LFO´s... what are in them?

Elektron Gear Octatrack
#19

I haven´t tried or used the LFO designer much, but will explore/experiment with it. Maybe something will come out of it, but probably not the initial ‘wish’. But that´s perfectly fine too. I´m just trying to learn more about the LFO´s (and/or other stuff) more in depth, bending the rules and/or limits of my mind regarding them.

Just found out this weekend that the OT doesn´t need samples or live input of sound to be able to create sounds. Really rewarding experience![/quote]
I should read everything before answering.
You nailed it in one sentence, @mkdsl ^^

BTW : what was the need behind the fact that you don’t want to use locks, but rather LFOs ? Would you mind developing this aspect ?
[/quote]
Basically it wasn´t a need per se, but more of a idea that came out of the experiment of what happened when just slicing an simple single kick.

As I did slice that and then played the slices in slice mode, listening to the sound of each slice. I ‘realized’ that there might be a potential of creating the impression of lots of variations of kicks where there in fact are no variations (sort of the impression of having an sample/slice chain). And that these variations could be made via different settings of ie the filter base/width as well as envelope (ie to get more snappy sounds etc).

But playing in slice mode doesn´t AFAIK utilize the p-locks, since they are just slices but NOT the actual steps in the grid sequencer which the p-locks are tied to. So my thought with the LFO´s where that since they can be connected to trig events, playing slices would give me the trig events to get the LFO´s moving and thus getting the variations from them. But it would also mean that there are no distinction between playing the same slice several times in a row or playing different slices. The LFO would run in any case. Not necessarily a bad thing as such, but I was thinking if I could somehow workaround that and get a more controlled behaviour. One could see it as a challenge of trying to get the OT to do something it´s not supposed or designed to do.

Will see what options the LFO designer gives me.

#20

Well… While in slice mode you can definitely plock a slice of your sample-chain to a step, it’s rather convenient and easier than the LFOs if all you need is some static assignation of a slice.
I see at least two things that are very handy with this :
[ol]
[li]You can for instance use the Slice Play mode to use the trigs buttons to play the slice of your chain and record it live.
E.g. if your slice is rythm oriented, you can tap your beat with the trig buttons, and it will record the slice trigged for each step.[/li]
[li]A nice thing to mess things up is[/li]
[/ol]
[ul]
[li]activate all the steps[/li]
[li]slice your sample-chain[/li]
[li]in the slice edit menu, choose “random trigs”

=> the slice are affected as plocks in random order ![/li]
[/ul]
You can also use the linear assignment to respect the original order.
All this uses p-locks on slices.

Another way to play with slices without the plocks :
use the crossfader to change the slice that is chosen for each step.
It works well with a sample-chain of different kicks, for instance.
Like : you sample one kick with different filter parameters, then use it as a sample chain.

Back to your LFO consideration :

I have absolutely no idea what you mean here.
Mind developping a little bit ?

#21

[quote="“Lying Dalai”"]
…

Back to your LFO consideration :

I have absolutely no idea what you mean here.
Mind developping a little bit ?[/quote]

miketheman wanted to have modulation applied to the sound that depended only on which trig key he was tapping. He asked whether an LFO could be made to do that, but LFOs move through their cycle in pre-determined steps and cannot alter their value depending on which trig key is being tapped.

#22

Still a bit blurry for me, excuse-me :dizzy_face:
Do you/does he mean : altering the sound of the sound (a kick for instance) throughout the whole sequence by pressing the trigs ?
Like : a combination of the trigs pressed <=> a certain slice is played/effect is applied ?
Or :

  • if trig X is on, it plays slice X with filter settings X,
  • trig Y is on > slice Y with filter settings Y
  • trig Z off > no slice played

I think my proposition does answer to the later. Doesn’t it ?

#23

Think realtime playing slices (in slice mode). No sequencer running with any p-locks.
Different trigs (slices in slice mode) = different slices and effect settings applied.

#24

How that, “no sequencer running” ?
No p-locks, OK, but what do you mean with no sequencer running ? What’s the goal ?

#25

:stuck_out_tongue_winking_eye:

The crash course goal where to learn more of LFO´s. For me basically, learning their limits I guess. To see if and how I can bend the ‘rules’.

The idea that came from the experiment of just slicing a single kick. Was to see if I, somehow by using the LFO´s and/or other stuff. Actually could trick the OT into giving a impression and behaviour similar to as when you are using a sample slice chain. But with the difference that there are no sample slice chain as such (with different drumhits). But just grabbing any random one short single kick. Sliced into many slices. And the effect settings to alter these slices into variations of kicks and/or other drums. That means: no preparations made on beforehand (like you normally do with sample slice chains).

Playing (but not recording) the slices in realtime by hitting the trigs (in slice mode) and by that having the LFO alter the effect settings from the trig events (= me playing the slices). AFAIK, the trigs in slice mode have no connection to the grid sequencer steps themselves (where the p-locks resides). But I do know of course that I always can record the trigs in slice mode first, and then alter the effects settings afterwards with p-locks at certain steps in the grid sequencer.

This idea could perhaps be described something like ‘LFO p-locks’ that occurs without the sequencer. And the question was if I could use one or several LFO´s to get such an controlled behaviour. It seems not possible, at least not without trickery outside the OT (which I don´t mind).

Slots or Slices
#26

And yes: it is totally overkill.

:imp:

#27

Wow.
OK.
Now I understand why I didn’t :dizzy_face:
What is this product you use, mike ?
Seems rather efficient !!
:smiley:

So…
ONE sample of a kick ?
Sliced into many slices ? With no overlapping I presume lile you want to grid slice to avoid manipulating…
OK.
You trig them manually with the Slice play mode.
No trigs on the sequencer.

The sequencer running free, with some LFO in the back ground turning knobs for you. Designed or not. On effects or whatnot (retrig is cool, for instance)

I tested this, seems OK for me.

Yet I’m not really convinced I could use this as a way to design a beat.

#28

:stuck_out_tongue_winking_eye:

To be honest: I´m not convinced either.

But it´s fun to challenge oneself. I´ve noticed that whenever I do that, I often learn something more along the way. And sometimes in the end, THAT have even become more valuable than the primary goal. It´s by trying to do stuff not ‘supposed’ to be possible you are able to learn about limits, either in yourself or in the unit you are working with.

As ie, when I recently learnt that the OT does NOT need any samples or sound to work with. It´s capable of generating it own ‘sound’.
One limit down, more to go… :alien:

Seems that you´ve now got the big picture. LFO´s working in the background probably needs to be a peripheral device (my Nord Modular G2 Engine would be perfect for the task).

#29

Interesting…
Would you mind developing a bit ? I don’t see how to do so.
Well, let me think…
I would use some noise generated by instance by normalizing empty record buffers.
Then select a little bit of it where the waveform is near the bottom and rather flat.
Then use Veets’ Catalist trick to draw my waveform from the LFO designer, and resample… And here we go, I got a wavetable synth.
Do you see another way ?

#30

Interesting…
Would you mind developing a bit ? I don’t see how to do so.
Well, let me think…
I would use some noise generated by instance by normalizing empty record buffers.
Then select a little bit of it where the waveform is near the bottom and rather flat.
Then use Veets’ Catalist trick to draw my waveform from the LFO designer, and resample… And here we go, I got a wavetable synth.
Do you see another way ?
[/quote]
Close. Just a bit more rough.

I made an example this weekend, that I didn´t put to much effort into regarding its musical qualities. But more explored the possibilities:
https://soundcloud.com/mikael-johansson-6/selfie-6

Lucky accident that led to this idea:
In another project I had a track with filter base/width settings tied to the fader. Plus an neighbour track with its own filter base/width settings also tied to the fader. Both filter had quite high resonance. But the differencies between the two filter were the actual range (base, width or both… don´t remember) that the fader controlled in each. I had an sample loaded into that track, but discovered that something else happened when I had pressed stop on the OT.

The sound resembled much of some sort of radio squeel/noise. I hadn´t even trigged the track, but just moved the fader only. But still the OT made its own sound. So then it striked me: self-generated noise. If I could just make an pulse and record that immediately and build everything else from it, you don´t need samples or live input. Rewarding and great fun!

#31

I really like what you came up with. This is hacking !!
:stuck_out_tongue_winking_eye:

Note : to easily clean your record buffer, you could use a Part that would be set up + dedicated to this and record void :slight_smile:

#32

Thanks.

Hmmm, record void you say? Never seen/used that. Is it an specific button combo or something else I have to do?

#33

Damn I figure now that I kind of tried to answer what was in your Soundcloud. I should get some rest. :zonked:

In fact, the best thing to erase all the tracks at once would be to get a special Pattern for this, where all Record buffer have no incoming source (no A B, no C D, no internal source).

I record void often, each time I’m as precise with configuring OT than I am to answer posts in this forum…
:dizzy_face:

#34

I’ve had some mixed results using a random LFO to choose between multiple round robin samples of the same instrument hit over different slices.

One issue I’ve had is figuring out how the depth works in that situation. Can somebody help me figure out how best to approach this situation?

A sample chain with 5 round robin kick hits, 5 round robin snare hits and 5 round robin closed hats. So:
1k - 2k - 3k - 4k - 5k - 6s - 7s - 8s - 9s - 10s - 11ch - 12ch - 13ch - 14ch - 15ch

In this situation I would plock a kick trigger to 3k, set the LFO to slices, random, and depth to 5.

I seem to get mixed success when I try this in various situations. Can somebody tell me exactly how the depth number works? Does it include the plocked (in this case 3k) slice in the number? Like, if you set the depth to 1 does that mean 2k, 3k and 4k are all possible results since each is one away from the plocked setting? I’m away from my machine but I seem to remember a depth of 1 only including the 4k with the 3k. So does the depth prefer to increase when it can only go one direction?

I remember having similar difficulties using a custom LFO that starts at zero and increases horizontally to the max. I wanted to use it to rotate between the snares, so I plocked 6s and set the depth to 5. Results were hard to understand with kicks continuing to be triggered.

Apologies for being away from my machine and unable to give concete examples, any insight into the depth number and any ideas for best performing this type of round robin slice triggering would be much appreciated.

1 Like
#35

The LFO is bidirectional, and the depth is effectively the number of slices in each direction that will be traversed. You will get odd results with a random LFO and with such a low depth.
More ideas here:


http://test.http://www.elektronauts.com/t/sample-chain-lfo-slice-question/12923
Random sample chain help
Tip: Randomly firing samples using silent slices

#36

I don’t actually have anything useful to contribute with, but I have followed the conversation with interest. Just wanted to mention that this is a great idea! A definitive must try.

#37

What you need to do for this is set your depth to 2 and make sure the slice is set to the middle of each group, this way it will choose a random number +/-2 from the middle point of 3 = random over the 5.
I do wish you could change the lfo to go unidirectional and set the polarity. i also wish you could set the phase, and send a reset. I also really wish that the OT has a mod matrix where you could set values based on other values. This would allow the OP’s request of setting another value per slice. Would also fix common requests like notes -> slices. But eh. I’m not upset with what is available :slight_smile:

In regards to OP, have you tried setting a ramp lfo to single, synced to the first step, set speed to 16x1 (i think will enumerate over 0->depth across 16 steps, if i have my memory, and maths right. set to 32, 48 or 64 for it to take that number of steps to get through the range.

#38

Hi again everyone!

Have been away a few days. I haven´t tried any of the suggestions, but will when time permits. In the meanwhile, thanks for contributing in whatever way. Lots of helpful stuff. Keep ´em coming!