that I can get quite alot done easier and faster. I typically use this cord for timestretching and vocal timing, and now for instrumental solo chopping. oh and also I can modulate my Q...
"{jU$t-n3Rv0U$-N-+h3-@Ll3y-W@y}"
sampleandhold
 
Posts: 581
Joined: Tue Apr 15, 2003 4:38 am

Postby illinformed » Thu Feb 19, 2004 3:37 pm

Thats a good one, I have midi templates of set up with the same sort of thing from 4 beat division to 128 division of a bar, it took a while to program but the results are well worth it, for some reason I always use contrller E (midi CC25) for 'sstart.

One thing I have noticed by doing this sort of thing with sstart is that I've found out it's not 100% accurate. A while ago I did some experiments using controller to quantize then sample start, I could never get it working accurately - a shame because it would have been ideal having you controller step through the sample. It works ok on shortish samples but the bigger they get, the more it goes out. I think it's just a general thing with sstart in cords to be honest. The interesting thing is that the 'sample offset' field 'voice modifiers' is sample accurate, no matter what the size of the sample is. This makes 'sstart and 'sample offset' 2 different things although they effectively do the same job. One workaround is that 'sample offset' can be programmed via sysex but this isn't really ideal.

Andrew
illinformed
 
Posts: 259
Joined: Thu Sep 11, 2003 2:12 pm
Location: London

Re: More fun with the 'sstart cord.

Postby DJFracture » Tue Jun 15, 2004 3:28 pm

sampleandhold wrote:what I found with this chord is, not only could I do that strange timestretching bit,


hmmm, sounds interesting. can you explain?

like creating a drill effect by drawing in midi notes at like 64th intervals and then modulating the sstart as the drill goes on? therefore the last midi note is playing the end of the sample and the first midi note, the beginning of the sample?
DJFracture
 
Posts: 11
Joined: Mon Feb 09, 2004 1:07 pm
Location: London

Postby sampleandhold » Tue Jun 15, 2004 5:06 pm

Exactly, like an old timestretch algo from the akia samplers. You can control length (the midi duration) and how fast long you want the sample to be (drawing a cc value over 150 measures, if one wanted)

sometimes it benefits the sample if you get nasty artifacts to add a decay on the adsr. when you do the time stretching thing the samples will bleed into each other and make the effect sound smoother. Just depends on the sound you want. it is a sort of real time timestretch program you can say. it comes close to the old timestretching effects heared in old rave and drum and bass choons.

then of course you can run the sample backwards to a degree. sample start is a really useful cord, sometimes when i am lazy i use it instead of chopping a break or something up, but like ill said you have to have a small sample or the accuracy goes out the window.
"{jU$t-n3Rv0U$-N-+h3-@Ll3y-W@y}"
sampleandhold
 
Posts: 581
Joined: Tue Apr 15, 2003 4:38 am


Return to EOS: The Lab