Recording buffer sample scrubbing "granular" setup?

isaac_anthony · 2023-12-14T21:51:23.103Z

I have input AB as my drums bus and inputs CD as my music bus.

If i set up a looper track to just record CD, instead of X-fading INTO the buffer looping, can you set it up to assign the x-fader to scrub the sample “granular” style?

So the drums are unaffected, but instead of whatever melody you have playing, all of a sudden, you turn it into “scanning” the sample buffer “granular” style?

How/can this be setup?

Shredrr · 2023-12-14T22:15:50.530Z

on Flex machines, Buffers are and can be assigned to Slot numbers for this purpose. You can place a trig replaying what the buffer has just as any sample in the flex slot list and if you “slide” thru that samples starting point your sample can feel like become granular. In example by using Playback-Pages params STRT, LEN, RTRG, RTIM and likewise on top using ATK, HOLD of the Amp-Page. Just be aware once the virtual playhead is behind the loop point the sample needs restart to scrub again… So your loop settings are of interest as well and therefore the length of your recording is also not without meaning.
As soon the length of a recording is known or can be assumed to be known, then the playhead has something to relate to and can certainly scrub thru it in the precision 64 of 64 chunks/slices or the fixed multiple of the time domain offered by its relation to tempo, depends on your chosen options for that flex machine on track.

• STRT, where to start playing
• LEN, how long can it play if available
• RTRG, how often does it repeat
• RTIM, in the loop frame size above 0.005x and below 8.0x (of the entire sample length i think)

So the goal is first to feed your buffer, seems you figured that out already.
And a buffer is just as good as any sample, even if not saved.
You can not scrub thru something that is not cached in the buffer.

there are other tricks as well. In example using endless FX2-Delay controlled by a trigless trig or scenes and apply the maximum feedback FB & SEND at the moments you want it and time TIME becomes the loop length . As you feed the delay buffer you practically just repeat, can sound granular even tho it is not granular at all.

Microtribe · 2023-12-14T22:27:38.188Z

@nedrush shows you how

https://youtu.be/y3bgvzQkzgo?si=DG1T8aTewdzLt7M2

error.eyes · 2023-12-14T23:27:07.228Z

The basic recipe is simple:

Turn off loop and stretch, and probably a few other SRC attributes, I forget. This is just to get the very basic setup up and running. Turn them on to get crazy if you wish.
Set RTRG to max or INF or whatever it is called
Set RTIM to some value (chef’s choice)
Set STRT to 0
Set Scene A to STRT 0
Set Scene B to STRT at max value
Trigger the sample as needed, keeping in mind envelop settings, sample duration, one-shot trigger, etc.
Move the fader

The idea is that the value of the STRT parameter only comes into effect after a sample is triggered. Once the sample is playing, STRT doesn’t do anything. Therefore, the sample needs to be constantly triggered and that’s where RTRG and RTIM come in. You could think of RTIM as “grain size” in the context of traditional granular synthesis.

I’m gonna check out if ned rush does it a different way.

sezare56 · 2023-12-15T00:06:35.347Z

isaac_anthony:

all of a sudden, you turn it into “scanning” the sample buffer “granular” style?

Part change…AED

This is on one track, you can playback 8 iterations of the same slot on up to 8 tracks, add lfos on start, and scroll AED for the 8 tracks.

lowph · 2023-12-15T00:11:23.888Z

I was reading this thread and remembering that scrolling technique you shared a while ago. I still use that!

sezare56 · 2023-12-15T00:14:11.356Z

I didn’t remember this one :

The Octatrack can be quite granular-esque [example inside] - #31 by sezare56

Same topic : constant radio recording, “direct” granular stuff…

GurtTractor · 2023-12-15T06:23:02.589Z

I recommend using 64 slices of a short sample for granular stuff, because that way you can automatically cut the ‘grains’ at zero crossings which minimises the sharp clicking you tend to get with other methods.

LFOs set to start time are a good idea too for a bit of movment, and I think if you carefully line up 2 or more LFOs with custom waveshapes you could probably generate some even more advanced effects…

isaac_anthony · 2023-12-15T08:07:42.829Z

i usually have my sample buffer at 2 bars.

why wouldn’t i use the entire 2 bar sample to scrub across?

Shredrr · 2023-12-15T09:26:21.903Z

because the scrub position sets in at what STRT parameter tells that is either a Slice from SL1 to SL64 (so 1 of 64) in slice mode or a number from 0 to 127 of 128 positions evenly spread over the entire recording or sample length. The longer the sample or buffered recording the bigger the chunks of 128.

GurtTractor · 2023-12-15T09:30:02.008Z

Yeah if you wanted to use longer buffers then the slice method probably isn’t what you should use, I just thought I’d mention it as an alternative way to do granular stuff on the OT.

error.eyes · 2023-12-15T17:50:31.858Z

That said, it makes for a pretty effect, i.e. on a plucky synth sequence

It’s good to know that this can be done with slices too. Also, turn loop on in ping pong mode and things get interesting. Otherwise, if slices are not needed, the basic approach is more flexible and smoother (or as smooth as the OT can get).

sezare56 · 2023-12-16T11:21:43.313Z

error.eyes:

turn loop on in ping pong

Yes, less clicky, use attack too.

I used pre-tuned tuned slices for granular stuff. All slices can be set at the beginning of the recording, so you don’t need a long recording. May do a demo when I receive my MKII !

Values and some explanations below. Octatrack Tips & Tricks (OT Tips) - #215 by sezare56

I tried to sync lfo in order to use it as grain envelope to avoid clicks, not perfect yet…