OT Special Tricks

avantronica · 2015-01-10T23:30:55.000Z

""license"":

You can also fake probability with a custom “pulse width” LFO, set to HOLD, applied to VOL.

This gives you x/16 probability - set x of the steps to 0, the rest of the steps to -128, and the depth to maximum. Set the speed to something funky, not a power of 2. A multiplier of 64 works best for maximum randomness.

Point taken, but speed is irrelevant when you use ‘hold’ (as long as the rate is higher than the trig rate) but in practice i find that you get perfect sample and hold style on super slow lfos anyway (mind you the random step lfo is kinda 16times faster than all other standard shapes), but yeah, good pointer #OTtip

license · 2015-01-27T06:30:40.000Z

""avantronica"":

[quote=“license”]You can also fake probability with a custom “pulse width” LFO, set to HOLD, applied to VOL.

This gives you x/16 probability - set x of the steps to 0, the rest of the steps to -128, and the depth to maximum. Set the speed to something funky, not a power of 2. A multiplier of 64 works best for maximum randomness.

Point taken, but speed is irrelevant when you use ‘hold’ (as long as the rate is higher than the trig rate) but in practice i find that you get perfect sample and hold style on super slow lfos anyway (mind you the random step lfo is kinda 16times faster than all other standard shapes), but yeah, good pointer #OTtip[/quote]
Not to be a pedant but the speed does in fact make a difference with HOLD, even if the LFO rate is higher than the trig rate. In effect you create a sort of moire pattern between the steps of the free-running LFO and trigs which hold the value of this LFO. If the values are too nice and round (like 4, 8, 12, 16, etc.) you get predictable patterns with audible repetition.
When you use a fairly ugly number for the rate, it’s sort of like running a 53-step pattern sequencer against a regular 16-step pattern. Repetition will still eventually occur but in context it is much less obvious.

avantronica · 2015-01-27T11:44:55.000Z

""license"":

[quote=“avantronica”][quote=“license”]You can also fake probability with a custom “pulse width” LFO, set to HOLD, applied to VOL.

This gives you x/16 probability - set x of the steps to 0, the rest of the steps to -128, and the depth to maximum. Set the speed to something funky, not a power of 2. A multiplier of 64 works best for maximum randomness.

Point taken, but speed is irrelevant when you use ‘hold’ (as long as the rate is higher than the trig rate) but in practice i find that you get perfect sample and hold style on super slow lfos anyway (mind you the random step lfo is kinda 16times faster than all other standard shapes), but yeah, good pointer #OTtip[/quote]
Not to be a pedant but the speed does in fact make a difference with HOLD, even if the LFO rate is higher than the trig rate. In effect you create a sort of moire pattern between the steps of the free-running LFO and trigs which hold the value of this LFO. If the values are too nice and round (like 4, 8, 12, 16, etc.) you get predictable patterns with audible repetition.
When you use a fairly ugly number for the rate, it’s sort of like running a 53-step pattern sequencer against a regular 16-step pattern. Repetition will still eventually occur but in context it is much less obvious.[/quote]

ok ok you just need to ensure that the variation potential is enhanced by not going too slow, which is exactly what i said in my post, i guess i was ignoring the specific context and commenting on behaviour of lfo and evidently thinking of a random LFO shape which is my default, so yeah, my bad for reading/replying too hastily
i understand what you’re advocating with a pseudo random binary state and that you only create potential for interesting variety by having a high irregular rate - the nice thing about using lfo designer is the ability to vary the duty cycle of the pulse and also to make it unipolar, it’s a good technique to have in the bag, very handy also if you want to play random but scale corrected notes too, which is how i use this technique in nord modular, it has its own random binary object ! Using ‘ramps’ with the designer will also allow you to get unipolar random, being careful to choose an irregular rate, to emphasise your point, thanks for the pedantry

MK7 · 2015-01-27T16:02:49.000Z

that one really turned into a nerd discussion I’m not sure if I understand what you’re talking about, at least not every detail of it, just the main message. good that you seem to have clarified something between yourselves

carry on guys, don’t let this comment interrupt you.

random/pseudo-random patterns? nice.

avantronica · 2015-01-27T16:18:54.000Z

""MK7"":

that one really turned into a nerd discussion I’m not sure if I understand what you’re talking about, at least not every detail of it, just the main message. good that you seem to have clarified something

carry on guys, don’t let this comment interrupt you.

random/pseudo-random patterns? nice.

what license wrote made sense, it’s just not what i read when i replied originally, it’s a terrific tip, but it’s actually trickier than you may think to force random scanning of a unipolar sawtooth
ie
////////\ where the low val is 0 and high 127
i find that the random 0-127 sample/hold vaule can be evolving but following a trend all too easily, unlike the random lfo, careful tuning of rate is necessary
the example posted above is sorta like this

-_______________ ie 16 vals, the rate at which this lookup table is scanned influences how many hi/low vale are encountered, you use the designer to manage the proportion of high to low, so it acts only positive (or alternately negative) and only 0 or 127, intermittently

there’s no nerd debate, whilst i didn’t say anything incorrect i just responded as if it was utilising a random lfo shape

it’s akin to having a sliced sample with a number of identical or similar samples and having one empty slice that will occasionally throw a different spin on a groove when slice is lfo varied

cooptrol · 2015-01-28T04:35:07.000Z

Reading this thread inspired me and I started thinking a trick that I will definitely try tomorrow. Everyone talks about connecting midi out to in to use the arp on audio tracks. What if you invert the idea and use audio ccs out to control midi ccs? This way you can control your external synth’s parameters using scenes and the crossfader, and override the lack of scenes in midi mode. Am I crazy?

sicijk · 2015-01-28T06:21:19.000Z

""cooptrol"":

Reading this thread inspired me and I started thinking a trick that I will definitely try tomorrow. Everyone talks about connecting midi out to in to use the arp on audio tracks. What if you invert the idea and use audio ccs out to control midi ccs? This way you can control your external synth’s parameters using scenes and the crossfader, and override the lack of scenes in midi mode. Am I crazy?

Actually you can control only one parameter with the crossfader…being it only a CC message for an external instr.
Scenes exist only for Audio track,sigh

data-purge · 2015-01-28T08:17:14.000Z

Gonna present my trick for humanistic step probability:

Setup LFO as Random, assign it to Vol, set speed to 0, synchronize it ti TRIG, set Depth to 0. Then set machine’s Vol to -50 for example (or p-lock it to step) then you can p-lock LFO’s Depth to ~50 to step and get some kind of step probability.

reeloy · 2015-01-28T13:46:54.000Z

…i love the probabilty option in my lxr…

and i was really surprised not to find such a thing in the elektrons, when i got them last year…

…so, i’ll give that one a try…but a dedicated probability parameter should’nt be that hard to implement, my swedish fellows…i’d love it…and many others, too…

license · 2015-01-28T14:36:16.000Z

""avantronica"":

[quote=“license”][quote=“avantronica”][quote=“license”]You can also fake probability with a custom “pulse width” LFO, set to HOLD, applied to VOL.

This gives you x/16 probability - set x of the steps to 0, the rest of the steps to -128, and the depth to maximum. Set the speed to something funky, not a power of 2. A multiplier of 64 works best for maximum randomness.

Point taken, but speed is irrelevant when you use ‘hold’ (as long as the rate is higher than the trig rate) but in practice i find that you get perfect sample and hold style on super slow lfos anyway (mind you the random step lfo is kinda 16times faster than all other standard shapes), but yeah, good pointer #OTtip[/quote]
Not to be a pedant but the speed does in fact make a difference with HOLD, even if the LFO rate is higher than the trig rate. In effect you create a sort of moire pattern between the steps of the free-running LFO and trigs which hold the value of this LFO. If the values are too nice and round (like 4, 8, 12, 16, etc.) you get predictable patterns with audible repetition.
When you use a fairly ugly number for the rate, it’s sort of like running a 53-step pattern sequencer against a regular 16-step pattern. Repetition will still eventually occur but in context it is much less obvious.[/quote]

ok ok you just need to ensure that the variation potential is enhanced by not going too slow, which is exactly what i said in my post, i guess i was ignoring the specific context and commenting on behaviour of lfo and evidently thinking of a random LFO shape which is my default, so yeah, my bad for reading/replying too hastily
i understand what you’re advocating with a pseudo random binary state and that you only create potential for interesting variety by having a high irregular rate - the nice thing about using lfo designer is the ability to vary the duty cycle of the pulse and also to make it unipolar, it’s a good technique to have in the bag, very handy also if you want to play random but scale corrected notes too, which is how i use this technique in nord modular, it has its own random binary object ! Using ‘ramps’ with the designer will also allow you to get unipolar random, being careful to choose an irregular rate, to emphasise your point, thanks for the pedantry [/quote]
Oh no problem at all man, I just wanted to clarify that, mostly to encourage further experimentation Sounds like I may have just confused other folks more though, oops…
Yeah, exactly, the LFO designer w/ HOLD is great for scales and things like that, you can use it as a kind of bank/database/array of pseudo-random values. Cool for all sorts of things that you might want to restrict to a specific set of values - pitch, timing, volume, etc.
I love the Nord Modular too… so happy about that Wine patch that came out recently, now I can edit my Micromod in OSX (and Linux)!

license · 2015-01-28T14:47:15.000Z

""reeloy"":

…i love the probabilty option in my lxr…

and i was really surprised not to find such a thing in the elektrons, when i got them last year…

…so, i’ll give that one a try…but a dedicated probability parameter should’nt be that hard to implement, my swedish fellows…i’d love it…and many others, too…

To me, a big strength of the Elektron paradigm is that the LFOs are so flexible and powerful that you can easily set up stuff like this with them. At first it might feel weird or hacky. But the alternative is a ton of menus and options for specific things like this, always there in the interface for you to jump over, and possibly always running in the engine, bogging down the CPU, even if you only use them 1% of the time.
For me the Elektron approach is a lot more elegant, I just reconfigure an LFO to do what I need for that particular sound. With this approach it’s also easy to experiment and come up with stuff that Elektron or I would never have thought of. Freedom through flexibility. Even moreso on the Monomachine where you can lock all the LFO settings per step!

MK7 · 2015-01-30T08:57:39.000Z

my 2 cents to probabilistic trig behaviour, although far less sophisticated than what’s been discussed above: map the LFO to STRT instead of VOL the trigs sound slightly different in this case and when STRT is very high the effect is the same as in case of VOL being very low.

@avantronica: thx for the additional clarifications

MK7 · 2015-03-02T04:28:49.000Z

A lot of insights to be found in the science lab february (2015) thread, just to keep potential readers of this thread aware of it:

http://www.elektronauts.com/t/ot-what-have-you-learned-during-the-scilab-challenge/10824

MK7 · 2016-03-07T15:47:52.000Z

Octatrack probability hack using MIDI arpeggios:

http://www.elektronauts.com/t/octatrack-probability-hack/18049

MonstreJumo · 2016-08-31T11:25:15.517Z

Great thread indeed!

My little participation :

Create pattern variations with “silent trigs” and scenes,

On a track, create some trigs with pLocked VOL to 0 (even doable between other trigs actually playing), and then have a scene that locks this track XVOL to MAX. When using this scene, the silent trigs will play, like you get another track running

Kind of scene locking partial mutes!

Use it with one shot trigs for lots of variations!

Cheers

Autopoiesis · 2016-09-03T05:45:12.847Z

Here are my favorite OT tricks that I haven’t seen discussed before, outside of my ramblings on forums.

Use the LFO designer to sequence notes, because you can modulate the order of the sequence with LFOs and you can use it for intertrack modulation. One example
Design quadrature-phase LFOs, use them for circular modulation of 4 parameters, and morph the speed or direction of the circular motion with scenes. For example, swirling a sound around 4 speakers. Also, you could use quad-phase LFOs to create a Shepard tone / sonic barberpole with just one sample of a sound whose pitch is rising - repeat the sample across 4 tracks, delay the sample progressively by “90 degrees” on each track, and route a quadrature-phase triangle designer LFO to each track’s amplitude. I don’t know of another hardware sampler that allows you to do something like this.
To get probabilistic Note On action, route a random LFO to note velocity (such that the LFO’s level and/or the base velocity level entail that the LFO will make the note hit 0 velocity sometimes) and turn off velocity sensitivity on whatever synth you’re sequencing. (You can still sequence velocity with a MIDI CC if your synth is flexible enough.)
“Latch” up to 7 notes on the arpeggiator by holding them down in Chromatic mode and then keeping them held while you switch back to Tracks mode. In TRUE arp mode, you can build up a sequence this way. You can remove notes by going back to Chromatic mode and re-pressing the latched keys (shown as still held down on the keyboard visualization). These arpeggio “sequences” can be stored to parts, but they will be wiped from memory as soon as you pause or stop the sequencer. This is kind of like how you can latch the performance pads on the Rytm by switching modes while holding a pad down.
You can do some bizarre physical modeling of flutes by putting the Chorus’s feedback at its max and bringing its delay close to minimum.

I’m pretty sure #4 is a bug, but it’s a fun one – please don’t fix it, Elektron!

I also love that you can ratchet notes in the microtiming page (press up!), as well as via the arp.

Ryan · 2016-09-03T06:00:49.491Z

Since the audio pool remains the same across projects, I intend to work between two projects. One for composition and and one for sound scavenging. The sound scavenging I just work on recording sounds and arranging samples to be spliced and rearranged. Therapeutic, I reckon. Just sound experiments and noisy nonsense with no need to make sense of anything. And then to take those globs of sound and arrange them in something slightly enjoyable will surely be an interesting process

Ryan · 2016-09-03T06:02:17.496Z

What do you mean “rachet notes”?

Ryan · 2016-09-03T06:09:10.743Z

Oh! A really cool thing I enjoy doing is sending selective arp to my MnM, A4, and AR. The AR-right now- is the least useful. But with the A4 and MnM it’s like opening up a retrigger function for some notes, but also with note variation added on. The midi capabilities of this thing are amazing, and I’ve actually learned a lot about the concept of midi just from using this machine. I thought it was cool as a sampler alone, the midi implementation makes it like the holy grail of music

Autopoiesis · 2016-09-03T06:10:12.721Z

Retriggered notes within the same step. Classic Berlin school trick.