IS FM synthesis your goto?

[WasteLand]

a whole FM thread on KVR, and i did read some stuff, but didn't react, must look at all the pages if there is really new info, an interesting another perspective.

i still use FM a lot, real FM, and PM, but i tend to forget that, strangely, mainly i use it in soft modulars. (well also soft-synths, for that matter)

some 'nice' modules that offer true FM, with TZFM, and even other connections for real FM... or FM and PM and AM (at audio rate of course).

so, i am not focused on FM itself anymore, it is part of most of the times of patches, that contain many different modulations at audio rate (or not...).

still educate myself about it, still learning, although i know my way around. FM8 is still my favorite FM (PM) focused synth, F'em, i have my eye on, but it is FM8 extended...

but the greatest combination i made, was AM/FM/PM, with partials, i.e. a module with a partial editor, building it up from scratch. 24 partials, and what a sound...

FM has become more a part of a chain.

but for example in FM8, the MSEG's are great, give me them...

[gentleclockdivider]

FYI , I was talking about phase modulation FM , no idea why you bring up the oddisey-Mini
I have a TG77 and know it inside out , I just didn't bother to mention every feature .
Phase offset , multiple feedback , and 0hz are imho the most important features which opens a lot of possibilities ( true pulswidth etc..)

And I just added "a few other things", to further expose the differences between SY FM series and the DX7. And I forgot to say we have more FM algorithms, and the possibility to create our own algorithms. Like you, I also have a TG77 (and a SY99). And I have been looking for an FS1R. Why? Maybe because I never considered a DX7 to be satisfactory.

Apparently, neither you.

It's wrong to assume I got the tg77 because the dx7 didn't satisfy me .
To me the tg77 is the pinnacle of FM synthesis and the dx7 is just great by itself

[imrae]

I'd be interested in a FM synthesizer that uses a different algorithm to prevent aliasing when modulating at extremely high rate.

Not even sure what this would be? Analog FM?

Anti-aliasing modulation at a high rate seems simple enough: filter the modulation signal. Nothing useful is going to come from the modulator above about 10 kHz, right? The hard part is surely preventing aliasing when modulating at an extremely high index. Then the problem is harmonics that come directly from driving the (co)sine function harder. This is the same (difficult) problem as anti-aliasing a distortion effect!

An analogue waveshaper is one solution; I love my FM Aid. But while this totally removes aliasing, it adds imperfections to the sine wave shape so your low-index sound becomes dirtier.

(I have been tinkering for a with ideas for an analogue VCO which uses waveshaping to do phase modulation with integer ratios. But maybe a semi-digital implementation would make more sense. )

[foosnark]

Anti-aliasing modulation at a high rate seems simple enough: filter the modulation signal. Nothing useful is going to come from the modulator above about 10 kHz, right? The hard part is surely preventing aliasing when modulating at an extremely high index. Then the problem is harmonics that come directly from driving the (co)sine function harder. This is the same (difficult) problem as anti-aliasing a distortion effect!

Bessel functions can predict the frequencies of the sidebands created by FM, and some 2-op synths will restrict the FM index based on that. But unless I'm mistaken, that only works for modulating a sine with a sine -- as soon as you add a third operator things get super messy.

Another fun thing with FM is that sidebands are generated both above and below the carrier and modulator, and can have negative frequencies -- equivalent to positive frequencies with reversed phase. Which means those frequencies "reflect" off of 0Hz and become inharmonic in the same way they reflect off of the Nyquist frequency when aliasing happens. No amount of oversampling, or even staying 100% analog, is going to change that.

[Mumdad]

(I have been tinkering for a with ideas for an analogue VCO which uses waveshaping to do phase modulation with integer ratios. But maybe a semi-digital implementation would make more sense. )

Have you seen the Motas-6? Semitones/cents rather than ratios but it offers phase modulation with VCOs. I'm nothing like technical enough to speculate on how it was done but I'm interested enough that I definitely want to check one out in person at some point..