Questions about XMF (Zebra 2)

[SLoec]

After a discussion on vi-control about XMF in Zebra 2, I would like to confirm the following points :

1. I tried to emulate "FM knob" of XMF by using a LFO to modulate the cutoff. Theorically, it should be the same, right ? Configuration was as follows : OSC1 (500 Hz sine) as input, a 100 Hz sine LFO. However, the spectrum at the output of XMF was quite different from the one obtained by replacing LFO with a sidechained signal (OSC2 - 100 Hz sine) and turning the FM knob. Apparently, it is because LFO is "dirty" (=not anymore a sine) at 100 Hz. If so, what is the maximum "clean" frequency of the LFO (by experiment, it appears to be around 30Hz) ?

2. Same configuration as above (using OSC2 and FM knob). The spectrum at the output of XMF looks exactly like the spectrum of a FM'd signal (with OSC1 as a carrier and OSC2 as modulator), however : why ? We are not at all doing FM (= modulating the frequency of an osc with the amplitude of another) here, but filter FM (= modulating the cut-off frequency of a filter with the amplitude of sidechained input, here OSC2), which is quite different, right ?

Maybe it's similar because when we modulate the cutoff frequency of a filter, it's kind of equivalent to modulating the amplitude and phase of each spectral component of the signal that goes through that filter, hence the kind of spectrum we would get from AM / FM / PM (= a central spectrum of the "carrier" signal, with around it the spectra of the modulator signal duplicated around it = the sidebands).

Also, maybe the spectra look very similar here because I'm using very simple sine signals and a simple LP4 filter (to better understand what's happening), but the difference may be more important with more complex signals and filters.

Is there some theory about spectral analysis of Filter FM'd signals (not FM'd signals, which are already explained by Chowning's 1973 paper !) ?

3a. About terminology. Why is the "FM" knob named that way ? Here, we are modulating the cutoff frequency of a filter, we are not doing FM (in the sense of modulating the frequency of a signal with the amplitude of another signal), right ?

3b. Terminology again. Why "cross-modulation" ?

Looking forward to reading more thoughts from people at u-he or experienced users, thanks !

[FakeNatty]

I dont know much about all the theories and papers on FM, but all I know is that if you use other synths that actually let you audio-rate modulate the filter cutoff or resonance of a filter, you get something more understandable than whatever Zebra is doing.

"Synths" I know that can do filter cutoff/resonance audio-rate modulation (not just lfos that extend to audible frequency) without some weird added tricks are: Bitwig, Ableton's M4L, Dune 3, MSoundFactory, Phase Plant.

All these also let you do amplitude/frequency cross-modulation easily.

It's better to set the modulation up in these tools yourself and see how it works.

[Howard]

3a. About terminology. Why is the "FM" knob named that way ? Here, we are modulating the cutoff frequency of a filter, we are not doing FM (in the sense of modulating the frequency of a signal with the amplitude of another signal), right ?

But we are doing FM. Chowning's "FM" is arguably more of a bad label than the one below this knob, and it certainly shouldn't invalidate the wider, older definition.

[Funkybot's Evil Twin]

Maybe it will help explain if we change the bold parts of the quote...

3a. About terminology. Why is the "FM" knob named that way ? Here, we are modulating the cutoff frequency of a filter, we are not doing FM (in the sense of modulating the frequency of a signal with the amplitude of another signal), right ?

See that? Modulating the frequency. Or Frequency Modulation. Or FM!