I’ve got a question that concernes sampling not only in Renoise. Let’s say i have a 2 seconds long “slicing an onion” sample. I know it’s quite harmonically rich so i can add it to hihats to make them sound cooler. But what if i want to make an instrument based on that sound? Tell me if i’m wrong: If the sample is recorded at 44100Hz, and i will cut a piece 168 of samples lenght(that would be the sample lenght of a C note sample) and loop it, will it sound harmonically correct to other sounds? And what if it would be 168 samples multiplied by 2,4,8,16 and so on? I know people do tune such sounds, but sometimes it’s really difficult and hard to hear, so doing it right is not easy.
When you cut sounds down to such a short length you are pretty much working with grain synthesis at that point. You are essentially creating your own oscillator with some kind of complex waveform type. Looping such a short sound will always give you some kind of tonal quality that can be easily tuned. 168 samples will work nicely, 336 (168 * 2) will also work nicely, 672 (168 * 4) etc. If you keep the loops pretty short then it should always have a nice tonal quality that you can work with, but since it’s so short it probably won’t sound much like the original sample. In other words, you will get some nice tone but you won’t be able to recognise that it came from slicing an onion.
The longer you make your loop, the more of the original sample content will be heard. If the original sample is very noisy, then a longer loop will simply sound more like noise and less like a tone. If the original sample is musical and contains a weird detuned note, then you will heard more of that weird detuned note and less of your nicely tuned C note loop.
You simply have to experiment a bit and find a nice balance between a tonal quality and keeping some characteristic of the original sound. It’s not always possible to create a perfectly tuned instrument, but you can still do some pretty interesting stuff using this technique.
if you take a ‘normal’ sample, it will often have some sort of variation in the pitch (unless you use a pure waveform i guess? or even then?). therefore, if you take just a grain, you will get that grain at the pitch where the ‘normal’ sample was at that particular moment in time. so the ‘normal’ sample actually contains loads of tiny samples which will a lot of times be lots of different pitches/sounds.
interesting stuff (this could be ‘duh’ to people, but it wasn’t to me… well now it is
Well, “pitch” is perceived as the periodicity of a sound (its frequency), so at the time scales we’re talking about here, all samples of equivalent length would be at the same pitch because they repeat at the same interval. If you were to move the play head back and forth in a longer sample where the pitch changes, you’d perceive it as a change in spectral content, not as a change in pitch. (Unless, as dblue noted, your loop is long enough that you get to hear the sound repeating on its own before your grain restarts).
Ok, cool thanks for clearing things out guys, but now other thing is the problem. How to do it in renoise? I mean, how to set sample lenght or loop lenght in renoise for let’s say my 168 samples? I know it’s possible to generate one with certain sample amount, but after cutting and pasting some portion from another sample, it resizes to make the copied portion fit in. When i experiment now with the granular synthesis, i record a sound onto my electribe sx and, there i can cut it to the exact desired sample amount, and then export it to wave, and then drop into renoise, but it’s a long and stupid way to do this.
So, anybody has some ideas how to make it happen in renoise? Renoise in fact is a sampler and it’s brilliant for such things like granular synthesis, but i can’t figure it out how to cut a sample to for example 168 samples.
Man i won’t count how many times the “sample” word is in that post.