could it be that the PW weirdness with the SAW waveform has the purpose to compensate for a missing modulation amount parameter?
That’s true that I misread you about pulse wave instead of pulse with, but it’s also true that when you are using a saw or triangle wave and pulse width is 0 you don’t hear any AM. This is very much not what I expect, though. It is disconcerting to ‘switch on the AM’ and not hear any effect for many configurations.
1 Like
That’s true that I misread you about pulse wave instead of pulse with, but it’s also true that when you are using a saw or triangle wave and pulse width is 0 you don’t hear any AM. [/quote]
Hey Anigbrowl, I thought you might have misread not only my instructions but also the manual and I can see your point and perhaps it is valid but I also see it as a design implementation that it is to work with only Pulse and Transistor Pulse. A good analogue for instance a Minimoog Voyager Old School it doesn’t have patch recall even though it’s still a Minimoog Voyager. The A4 even though analog does in fact behave some what differently to other true analog devices.
The above could not be more true and is exactly why, after staying away from the A4 for a long time and coming back to it as it’s own thing, it is quickly becoming my favorite machine - even over the OT. It doesn’t behave “like” any other analog I’ve ever played. It has a “thing” and learning it is the key to finding out all the amazing things you can do with it.
1 Like
That’s well put.
1 Like
From the manual:
AM1 (Osc1 AM) engages Amplitude Modulation on Oscillator 1, which means that its signal is switched on/off (multiplied) by the pulse wave of Oscillator 2.
[…]
Note that the waveform and pulse width of Oscillator 2 also affects the result, even if that oscillator is not heard.
Regardless of what waveform is selected, it’s always the oscillator’s Pulse wave that AM-modulates the other oscillator. In other words, this modulation simply means that the Pulse signal switches the other oscillator’s signal on and off.
If the Pulse wave has around 50% pulse width, the effect is normally easily heard, since the modulated waveform is chopped up quite a bit. But if the pulsewidth is at or near its extremes (0% or 100%), it results in either unaffected or completely muted sound, since the signal stays either on or off. This is why it’s not always obvious what the AM does, if you just engage or disengage it and listen. Therefore the manual says that the pulse width setting affects the result.
So what is the pulsewidth when Pulse is not selected as the waveform? Is it the same as the PW knob says, so that it would be the same as if Pulse were selected?
In fact, no! It is more confusing than this. And therefore the manual says that the waveform choice also affects the result.
For instance, the pulse waveform is also used internally to create the PW-modulatable “double saw”. In order to have a normal Saw waveform in the middle of the PW knob setting, the actual pulse width of the Pulse wave changes from 0% to 100% and back to 0% when turning the PW knob with Sawtooth as the chosen waveform. This is different to the 0%-to-100% behaviour when Pulse is selected. (The reason is that the mid position of the PW knob should always give the normal version of the waveform selected. Selecting between basic waveforms shouldn’t have to include setting the PW parameter differently.) A side-effect of this is that turning on the AM might sound identical if the PW is set to mid position when Saw or Triangle wave is selected.
That’s well put.[/quote]
Yes, the AM is an example of something that isn’t instant gratification. Find the useful combinations of all the dozens of parameters that affects the sound of this complex modulation, don’t expect magic to happen just by switching it on! 
(The AM modulator might be a bit restricted and illogical due to all this, but note that the AM-modulated signal is whatever waveform selection done on the oscillator page. It’s not restricted to the oscillator waveforms, it can also be the Ext inputs, the Neighbor synth track, or the filter feedback path!)
14 Likes
Yep - I played with the AM stuff exclusively this morning and both my “harsh/weird” sounds and drum sounds are taking on a whole new depth of dimension.
This box DOES NOT, in many ways, act the same as any other analog synthesizers. It can mimic certain sounds that some classics are known for, but the joy to be had here is to learn the mystique of the A4 and create your own palette of sounds - many of which will be “new” from an “analog”.
I, too, was disenchanted the first time around with the A4 - I will admit it.
I approached it as just another analog 2vco, 2vcf, 2lfo, 2eg machine and found it didn’t do what I thought it would. Things came up and it went away for a while.
This time around, I went immediately at it as something I was completely unfamiliar with and needed to test it out on a basic level to see how all the “parts” interact. This is becoming a very fun machine to make sound with.
Others would benefit from the same approach. Learn from us oldsters.
3 Likes
Thanks for posting /mr. That was an insightful response even though I must admit I had to read it two or three times. 
1 Like
Damn, I just hit some button combo to lose my reply! (func+cue?) 
I noticed that it worked with external audio and that modulating PW was getting some nice sounds, now I know why.
Hmm, like so much elektron gear, it makes sense while i’m using it, but I’m not really verbally understanding what’s happening. …For me that’s a good thing as it motivates me to explore.
Pitch and tuning make a difference too don’t they? Seems like throwing some autobend in (and the different sync options) has interesting results.
I dunno, it seems I can get pretty ring modish sounds using both AM1 and 2. Next time i’m in front of the a4 i’ll try specifically and make a note of what’s going on.
i’m no expert but i think ringmod is quite well defined…
basically you do AM, but then cancel out the original signal that’s being modulated, leaving only the harmonics produced by the amplitude modulation.
Currently the closest I can get is by using the 2nd filter in notch mode to cancel out the signal’s fundamental. But a lot of the original harmonics are still present… so no true ringmod.
However, it already gives some really interesting sounds!
It would be great if it was possible to cancel out the original signal entirely.
Not sure if the architecture would allow such an add-on to be possible via software update. Just wondering. It’s already pretty cool as it is!
and yea thanks /mr for clearing things up!
1 Like
It’s done by the hardware, so no, Software Update - you can’t touch this ™.
The notch method you describe is a creative idea. In fact, pure ringmod is not always more useful than AM (ringmod+original). Some synths, like the Yamaha CS-30/50/60/80, have a continuously variable ringmod amount - from unaffected, via AM, to pure ringmod - and it’s far from every sound that sounds best with the ringmod maxed out. Sometimes it’s nice to have a little tonality left in the otherwise atonal and metallic sound.
1 Like
This reminds me. What’s up with the phase? I can’t get it to work right. LFOs should cancel out when set opposite right?
thanks, good to know… seems some things in the sound generators are done in hardware, and some are hybrid, like the suboscillator tunings (as the 5th subosc was added via OS update)?
but as this is off topic, i was wondering, what do you think of something like a technical Q&A session with you guys, regarding the architecture of the Analog Four?
basically a set of questions from the community, and answers from Elektron R&D? Maybe just a separate thread in the forum would do. there’s obviously the risk that it will just end up as a feature wishlist…
well… i know i’ve got some questions about oscillator & filter characteristics, and also some curiosity about what the FPGA does…
The phase of what?
And what should be cancelled out?
Wow, very interesting read!
Let me add some dry theory about AM: http://www.soundonsound.com/sos/mar00/articles/synthsecrets.htm
The phase of what?
And what should be cancelled out?[/quote]
I was thinking if I had LFO A +B with opposite phases modulating the same parameter it’d do “nothing”.
Or I was looking to hear if there would be interesting distortions by modulating phase at audio rates, while already using the LFO at audio rates. It often seems I can only hear phase changes when manually adjusting the setting.
I bring it up because I find that using the LFOs at audio rates on the oscillator’s pitch, pulse width, waveform, anything! really makes the effects of the AM more noticeable.
There are some other things don’t seem to respond to audio rates too? For instance, sync amount doesn’t seem to respond much?
You mean LFO 1 and LFO 2, and setting one them out of phase with 180 degrees?
Provided that you use sine or triangle LFOs (the other ones don’t cancel each other out), and trig the LFOs simultaneously (to give them a well-known phase difference), and modulate the same parameter with the same amount, I’d expect the result to be quite close to nothing, yes.
What do you expect to hear when modulating sync amount with audio rates? I’d be surprised if it sounded very radical, since sync amount is only used to determine what happens sync-wise when an oscillator resets. Everything that happens to the sync amount between these sparse occasions are irrelevant.
1 Like
Thanks! That makes sense. On further experimentation I was able to hear sync 2->1 amounts at audio rates, but it sounds way cooler using an envelope. 
I was reading about casio phase distortion and want to see if any of that applied… 
…
With AM sometimes I find it easier use the velocity mod to change the pitch of one (or more) of the oscillators and parameter lock velocity. It’s fun to play with having different pitch depth amounts for each oscillator.
1 Like
You can do this if the modulating oscillator is set to pulse or tranny pulse then turn the PW knob anti-clockwise, the further you turn the less you hear the original signal.
In the positive direction, the more you turn the more you hear the original signal, so it’s good for controlling AM amount.
That’s how it works on mine so if it’s malfunctioning I don’t want it fixed!
Thanks for replying to this thread, /mr. It’s always interesting to read about the intricate details of this wonderful machine.
1 Like