FM Tips & Tricks

Good tips. Cheers

I’ve been there 35 years ago, ( and somtimes still now ) take your time… the biggest part of your life will pass on the road , not at your goals.
The A4 has the fantastic advantage of having the step-sequencer Inside >> That’s a fantastic tool to learn all about synthesis without getting lost.
Just make a pattern with a number of indentical SIMPLE steps.
Examp : - Note length 1 step
- Amp and Filter Env > Attck 0
> Decay 0
> Sustain 64
> Release 1
- FX -setting > All at 0
- Oscillators >>> A Any > Level 90 - no syncing - No pulsewith -
>>> B Any > Level 0 - no syncing - No Puls-sewith - No Vibrato - nothing at all.
- LFO’s > 1 & 2 = 1x at Spped +1
etc…; etc;;;

The poin is to have a sound that’s as boring as possible and to make a pattern with it that has 4 or more identical steps at equal distance > I wouldn’t make it too long to start .

Now you can test all parameters 1 by 1 at ease and test their interactions WHILE THE SEQUENCER IS RUNNING
Have the discipline to keep it simple = limit yourself to 1 page of parameters until you really grasp what they are doing and how they interact. Don’ go to another page until you really get it ! WRITE DOWN what you’ve learned - even better > MAKE DRAWINGS of what happens.
Envelopes will be the biggest challenge - I’m still Learning - for shure when using multiple envelopes send to different parameters + time-variants by velocity etc…
*
I think the sequencer is one of the greatest Tools to learn at ease - if you don’t wanna go to fasT.
*
Don’t hurry - Enjoy slow time Learning. :wink:
Ones a CLIKC is maid in your brain consernig a parameter and what it really does you’ll never loose that knowledge again -it’s there when you need it - and you’ll slowly grasp the deeper interaction with other parameters. You get lost by going to fast :joy: first learn to walk - Before you know you’ll run a Maraton :astonished:

2 Likes

I’ve been there 35 years ago, ( and somtimes still now ) take your time… the biggest part of your life will pass on the road , not at your goals.
The A4 has the fantastic advantage of having the step-sequencer Inside >> That’s a fantastic tool to learn all about synthesis without getting lost.
Just make a pattern with a number of indentical SIMPLE steps.
Examp : - Note length 1 step
- Amp and Filter Env > Attck 0
> Decay 0
> Sustain 64
> Release 1
- FX -setting > All at 0
- Oscillators >>> A Any > Level 90 - no syncing - No pulsewith -
>>> B Any > Level 0 - no syncing - No Puls-sewith - No Vibrato - nothing at all.
- LFO’s > 1 & 2 = 1x at Spped +1
etc…; etc;;;

The poin is to have a sound that’s as boring as possible and to make a pattern with it that has 4 or more identical steps at equal distance > I wouldn’t make it too long to start .

Now you can test all parameters 1 by 1 at ease and test their interactions WHILE THE SEQUENCER IS RUNNING
Have the discipline to keep it simple = limit yourself to 1 page of parameters until you really grasp what they are doing and how they interact. Don’ go to another page until you really get it ! WRITE DOWN what you’ve learned - even better > MAKE DRAWINGS of what happens.
Envelopes will be the biggest challenge - I’m still Learning - for shure when using multiple envelopes send to different parameters + time-variants by velocity etc…
*
I think the sequencer is one of the greatest Tools to learn at ease - if you don’t wanna go to fasT.
*
Don’t hurry - Enjoy slow time Learning. :wink:
Ones a CLIKC is maid in your brain consernig a parameter and what it really does you’ll never loose that knowledge again -it’s there when you need it - and you’ll slowly grasp the deeper interaction with other parameters. You get lost by going to fast :joy: first learn to walk - Before you know you’ll run a Maraton :astonished:
[/quote]
Thanks for the advice, i’ll keep it in mind. I do make the occasional decent sound but its usually just a happy accident whilst noodling.
I feel like I’ve derailed this thread now so lets get back to the topic. Im going to try some of these tips for FM type sounds! :slight_smile:

good tips here.

Im starting to get a feel for how to get similarities in programming sounds similar to how i would go about FM on my Dx’s

I never gave it any thought before , but Id be interested to see the possibilities if the user envelope (env2) could map to lfo 2 parameters (in addition to lfo 1). this would almost serve extra functionality similar to changing algorithm… similar, but not exactly.

I think it helps if you make something clear to you:

  1. Envelope modulates your sound over time with a curve.
  2. Modulating the depth of that envelope with an LFO , makes that change with the applicable to your sound with the frequency you set to that LFO.
    (Set the speed to 1/4, and every fourth note, will be affected, depending on the restart of the LFO.)
  3. Modulating an LFO with an Envelope will blend in, the LFO target over time.
  4. LFO fades, can be used to sculpt the sound beginning /end.

I do FM’ish sounds first as “normal” sounds - i.e. i make sure that the patch sounds good without FM.

Then I add FM, probably I use an envelope which modulates an LFO, to bring in more or less FM to the sound at a certain point in time.

I think its good to have solid starting points, but experimentation also yields cool results. The best things I did discover by accident. Take notes - when something sounds cool - keep it, and take a note what is a cool effect on that sound, use that for a performance macro.
Be very sparsely with FX - you can make sound a lot of things good with echo - but you will need that FX for escalation of that probably later. Make it first as good sounding as possible, and then use the FX as icing on the cake. (Controlled by performance parameters.)

I guess you mean 1k in the LFO pitch mode (the last new 12 options), right?.

What about OSC tune?. I don’t find good ratios between carrier (OSC) and modulator (LFO)…

I’m trying to make harmonic sounds, but they sound awful :slight_smile:

I think I have to keep studying hard :smiley:

Thanks!

2 Likes

Hi!
Does anybody know the actual mathematical function for the frequency of the LFO in “pitch fork” mode (as a function of x, ie the frequency currently played by OSC1)?
If 32*1k is truly the 1:1 point, then is it simply:

speed*multiplier*x/32000

Alternatively, can anyone think of a way to hear the LFO so that it can be tuned by ear?
I was thinking maybe modulating the value of a CV track with the LFO… not sure if that would work.

Thanks

1 Like

Hi there !

I’m since a long time trying to make some Analog four patches with Frequency mod but I’m not really good at.
I’m searching some ressources to make some other presets/patches than FMdrums.

Can you please help me ?
Any setting advice, or LFO/env and routing is welcome… Seriously…

Thank you a lot

1 Like

above

and

https://www.elektronauts.com/search?expanded=true&q=fm%20%20%23elektron-gear%3Aanalog-four

3 Likes

UP ! Any routing idea ?

Actually you don’t need a mathematical function. It’s quite simple.
LFO1 is synced to OSC1 tune. LFO2 is synced to OSC2 tuning. LFO setting to achieve a relation of 1:1 is Speed 32 / Mult “Fork” 1K. This means if you play a note with 440hz the LFO will run at the same frequency, too. If you increase the LFO rate to 2K it will run at 880hz.

Now how to “hear” an LFO? My answer would be amplitude modulation (AM)

  1. Build something close to a sinewave. To ensure you’re hearing only the fundamental.
    2. Turn down Amp Vol to 0. (not really necessary)
  2. Set an LFO to modulate Amp Vol and Depth to max.
  3. Set LFO Speed to 32 and Mult to “fork” 1K, waveform to Sine.
  4. Check an analyzer to verify you’re now not only hearing the fundamental (eg 110hz) but also evenly spaced harmonics above and below it, because that’s what AM is basically doing at equal rates. (eg 110hz, 220hz, 330hz …)

There will be some oddly spaced harmonics, too. But in general you can check/confirm this way where the 1:1 point for LFO to OSC speed is set.

PS: Apologies for reviving such an old thread which I usually wouldn’t. However seeing the LFO syncing question popping up over and over again I thought it might probably be helpful to try and shed some more light on this.

Edit: Please shout if anything is unclear. I tried to explain the best I could right now!
2nd edit: This is more obvious when checking with lower frequencies as these seem to be a little cleaner.

6 Likes

A question came to my mind the other day… maybe someone technically more proficient has an idea around this.

Let’s consider the following:

  • OSC1’s pitch is modulated by LFO1 set to [fork]32/1K

Does this mean that when pitch is modulated up, LFO speed increases simultaneously, resulting in intensified modulation of OSC1? And can this be understood as technically equivalent to a feedbacked operator?

1 Like

i believe (though have not verified) the fork timing is based on the base pitch. pretty sure it’s not fedback.

Ah, yes. This might be true. I will try and see if I can find a way to test this. :slight_smile:

1 Like

Okay. So here’s what I did…

OSC2 is the only audible sound source. OSC1 & OSC2 set up similarly with tracking enabled. Both LFOs had similar settings, too.

a) No modulation

b) LFO1 modulating OSC2’s pitch

c) LFO2 modulating OSC2’s pitch

Observation: There seems to be a significant difference in number of harmonics (up to noise) being produced when using LFO2 to modulate OSC2’s pitch. If that is feedback or something else might be answered by someone else. Either way you seem to “lose” the fundamental, which was in this case at around 82hz.

2 Likes

Super interesting! Thank you this sparks some ideas in my head already. Will go to use my a4 asap :slight_smile:
I love analog FM

1 Like

Btw are you using pitch modulation or frequency modulation with the LFOs? There’s two different options and it’s a little unclear the difference in behaviour. I suspect frequency modulation is intended for this kind of stuff.

Playing with with it today, it seems like if you use LFO Pitch Modulation you get feedback. But if you use Frequency modulation you don’t get feedback.

1 Like

In the example above I used pitch modulation, which is from my understanding exponential while frequency modulation is linear. If I remember correctly they do behave differently with regard to “feedback”, but I’ll check and confirm if I find some time for more experiments…

1 Like

For another flavour, try not using the Oscs as sound sources at all. Instead use Filter 2 as a sine wave (res fully cranked) and use the LFOs and Filt Env to control the filter freq.

6 Likes