# Self-Controlling Lfo For More Randomness

just figured out tonight that you can set the LFO to control its own parameters. i found it useful for creating a random-random, where the LFO’s random shape controls its own frequency. because you choose the frequency, you get the frequency controlling itself, so uhhh meta-frequency? anyway, the random gets random’ed (assuming you’re putting the waveform on ‘random’ of course)

the only problem here is that aside from looking nifty and being meta, it doesn’t allow you to control anything with it anymore. obviously a Hydra should be called in here to solve this problem. just connect the LFO output to Hydra Input, and send one of the Hydra outputs back to the LFO frequency. now you looped it through the Hydra, so you can have one of the Hydra’s other outputs do whatever you want. also, you get the added benefit of being able to have some control over the LFO frequency too, by limiting its width (and thus its speed) (ok you could do this from the LFO directly already but ok)

afaik, this is the most efficient way to do this, because you have more outputs available than when doing a simple LFO(1) random > LFO(2) frequency random.
on the other hand, the LFO > LFO method does provide you with the option to combine waveforms, but then again, who’s to stop you from adding another LFO into the equation?

You have to keep in mind here, when you want to alter several params at a time with random data and then link more than 1 parameter to a single LFO’s output, the result will always be LESS randomness, not more. When you for example connect filter cutoff AND resonance to a hydra fed by a single LFO, 25% on the cutoff will also always mean 25% on the resonance. No matter how you define param ins and outs on the hydra, flip or alter data ranges per param. The result will always be synced to the other params. And that means always less randomness.

For a maximum of randomness there is no alternative to using a separate LFO per param.

In theory you could of course alter the LFO target of a single LFO via pattern commands and reset the LFO then again for each param, all on the same pattern line. But I’d think a separate LFO per param is the way easier solution then.

Nice idea. Setting the Hydra to control LFO type is a lot of fun. Also using a Keytracker to control the LFO means you can get close to having a uniquely weird sound for every note you play, similarly with Velocity tracker.

If you have the time, can someone please share an example xrns of this method so I might have a better time figuring this out? Thank you.

Here’s a fun little example I’ve just made:
dblue-its-windy-outside.xrns

Imagine sitting in a drafty house while the wind is blowing like hell outside

@bitarts: yes, i see what you mean and you’re right on that. this example was more about the loopy concept of having an LFO control itself. the matter of ‘how to achieve the most random result’ is indeed one where you’d best just use separate randoms, thus separate LFOs.

@jupiter: thanks! indeed, you can add whatever else you want to the chain to make the thing even weirder

@dblue: thanks for providing that example, saving me time