FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 

Why self oscillate a filter?
MUFF WIGGLER Forum Index -> Modular Synth General Discussion  
Author Why self oscillate a filter?
It seems that many VCFs boast of the ability to self oscillate at high resonance settings. Can someone explain to me how this is a good thing? I'm getting to know my Rossum Evolution, and frankly, I'm finding it's tendency to self O pretty annoying. What am I missing?

If you're asking about specific uses besides just "sometimes people like that sound," one technique is to set up 1:1 tracking between pitch CV and filter cutoff and tune the filter cutoff to a specific interval above the oscillator fundamental(s). Since high filter resonance increases the rate at which frequencies above the cutoff attenuate, this is pretty close to cutting off frequencies above the filter cutoff while also introducing an oscillator at that pitch.

As for a specific application, you can use this to do a simple yet effective subtractive approximation of a Hammond organ - set the oscillator to a sawtooth, crank up the filter resonance, set up pitch tracking for the filter cutoff, and tune it so that the cutoff lies at either an octave and a fifth or two octaves above the fundamental. This will give you a fair simulation of the 888 or 8888 registrations (tweak as desired!)

Of course, if you want to make crazy filter FM tones, that's also a good time to push the filter into self-oscillation - the more the resonance is cranked, the more audible the effects of filter FM will be.
The Junglechrist
Sine wave from filter sound better to me to make Kick drums sounds than VCO, but I never realy tried to know why I have to admit.
Thank you. That's all good food for thought. I'll see what I can come up with.

@The Junglechrist
Agreed. the S.O. sine from filters feels mighty smooth to me too. But is it a cost effective use of a $350 module, when it is designed to do a whole lot more that just spit sines waves?

I guess the subtext to my post is that I want my filter to add growl and snort to my bass lines, and instead I'm getting a lot of shriek and whine.
SuppleWhat wrote:

I guess the subtext to my post is that I want my filter to add growl and snort to my bass lines, and instead I'm getting a lot of shriek and whine.

The Evolution filter is a mighty and versatile beast, it is certainly capable of growl. I don't know what you mean by snort, but I'm sure it can do that to.

Key with growl is a low resonance setting and filter distortion. Normally you use a high input level into the filters audio input to get it to distort. Oscillators can output 5V, but there are ones with 10V-12V outputs. If you have a 5V oscillator, just add something like a Doepfer A-183-3 amplifier.

An other option is to patch the filter output back into the input. You need a multiple and mixer module for that. AKS Minimoog trick.

Audio rate modulation of both the filter cutoff and resonance are fun to play with. You can use the audio input oscillator for that, but other oscillators with different pitches completely change the filter characteristics. Experiment! Edit: a little bit goes a long way.

And finally, the Evolution has the Species control, that controls the amount of distortion in the filter core. Read amount of growl. Use audio rate here to in combination with the Genus and audio rate modulation of that.

Start with just 1 oscillator into the filter at 40-60Hz. Set everything to 0.
Turn up the cutoff and listen.
Turn up the species now you should get decant growl.
try the rest of what I wrote.
supposedly the sine is very pure - apparently some FM enthusiasts don't use VCOs - just sine waves from filters.

Bare Modular had been using my AI004 OTA filter as a sine oscillator both for bass sounds and higher tones and I love it!
I agree with the "purity" comment. If I look at the sine wave from my .COM Q106, it has a tiny bit of a hitch in the shape of it when observed with an oscilloscope. It's not a perfectly smooth sine wave. I've not closely examined my Moon oscillator to see if its sine is more perfect, though I hope it might be, since it was more expen$ive. However... even knowing this, I've not deliberately used my filter as an oscillator source.

I'll have to try the FM idea, to see if it works better, with smoother harmonics.
The purity will depend on the signal. A VCO's sine wave has to be shaped from a triangle wave, so there will always be some offset to it, and depending on how your trimmers are set/the design, it can be better or worse.

The purity from a filter (using the AI004 as an example here as I've spent months looking at it in a scope) is pretty dang pure right off the bat.
I like to feedback a self resonating filter into itself to get really beefy tones that make awesome kick drums and the like. I also like saw core oscillators for their rich harmonics, but most of them don't have sine outputs so I use a self resonating filter for that duty.
Pinging filters on the verge of self oscillation is a beautiful thing! thumbs up
self oscillating filters often make nice sine waves. if you add a envelope to the cutoff you can make dope bassdrums or use it as an oscillator to play nice synewaves or use it to fm anoteher osc:)
It can be fun to experiment with different filters that are self-resonating but also have another signal coming in at a slightly different frequency.
Some filters will try to entrain to match the incoming signal, and right on the edge of that can be some nice timbres for drones.

Some filters also have a little something extra that makes self-resonating interesting, like Three Sisters where you have three sines (and can use one to FM the others, etc.) or Cinnamon where you can flip switches and get dirty sawtooth and other stuff out of it, depending.
I had the same thoughts when I started playing with my first filter (Ian Fritz Threeler) not that long ago. A couple observations that may be specific to that filter:

If you're using it as an LPF to gently roll off some higher frequencies and then crank the resonance you're going to get an awful squeal because the frequency on the filter is cranked. Try turning down the freq - it might cut out the input signal but you may find you like the sound.

The input level is highly interactive with how the filter resonance interacts at self resonating Q amounts... play around with the input level as milkshake mentioned.

Keep that Q level low and raise incrementally when approaching self oscillation. On the Threeler Q at 9 oclock is a healthy amount, at 12 it's generally getting nasty and unruly. Past 3 oclock a tear in the spacetime continuum appears and all hell breaks loose.
For kick drums of course, i mean are you into heavily ambient/non drum/ drone music or what? hmmm.....
I like using filters as oscillators for kicks as stated, in full resonance or just barely self-oscillating when pinged. This is how you get the famous buchla bongo sound. I use my Korgasmatron II as a sound source about as often as I use it for filtering.

It's also nice to feed the self oscillating filter with noise, you get a pitch from the self resonance but you also get a "airy" texture. This is how the famous Tomita whistle lead is made.
Dr. Sketch-n-Etch
Actually, filters don't "naturally" produce flawless sine waves. Almost all of them must be amplitude-limited in some way or they will simply form large waves which clip against the rails and become almost like square waves. Also, even when the sine wave is limited, it must typically be filtered to eliminate fairly severe distortion. For example, SVFs which self-oscillate typically only put out clean sine waves from the LP output. Cascaded OTA (or VCA) filters should only be tapped for sines from the fourth stage. The first-stage sine wave if typically pretty distorted, but the fourth-stage sine will be very clean (less than 0.1% THD, typically). If a filter does produce a clean sine wave without limiting, then it is right on the edge of oscillation and the oscillation will not be stable.
Hey Dr Sketch,

Is there a way to CV modulate the resonance on the korgasmatron II? I've tried multing the output to a VCA and back in the input but when I fade it in it causes the resonance pitch to drop which is different than when I turn the knob.

Love your work! Thanks
Dr. Sketch-n-Etch
tito_tunes wrote:
Hey Dr Sketch,

Is there a way to CV modulate the resonance on the korgasmatron II? I've tried multing the output to a VCA and back in the input but when I fade it in it causes the resonance pitch to drop which is different than when I turn the knob.

Love your work! Thanks

Yes, but only with the expander module. The new Morgasmatron has built-in VCQ, but the input is unattenuated and sums with the position of the Q pot.
SuppleWhat wrote:
Can someone explain to me how this is a good thing?

Because it can sound nice in the right context. I came late to the “pinged filter party” but am now pretty much obsessed! Not every filter sounds great doing it - my favourite for that at the moment is Three Sisters.

There is a lot of control on Evolution so you “should” be able to find your personal sweet spots. Personally I found pulling the filter frequency low as it starts to peak gives more pleasant results so you don’t get that squealy feedback finger nails on a blackboard type sound. Running a low-ish frequency square wave as source with Q peaking less than an octave above does nice things for my ears. But I am fairly deaf!
As a side note:

A lot of old drum machines (808 78 606 etc) use nearly self resonating filters as sound source. The cutoff is the frequency of the sound, the amount of resonance is the decay of the sound.
MUFF WIGGLER Forum Index -> Modular Synth General Discussion  
Page 1 of 1
Powered by phpBB © phpBB Group