Does anybody know how the operators of the fmpart of the moonsound works

Ik have experience whith fm synthesis but only whith the audio and msx music, which is very straigtforward. But whith the other two operators the possibilities seem endless.

Can anyone give me some ideas of how why and what

thank you very much if so

It's not much different than the msx-audio/fm-pac tho. With the FM-Pac/MSX audio you have 2 operators, and with the Moonsound you have 4. So, you can make complexer sounds, sometimes richer. I personally would always use 18 2-op voices instead of a number of 4-op voices together with the remaining 2-op voices. The only difference compared to msx-audio/fm-pac is that you have 8 waveforms per operator, and afaik these are the same as those in the TX81z from Yamaha. Also, there are a few algorithms, but the only one I'd find interesting is the [4]->[3]->[2]->[1] algo. For the rest I'd personally use the 2-op voices. The *big* problem with MBFM is that its channel volume and brightness commands simply refer to operator 1 and 2, *even* when there are more carriers, even when the op2 isn't a modulator. This is truly dumb application design, and in many cases it renders the 4-op voices partly or completely useless when you like to change things during your song.

My advice? Stick to 2-op voices and be done with it.

Here's why 2-op mode rules:
- You can still combine 2 2-op voices using 2 channels, while still being able to use those channels for something else later, e.g. 2 single independent voices instead of a 'dual voice'.
- 2 2-op voices give you a [2]->[1] & [4]->[3] algo after all (there's 1 4-op algo like this), but with *2* feedbacks instead of 1!
- in case you still use a dual voice config on 2 channels, you can pan these channels apart: left/mid/right
- in case you still use a dual voice config on 2 channels, you can detune these channels apart
- in case you still use a dual voice config on 2 channels, you have all the controls for volume and brighness, pitchbend, vibrato per channel.

The only reason you'd ever want to use a 4-op voice is when you really want a [4]->[3]->[2]->[1] algo, because you can't fake that using 2-op voices.

so what you're saying is that i try to forget the 4 opvoices en try something and use the 2 op voices only
pity

i really was hopeing that i could use everything on the moonsound

Well, my last line is the essential line here. One of the 4-op algorithms is a [4]->[3] & [2]->[1] algo, which is also a very common algo you can use for a lot of voices. But this can also be realized using 2 2-op voices, with which you gain all the stuff I mentioned above in that list.

Everything that can't be simulated using 2 2-op voices will *have* to be a 4-op sound, which includes all kinds of odd limitations.

Note that with 2-op voices you already have more potential than you may think, because of the 8 operator waveforms, so, don't think too light about the 2-op voices..

i didn't
I just tried to edit some voices and you were right.

i really have to get over the limitations of the 4 ops
and use the 2ops with the new different waveforms
en who knows what happens