FM synthesis tips and tricks

I’ve heard of a workshop that @Ess did on the Digitone which was live-streamed. I heard he demonstrated how to get harsh and gnarly tones out of the DN.
If anybody has a recording of this please share it!
He talks about the Cycles in this one, and the tones he generates in his set are otherworldly! This one blew me away

2 Likes

No, randomize only works on all the parameter (syn1, 2, amp, lfo etc.) pages. Trigs can have probability but no way to randomize the sequence. You
can randomize the trig page though :slight_smile:

3 Likes

I had an important (and belated) realisation after reading these posts - that there’s a fundamental correspondence between FM synthesis and Fast Fourier Transform which decomposes back into its constituent sine-waves.

4 Likes

Yes, the Model:Cycles sets in that video are great. It’s also an excellent device. The Tone Machine is particularly fun. And many of the kick and snare presets are very good. Unfortunately, many of the synth presets are at maximum volume so they really blast your ears when using headphones. But in a way that’s an encouragement to make one’s own sounds.

The most important thing to tweak when programming FM synths

2 Likes

Watching this thread. :nerd_face:

1 Like
6 Likes
2 Likes

For those of us used to analog synthesis, the main concept around FM that was useful to me to understand was the following:

With analog synthesis we typically take a rich, buzzy waveform (such as a sawtooth) and remove frequencies using a filter (e.g. a lowpass to remove the highs). For example you can create a plucky sound with the filter envelope where a fast attack and short decay provide the bright transient, and a low sustain provides a filtered body with less brightness.

With FM synthesis it’s the opposite - you typically start with a sinewave that is not bright at all and add brightness by modulating the sinewave with another waveform (again, often a sinewave). The frequency of that 2nd sinewave affects what that brightness (e.g. resulting waveform) sounds like (by modulating the 2nd sinewave with a 3rd sinewave you introduce even more complexity).

Then instead of using a filter to create the plucky sound from the first example, you use a volume envelope on the 2nd sinewave so that it changes how much it modulates the first sinewave (the one that you hear). Using a similarly shaped volume envelope as we had with the filter envelop in analog synthesis, you get roughly the same effect. The big difference is that analog synths introduce resonance, which you can’t really achieve in FM synthesis (at least I don’t know how to).

Hopefully some of what I just explained made sense!

6 Likes

That’s a reasonable conceptual framework but if using it to explain to other people I’d suggest a small (admittedly pedantic) change. Rather than saying “analog” I’d say “subtractive” to alleviate potential ambiguity. After all, tons of digital synths use subtractive synthesis and there are non-subtractive types of analog synthesis.
Also, while I also associate fm with modulating sine waves most modern (and many older) fm synths have selectable waveforms per operator. The dx7 was all sine and it’s still the zero point reference for fm.
I don’t mean to sound critical or anything, just trying to add.

4 Likes

Yamaha actually used Phase Modulation in its DXs, and called it FM to simplify communication.
PM difference was pretty vague to me before I see this animation at 6’30’’. Digitone also use PM.

3 Likes

Watch this first, the most easy to understand explanation would be on simpler FM chips

than go to this channel https://www.youtube.com/@powerdx7383

2 Likes

This first, simpler ! :wink:

1 Like

I found this very helpful when I watched it. I’ve got a rudimentary understanding of how it works but the real deal is figuring out what I can do with it. This was useful in showing the specific possibilities.

IMHO, you can’t understand FM as long as you’re looking at waves. You have to see it in the frequency domain or it’ll make no sense.

2 Likes

I couldn’t find a relevant thread and I am not going to start one for this simple question.

Can someone explain or try to explain why FM Synthesis is so good in creating piano-like sounds? What makes it so special?

1 Like

I’m assuming you’re talking about electric piano sounds? I’ve never heard anyone rave about FM’s version of an acoustic piano, although people have approximated a piano with FM.

The classic FM electric piano tends to work well for 3 reasons:

  1. FM is good at bell-like, metallic sounds, the sonic basis for an electric piano. You can make that kind of timbre with 2 oscillator FM (“2 operator” in FM talk). In other words, one sine wave is modulating (or if you prefer, modulating the phase) of another sine wave for every note.
  2. Most FM synths have an architecture with at least 4 digital oscillators per note. So you can easily have 2 FM pairs, each pair making the same bell sound; then you detune one FM pair from the other to make the sound richer.
  3. Another thing FM does well is woody, thumpy sounds. With a 6 operator FM synth (like the old Yamaha DX7, which started the FM EP craze), you could use a third FM pair to imitate the thump of a hammer striking a tine in a Rhodes, which was a subliminal part of the sound.

(People sometimes allocate their 6 operators differently to make an electric piano timbre; the above is just one way. In a 4 operator FM synth, the thump or the second, detuned bell has to be left out—but you can still get a recognizable electric piano, even with that limitation. What shouts “EP!” is the bell-like sound, and chorus to taste.)

Hope this helps.

=edited for accuracy=

4 Likes

Yes! Thank you. Pretty much all VSTs i have have E piano presets. But when i listen to presets on YouTube, it’s always FM Synths that impress me!

1 Like

This is my favorite FM tutorial. So good.

1 Like