I struggle with the AM. I think I understand how the feature works, but I haven’t found much use for the hard, nasty sounds I’m getting from it. I’m pretty sure there is more to the AM feature than that. Do anyone have any tips they feel like sharing?
1 Like
try adjusting the oscillator you are using to do the AM
so if you are going osc1 into osc2, adjust osc1, shape, slow it way down, etc etc…
1 Like
Effort post ahoy!
[color=rgba(0, 0, 0, 0.701961)]I can’t say I’ve played with AM on the A4 a whole lot but I do have experience with what the MS-20 calls ring mod, and the behavior is similar if you set it up right so it’s a good starting point for experimentation. Here’s how:
[color=rgba(0, 0, 0, 0.701961)]First, go to the sound options page and turn off oscillator drift. We want a consistent baseline while we’re learning what the hell we’re doing.
[color=rgba(0, 0, 0, 0.701961)]Back to the osc settings:
[color=rgba(0, 0, 0, 0.701961)]OSC 1: Square wave, knobs at default, level at 0.
[color=rgba(0, 0, 0, 0.701961)]OSC 2: Square wave, knobs at default, turn it up.
[color=rgba(0, 0, 0, 0.701961)]AM1 off, AM2 on, TRG on.
[color=rgba(0, 0, 0, 0.701961)]Kill the filters.
[color=rgba(0, 0, 0, 0.701961)]Play a note (or better yet set a single infinite note in the sequencer) and what you’ll effectively hear isn’t too far removed from a pulse wave. (If you think you know AM, why might this be? Think of what’s going on with the waveform.)
[color=rgba(0, 0, 0, 0.701961)]Now, the key to ring modulation is that it’s real, real sensitive to detuning. After a few cents the detuning turns into an enharmonious clangor – that hard sound we associate with rings. Let’s experiment then with what happens as we detune Osc 1 (remember, Osc 1 is producing no sound but is modifying Osc 2). Slowly turn the FIN knob up just a bit and you’ll get a harder version of oscillator beating. After about +30 or so you’ll start getting that clangor. Keep going – you can keep turning the fine tune up and the TUN knob will also incrememnt when you’ve gone a full semitone. At a full semitone it’s just some weird mess, but keep going and you’ll find spots where it sounds relatively stable like a hard oscillator sync, such as at +5 and +7 semitones. By the time you’ve gone all the way to a full octave up it sounds like a pulse wave again, except thinner.
[color=rgba(0, 0, 0, 0.701961)]At this point you have a basic idea of what it does and how to control it – keep it around an octave or fifth for that sync-ish yet musical sound, farther away gets you less tonal stuff. Now try some experiments:
[color=rgba(0, 0, 0, 0.701961)]AM is sensitive to the modulating osc’s pulse width. Try different PW settings on Osc 1 and PWM. It sounds different when you go left than when you go right, although I don’t really know why.
[color=rgba(0, 0, 0, 0.701961)]Modulate the tuning of Osc 1 with envelope 2. Setting Osc 1 to a 'stable (octave or fifth) setting and the envelope sustain/release to 0 will get you a clangy transient at the start of a note.
[color=rgba(0, 0, 0, 0.701961)]And of course play with an LFO. Set the retrigger to TRG to lock the clangy parts to the rhythm. Try both positive and negative pitch mods for Osc 1. The Exponenial and Ramp LFO waveforms are especially fun here. My favorite is speed at -16, multiplier at 36x and an exponential waveform, with a depth around +12. Or the same settings with a square LFO for a dirty take on that 80s-esque synth pop bass.
[color=rgba(0, 0, 0, 0.701961)]Finally, when you’re done with that you can start trying out other waveforms. This is beyond what the MS-20 can do (it can only ring mod a pulse against a square) so I’m in uncharted territory here. Just remember to go slowly when experimenting, changing only one parameter at a time until you’re sure how it affects the sound. There’s a zillion ways you can tweak things even without getting the filters involved, and it’s easy to lose track of what each parameter is contributing.
[color=rgba(0, 0, 0, 0.701961)]Have fun!
7 Likes
Addendum: Turning off keytrrack on Osc 1 (so it’s a constant ‘pitch’) and dropping it down as low as possible is rather instructive – the note comes in and out like an arpeggio, as its amplitude is zilched out on the downswing and increased on the up. This is why the ring mod kills the sound when the pulse width is all the way left (when the pulse is always down), and does nothing when it’s all the way right (when it’s always up).
1 Like
This has been very helpful! I’m getting results I like now, so thanks a lot for taking the time to type it.
1 Like
No problem – I learned a lot typing all that up. Now I need to take my own advice and apply it to Oscillator sync, which I’ve never really been able to wrap my head properly around. I know the theory but can never grok what actually affects what.
next up, check out the feedback oscillator 
I’ve been p-locking AM per step on some crazy short saw leads lately and getting great results, especially in combination with jumping up an octave or two and adjusting oscillator sync mode all on the same step.
It’s getting me a bit closer to my “a-ha!” moment with Analog Four.
there is a really good thread on here somewhere about the am features on the analog four. the pulsewidth on the oscillator doing the modulation is the key component, and the behavior differs between the oscillator types.
search for that thread, as there are very detailed descriptions of the behavior there.
Yeah, I read a nice thread about AM a while back. This one, right? http://www.elektronauts.com/t/amplitude-modulation/431
A nice read for understanding how the AM feature is implemented.
My problem was not understanding the feature, though. I just wasn’t able to create any nice-sounding patches with it. Hardwired 100% pulse modulation just seemed crazy at first. But dumbledogs tips helped me get past that.
…and yeah, I love the feedback oscillator. Having tons of fun with that. 
I must admit I was a bit underwhelmed by the a4 sound engine when I first bought it, but I slowly realize there is much more to discover and explore than I first thought. it is my lack of skills that is limiting me, not the instrument.
1 Like
I like switching off the tracking on one of the oscs, good for getting weird.
Yes - this can be a great way to just change the harmonics in a drastic way - for shure if you change the note -root at some trigs.
Some good tips here so far
For me a handy way of thinking of AM is as extremely fast tremolo. So fast that it enters the audible range. You are using the varying amplitude of one waveform to modulate the amplitude of another. Doing this really quickly (audio rate) introduces frequency components that may not have been contained in either of the initial frequencies.
AM works by creating sideband frequency components at the sum and difference of the base frequencies. The modulator being fed into the carrier (ie. OSC2 --> OSC1) is what creates these side bands.
If the pitches of the two OSC are related numerically than the result will be a harmonic tone related to the base frequencies with a well defined pitch. If they are not related then it will be inharmonic and will not have a well defined pitch.
This means that if OSC1 is set to 220Hz (pitch of A)
and OSC2 is set to some even ratio (multiple or division, so like 110Hz)
The sidebands created will be at 220 + 110 = 330Hz (the sum of the two frequencies)
and 220 - 110 = 110Hz (the difference of the two frequencies)
This will be a harmonic tone.
I realise the A4 does not deal in Hertz but operating in the musical scale is just as easy. Use octaves or harmonically related intervals such as perfect fifths etc as the base freq for each OSC and then subtracting or adding multiples or divisions of these.
You can entirely turn down the volume of whichever oscillator you are using as the modulator. This means you will only hear the effect it has on the osc being modulated. So that depends on whether you are using AM1 (modulates OSC1 with OSC2) or AM2 (which modulates OSC2 with OSC1)
You can of course do both at once but I find it interesting to start with simple waveforms such as the triangle (the simplest the A4 has). The A4 has no sine waves but you can approximate one by taking a bit of high end off a triangle wave with a non-resonant Lo pass filter
And then as mentioned check out the Feedback oscillator in OSC1 
I’ve been using the A4 AM quite a lot recently. I’m enjoying the sounds I’ve been getting through experimenting with it. The tips in this thread were really helpful as a starter.
I’m glad you posted, I have recently acquired the Analog Keys and its AM has remained a bit mysterious for me, this thread is really helpful. I’m off to dig into some AM drums now 
1 Like
Love how posts from over 4 years ago still help new people… thanks for chipping in.
2 Likes