e2s: single cycle waveforms?

mabi · Post by **mabi** » Wed Aug 05, 2015 5:41 pm

I have been trying to use single cycle waveforms as samples, wondering if anyone else has gone through this... import the wav, edit the loop point to 000 to make it fully loop.

When I play in keyboard mode it sounds ok, but in a sequence there is lots of clicking going from step to step, not usable.

Tried the different polyphony settings.

Tried editing the wav before importing it, to double it and make the loop the halfway point, adding a fade-in at the beginning to try and smooth it more.

Tried different envelope settings. With no envelope you don't notice clicking I think just because there is no attack.

Wondering if anyone has thoughts on how to make them play correctly, or if the e2s just isn't built for single cycle wavs.

cheers!

apapdop · Post by **apapdop** » Wed Aug 05, 2015 6:14 pm

Not tried it. The built in oscillator types sound good though...

Tom 62 · Post by **Tom 62** » Wed Aug 05, 2015 6:20 pm

Strange. No clicks here. Switching the Amp EG on or off makes no difference. Maybe I'm deaf

I use some single cycle Moog samples from the Elektron website (if I remember correctly they are offered as free sound sources for the Analog Rytm).

AxeBreaker · Post by **AxeBreaker** » Wed Aug 05, 2015 7:00 pm

Are you using headphones? The headphone jack has some clicks the 1/4 jacks do not.

mabi · Post by **mabi** » Wed Aug 05, 2015 7:49 pm

Trying some of the factory patterns with the new samples there is no clicking, just with a basic rhythm pattern I made. I will have to see what was different about my pattern.

Thanks for the suggestions. I found the moog samples on the elektron site. I've been browsing through the massive adventure kid collection:

http://www.adventurekid.se/akrt/waveforms/

That is troublesome if there are artifacts in headphones but not line out, I hope I don't hear that happening.

I think the first reason I was interested in the new electribes was their being compact and battery-powered, so I can play on it wherever I happen to be, using headphones.

dschroeter · Post by **dschroeter** » Wed Aug 05, 2015 8:53 pm

Here are some other one-cycle resources and info about how to make them:

http://www.korgforums.com/forum/phpbb2/ ... 30523b6dbe

Are you sure you are starting and ending at a zero-crossing? Audacity or similar will allow you to zoom into the individual .wav blocks, they should begin and end at 0.

mabi · Post by **mabi** » Wed Aug 05, 2015 9:14 pm

Cool, thanks

So far I only have been using the adventure kid and elektron samples, those seem to have zero crossings correct.

What about tuning? I am guessing imported samples are assigned root note C4.

It doesn't appear you can edit a sample's root note, are you tuning them by ear with the sample tune parameter to get them to C4?

musicmagus · Post by **musicmagus** » Wed Aug 05, 2015 11:24 pm

With single cycles you can use the loop point to change tuning as well as the tune parameter.
When you shorten the single cycles the pitch goes up.. Lengthen them and the pitch goes back down.

musicmagus · Post by **musicmagus** » Wed Aug 05, 2015 11:27 pm

P.s. I am normally running audio out the phones jack and I don't get any clicking. My OS is 1.05

sauce · Post by **sauce** » Thu Aug 06, 2015 5:24 am

dschroeter wrote:Are you sure you are starting and ending at a zero-crossing? .

For CSW, zero-point crossing is less important than when looping a longer synth note. Irregular loop points actually provide a harsher buzzing oscillator.

mabi wrote:I am guessing imported samples are assigned root note C4.

The Adventure Kid waves are mostly tuned to D for some reason.

musicmagus wrote:With single cycles you can use the loop point to change tuning as well as the tune parameter.
When you shorten the single cycles the pitch goes up.. Lengthen them and the pitch goes back down.

True. On the ESX any sample looped at 675 bits will be tuned to C. I haven't had a chance to test this on the ES2 yet.

mabi · Post by **mabi** » Thu Aug 06, 2015 4:25 pm

OK, figured out what I was going for... The sampler's native rate is 48k, and it assigns an imported sample to middle c, so a properly pitched waveform to import would be

48000 / 261.626 (middle c) = 183.468 samples.

To get more accurate in the rounding off, you could double that to 366.936 samples, rounded to 367, for a double cycle waveform recorded/rendered at middle c.

I think that gives me the guideline I wanted to edit waveforms to be optimal.

mabi · Post by **mabi** » Thu Aug 06, 2015 9:29 pm

With free time at work this afternoon, resampled the moog ones and some of the adventure kids ones to 48K, middle C. (Not sure if all of them are actually C4, but they are all C.)

Had to do the moog ones individually because they are different lengths, but the adventure kid ones are all 600 samples long, so I could batch process them in adobe audition. Oddly satisfying to watch a mac pro tear through processing 1500 wav files at once.

Not sure if it is kosher to post them without the permission of the original authors...

AxeBreaker · Post by **AxeBreaker** » Sun Aug 30, 2015 3:41 pm

I tried doubling an adventure kid waveform at 44.1, and then speeding it up to middle C as reported by audacity. It ended up being C4+5, which I tuned down on the electribe to + .8.

Did you get closer with the Adobe software? Also if you want to host the re-formatted adventure kid files, he'd probably do it for you and give you credit, if other hardware/software is any indication.

mabi · Post by **mabi** » Mon Aug 31, 2015 3:52 pm

AxeBreaker wrote:I tried doubling an adventure kid waveform at 44.1, and then speeding it up to middle C as reported by audacity. It ended up being C4+5, which I tuned down on the electribe to + .8.

Did you get closer with the Adobe software? Also if you want to host the re-formatted adventure kid files, he'd probably do it for you and give you credit, if other hardware/software is any indication.

What I did in audition was use the stretch and pitch filter, choosing the izotope algorithm, setting the new duration to 367 samples (how I arrived at 367 in my comment above), and checking the box to lock stretch and pitch shift.

Before I did that I converted the sample rate to 48k to match the native sample rate of the electribe.

What you wind up with is the single cycle precisely pitched to C, so no tuning necessary on the electribe.

Since all the adventure kid samples are the same duration, I was able to make a batch process to run that conversion and filter on all of them at once.

I will have to send him a message to see about what you suggest.

mabi · Post by **mabi** » Sun Nov 01, 2015 4:04 am

Adventure Kid said cool to sharing the modified samples....

https://drive.google.com/file/d/0B4IlC5 ... sp=sharing

enjoy responsibly