FMing the filter

Post your patches here! Put the format(s) in the subject.
Post Reply
User avatar
Kosmikos
Super Deluxe Wiggler
Posts: 1867
Joined: Thu Nov 26, 2020 1:21 pm
Location: London, U.K.
Contact:

FMing the filter

Post by Kosmikos »

One function I really like on my Rev2 is the filter AudioMod. It just brings a bit more fizz to the sound. It’s a really nice feature, but I haven’t seen it on other synths.

I’ve recently thought, surely that has to be an easy thing to do in modular, or even with other synths, but what is it? So back to the manual (which I haven’t re-read in a while), and it’s a simple FM modulation of the filter by the oscillator. :doh:

I really like a bit of FM in my patches. I actually never thought of FMing the filter to be anything but the result of random patching. But it seems such a basic thing to do with hindsight.

I’d be curious to hear more about this simple technique, and its effects when the resonance is not too pronounced (high resonance and it’s just plain FM bell sounds …etc.). It seems it just adds a bit of dirt to the sound, but I struggle to understand what it actually does from a physical standpoint. I can understand FM, it just “wobble” the carrier waveform, but in case of a low pass filter with low resonance, it doesn’t really oscillate at this point, so it’s like a rapid modulation around the cutoff frequency. I’m not sure what that means to the harmonic content. It’s like FM wobble, but with an added timing dimension. Maybe similar to a three FM operator setup?

Does anyone have a better understanding of this? Do you use it in your patches?
Kosmikos’ tracks on Soundcloud: https://soundcloud.com/user-254876270
User avatar
starthief
Super Deluxe Wiggler
Posts: 6712
Joined: Tue Oct 04, 2016 10:39 am
Location: St Louis, MO
Contact:

Re: FMing the filter

Post by starthief »

I think of it as kind of a hybrid of AM and phase modulation (since filters cause a phase shift near the cutoff frequency).

I do like filter FM quite a bit. Sometimes I route the filter output to the freq CV and get fun results.
hammerhead
Common Wiggler
Posts: 105
Joined: Sun Oct 16, 2022 8:12 am

Re: FMing the filter

Post by hammerhead »

It's not like FM wobble because you only change the amplitude of the input's harmonic content (and above the cutoff frequency, also the phase) but not the input's fundamental.
Applying FM to an entire input signal - same amount for all partials including the fundamental - is only possible in delays/ flangers.
Roughly speaking filter FM is more similar to AM (which at full index results in two modulation sidebands plus the original carrier and modulator signals). How different it sounds from AM depends on the filter's slope. Obviously a 6 dB/ octave one-pole filter (like a lowpass gate in "both" mode) deviates less from AM than a 24 dB/ octave four-pole filter.
But again, this is just roughly speaking. The phase shift and other of the filter's properties mean it's not AM. It's filter FM - try to take it for what it is, and try out flanger vibrato/ FM if you can to experience the difference you seem to be unclear about wrt harmonics vs. fundamental/ AM vs FM.
User avatar
goldplate
Common Wiggler
Posts: 204
Joined: Sat Oct 10, 2015 9:12 am

Re: FMing the filter

Post by goldplate »

You make the cutoff moving very fastly over a range of frequencies specified by the amount of modulation you set. It can sound voicy as it makes like if there is more than one resonance added to the sound.
Jee
Veteran Wiggler
Posts: 714
Joined: Wed Sep 16, 2009 2:50 am
Location: France

Re: FMing the filter

Post by Jee »

I’m not much into the physics of it but in terms of musical use FM’ing filters with audio sources has a lot of applications.

As you mentioned, with low resonance you get some dirtier sounds especially if you use bass sounds as modulator. Higher resonance will give a more bell-like or formant-like sound, which is emphasized further with higher frequency modulators.

To me there are 2 main approaches deriving from one question : do I want to keep tuning or not ?

Thing is that you will get the sound processed by the filter (let’s say a vco) and the modulator to act as 2 audible VCO, through the cutoff modulation.

If I want “properly” FM’ed tuned effect (pitch of my signal still in tune) you’ll need a V/oct input on the filter and tune the frequency equal to (or a multiple of) your carrier (original sound source) and get the filter to follow the source. So basically the vco processed, the modulator and the vcf cutoff follow the pitch all together. This leads to interesting tuned gnarly sounds. The more FM you apply, the more drift you introduce due to phase shift (I suppose).

Going a more random way and using random cutoff and light modulation will still work to some extent but you’ll get some notes out of tune the further appart the frequencies are.

Eventually if none of the elements are tuned, you’ll get some cool FX too, but not tuned.

Usually I tend to add “pinches” of FM using envelopes and get some grit on my sounds on the attack (or at the end of my sound). I hardly use a continuous modulation. It allows also some temporary drifts that can sound pretty cool and gives life to a patch while keeping it musical.
User avatar
Kosmikos
Super Deluxe Wiggler
Posts: 1867
Joined: Thu Nov 26, 2020 1:21 pm
Location: London, U.K.
Contact:

Re: FMing the filter

Post by Kosmikos »

Thanks for those answers, lots of food for thoughts. :hmm:
I’ll try experimenting with my Joranalogue Filter 8, which has both exp FM and linear FM inputs, and also self patching the filter. I should be able to get a much wider range of effects with it than I’m getting from the Rev2.
Kosmikos’ tracks on Soundcloud: https://soundcloud.com/user-254876270
hammerhead
Common Wiggler
Posts: 105
Joined: Sun Oct 16, 2022 8:12 am

Re: FMing the filter

Post by hammerhead »

Jee wrote: Sun Nov 20, 2022 2:45 am Usually I tend to add “pinches” of FM using envelopes and get some grit on my sounds on the attack (or at the end of my sound). I hardly use a continuous modulation.
Good point. A focus on this is what set Buchla and Serge apart from Moog, ARP etc in their days. On a 2600, even though it has a host of attenuation sliders, you quickly run out of ways to implement VC over more than one modulation depths. On an Easel the audio modulation paths have attenuators and CV inputs. The difference this makes for how you approach sound design, and the habits you get into, are huge.
I'm not sure if all of the many West Coast inspired brands in Euro take this into account this when they design their frontplates.
ABB
Common Wiggler
Posts: 130
Joined: Sat Dec 05, 2020 4:19 pm

Re: FMing the filter

Post by ABB »

I use the Filter8 also for FMing the filter with the linear FM CV input. I patched it with the BB notch out of the filter which gives some nice gritty variations of a signal, though it leads to a less deeper frequencies if used for a bass, but sounds nice for me:)
LunaticSound
Ultra Wiggler
Posts: 948
Joined: Tue Dec 25, 2018 3:50 pm
Location: Hamburg

Re: FMing the filter

Post by LunaticSound »

Keep in mind you can FM the filter with a lit more than just sines. Noise adds fuzziness, square waves make the resonance jump between two frequencies, samples do crazy shit, as does feedback.
User avatar
Kosmikos
Super Deluxe Wiggler
Posts: 1867
Joined: Thu Nov 26, 2020 1:21 pm
Location: London, U.K.
Contact:

Re: FMing the filter

Post by Kosmikos »

I’ve done a modular recording with a drone using some FMing of a Xaoc Zagrzeb, if anyone is interested. I was feeding it a WMD/SSF Spectrum triangle wave and FMing the filter with the -1 Sub triangle out (I think). -2 sub didn’t seem to work as well for some reason. It’s nothing much out of the ordinary, but with high resonance, some stereo spread, and a nice reverb, it sounded rather good to my ears: :hyper:
Kosmikos’ tracks on Soundcloud: https://soundcloud.com/user-254876270
User avatar
enzyme00
Common Wiggler
Posts: 204
Joined: Fri Mar 12, 2021 9:24 am

Re: FMing the filter

Post by enzyme00 »

LunaticSound wrote: Tue Nov 22, 2022 8:09 am Keep in mind you can FM the filter with a lit more than just sines. Noise adds fuzziness, square waves make the resonance jump between two frequencies, samples do crazy shit, as does feedback.
Love me some square waves for FM'ing filters. Go wild and stick the modulating square signal through a VCA and then use a square LFO synced to some division of your clock to control the VCA. Squares within rhythmic squares 😁
My Modular + Matriarch / My Youtube
User avatar
Snufflepuff
Veteran Wiggler
Posts: 580
Joined: Sat Dec 19, 2020 11:53 pm

Re: FMing the filter

Post by Snufflepuff »

Just to be clear here, are we talking about frequency modulation, which is really only possible when the filter is self-oscillating at a sine tone, etc and thus producing unique pitches at separate cutoff frequencies, or frequency cutoff modulation, which is distinct from frequency modulation? It is a type of audio-rate modulation, but the mathematics (and results) are different than FM.
User avatar
Kosmikos
Super Deluxe Wiggler
Posts: 1867
Joined: Thu Nov 26, 2020 1:21 pm
Location: London, U.K.
Contact:

Re: FMing the filter

Post by Kosmikos »

Snufflepuff wrote: Tue Nov 22, 2022 7:35 pm Just to be clear here, are we talking about frequency modulation, which is really only possible when the filter is self-oscillating at a sine tone, etc and thus producing unique pitches at separate cutoff frequencies, or frequency cutoff modulation, which is distinct from frequency modulation? It is a type of audio-rate modulation, but the mathematics (and results) are different than FM.
Good point, I was thinking more about cutoff frequency modulation (CFM then?), but since the Rev2 AudioMod knob doesn’t really make this distinction, and both are essentially achieved using the same process, I never really thought much about it.
You can often push through self resonance with the input signal when on the threshold, and the filtered signal is still there too, so it certainly seems like a difficult effect to entangle in practice.
Kosmikos’ tracks on Soundcloud: https://soundcloud.com/user-254876270
Jee
Veteran Wiggler
Posts: 714
Joined: Wed Sep 16, 2009 2:50 am
Location: France

Re: FMing the filter

Post by Jee »

Snufflepuff wrote: Tue Nov 22, 2022 7:35 pm Just to be clear here, are we talking about frequency modulation, which is really only possible when the filter is self-oscillating at a sine tone, etc and thus producing unique pitches at separate cutoff frequencies, or frequency cutoff modulation, which is distinct from frequency modulation? It is a type of audio-rate modulation, but the mathematics (and results) are different than FM.
Heuuu… no? Cutoff being the frequency of the filter, cutoff modulation and frequency modulation are literally the same thing, no matter if the filter self oscillates or not.
tru7h
Learning to Wiggle
Posts: 27
Joined: Fri Jul 01, 2022 5:00 pm

Re: FMing the filter

Post by tru7h »

i find filter fm to be especially useful when i'm already doing an fm patch. using the same modulator that i use for the oscillator to slightly modulate the filter can often (with the correct polarity of the modulator) tame overly harsh fm sounds. you're basically filtering down just the overly bright/harsh part of the wave and leaving alone the rest of it.
User avatar
Kosmikos
Super Deluxe Wiggler
Posts: 1867
Joined: Thu Nov 26, 2020 1:21 pm
Location: London, U.K.
Contact:

Re: FMing the filter

Post by Kosmikos »

tru7h wrote: Thu Nov 24, 2022 8:03 am i find filter fm to be especially useful when i'm already doing an fm patch. using the same modulator that i use for the oscillator to slightly modulate the filter can often (with the correct polarity of the modulator) tame overly harsh fm sounds. you're basically filtering down just the overly bright/harsh part of the wave and leaving alone the rest of it.
Good tip :tu:
I’m putting together a Eurorack case with two through zero oscillators for FM and fun stuff, I’ll definitely try this technique once I get it done.
Kosmikos’ tracks on Soundcloud: https://soundcloud.com/user-254876270
Post Reply

Return to “Synthesis Techniques”