On some complex oscillators, the modulating oscillator can directly wiggle the wavefolder knob, like the DPO.
1 Like
Yeah I think the confusion is over where the modulating happens - like it wouldn’t be possible to tap into the space between the oscillator and the wavefolder and somehow modulate it there - if you modulate the oscillator, you do it at the frequency input of the oscillator (or later in AM section, but still).
1 Like
It isn’t obvious to me that the order doesn’t matter with something like this. The order of distortion pedals (for instance) sometimes matters.
I guess I’ll see if I can mock something up (I am still pretty unskilled with Audulus) to see if wavefolding and amplitude-modulated signal sounds different from amplitude-modulating a wavefolded signal. My gut feeling is that it does make a difference. (I think the visual layout of the easel and the Buchla rack modules seems to group the wavefolding with the the carrier oscillator rather than separately, but that isn’t definitive.
Re: reverb. I understand that the reverb is an important component. So, I have tried to take that into account in playing with WestCat and Rasums. I’ve also explored the Arturia demo version with and without the reverb to get a sense of what the reverb is contributing.
You could well be right. I guess an empirical test will be the way to go to see if it even matters.
1 Like
This would sound different. I didn’t know you were talking about AM. My guess is if you flip it into AM, it applies that modulation between the oscillator and the wavefolder. If you do FM, it modulates the frequency input.
cortini drone with easel touchplate controller.audulus (378.7 KB)
I took a swing at making a cortini style drone with the WCO and found that hard sync and filtering off the high end were important, then I added a little drive and some reverb. I also made a easel-style touch controller with wide-ranging pitch control. I also added a regular modulation out and individual knob outs, so you can use it like a performance patch macro, just like the easel.
edit - I added a quick demo video of this patch to the top post.
5 Likes
Sounds as close as you’d expect putting Audulus up against ~$3k of gear! Great job
2 Likes
Also the WCO intentionally looks for new tones with the complex oscillator format, so I’m not shocked it doesn’t perfectly replicate the Bucharest (lol, autocorrect) Buchla easel. But with a little time I think @espiegel123 can draw out the desired sounds.
1 Like
@robertsyrett: Thanks so much for putting that together. It sounds great.
I really appreciate that you did that and you and @biminiroad being patient with my questions.
The patch sounds great. I suspect (but don’t have the chops yet to prove it) that with some judicious tweaking of the internals that WestCat could be made even more awesome and responsive.
@biminiroad: I am not so much looking at it from the perspective of a precise clone, and I think that with a little judicious adjustment to the internals that the excellent WestCat Oscillator and Rasmus can be made even more awesome.
Something that I notice in Todd’s tutorial video (which one can also hear in Arturia’s software Buchla clone whose demo I am exploring) is that the range of effect/responsive of the AM slider is quite a bit different from what we have hear as is the degree of wavefolding.
I don’t yet understand enough about what goes into a good AM patch to know how to tweak the awesome stuff you guys have done, but I think some fine-tuning could go a long way.
Listen to the the depth of impact as Todd brings up the modulation fader in the video.
I am going to try to figure some stuff out to see if I can make a useful contribution – but I am pretty ignorant of how all the DSP stuff works behind the scenes.
I hope you won’t mind a few more questions about CO architecture in a hardware synth. If we think of the CO as being made up of two oscillators (let’s call them Modulator and Carrier) plus a wavefolder, my questioare:
- what sort of inputs (internally) does the carrier take?
- Is FM being done by modulating the frequency input voltage to the carrier oscillator?
- how is the amplitude modulation done? Does the oscillator have an input voltage that influences amplitude or is there a VCA that comes after the oscillator that is amplitude modulate.
Thanks so much for all the info and great patch.
1 Like
Usually the sine output. The continuous change of a sine wave leads to the most interesting sidebands in FM. In the WCO I have a sine that feeds back on itself for the modulator though, so It also offers rounded saw and ramp waves to the carrier.
Usually FM is referring to linear FM. The easiest way to do that in Audulus is to have an index (say 800Hz) an multiply the incoming audio by that index and add it to the Hz value going into the oscillator core (usually a phasor node). So when the FM knob is all the way turned up the signal can vary the frequency up to ±300 Hz. When this happens at audio rates it looks more like wave shaping than a continuous change in frequency because it’s all happening so fast.
You first have to adjust the modulating signal so that it is all positive values and then you multiply the output of the carrier by the modulator signal. The multiply or the level node can be thought of as a VCA in this context.
also @biminiroad
I am making progress in this exploration. So back to my earlier question of whether the order of the wavefolding and amplitude modulation makes a difference, I can now say with confidence that it does, and the amount of difference depends on the method used to perform the amplitude modulation.
I put together an (ugly) patch that lets you switch between 7 versions of AM (6 are done by feeding the modulator into a VCA’s CV input and the seventh is the algo pulled out of West Cat’s internals). I used RS’s wavefolder module. There is a toggle that lets you swap the order.
The patch is pretty primitive but useful I think. You might find it interesting to explore. Let me know if more explanatory text on how to use it would be useful. I think it probably pretty obvious.
What really strikes me is how different the AM result is with the different VCAs. To my ear, a couple of those VCA’s provide a particularly beefy result.
Here’s the patch in case you are interested in checking it out:
AM WF Order Toggle 0.2.audulus (259.7 KB)
In using it, I also notice that the wavefolder colors the signal even when it is turned all the way down. I wonder if it might be worthwhile to slightly complexify the way the fold amount knob works so that when it is all the way at 0, you get the original signal without coloration.
2 Likes
Super cool patch! Was just lost in it for 5 minutes. I think I’ll add in those VCAs tonight to the new library 
2 Likes
They are pretty awesome and really provide some great variations in palette. Now, I have to figure out which one I want to use for AM in the CO that I am putting together (which I will, of course, post if it ever gets into a usable state).
1 Like
@robertsyrett and @biminiroad : I am kind of fuzzy on how linear FM works (as opposed to the other type – expoenential?) if one feeds the output of one oscillator to the frequency CV input of another is that linear FM?
Do you have any Audulus patch recommendations (preferably not to hairy) that would be good for me to study? Now that I have a bit of a handle on getting the AM responding along the lines of what I am looking for, I want to explore adding FM to my version of a West Coast oscillator. (Would be totally happy if any of my results made there way into Rasmus or WestCat or a related system).
The normal frequency input for most Audulus VCO modules is 1 per octave, and for Eurorack it is 1 volt per octave. Both of these inputs are exponential in nature. For example for the Audulus standard 0 = 440Hz, 1 = 880Hz, 2 = 1760Hz etc. If you apply a modulation signal to this type of input, it will also change the frequency exponentially. A linear FM input on the other hand makes a linear change to the output frequency so if for example 0 is 440Hz and 1 = 550Hz then 2 would equal 660Hz etc. The actual ratio between the modulator change and the frequency change is usually adjustable (modulation depth or level).
1 Like
Thanks for the explanation. I just found a good demo on YouTube demonstrating this. So, would I be correct in believing that the FM mode of the music easel is exponential FM?
Is Yamaha DX-style phase modulation essentially linear FM?
Do you know if there is an already-existing Audulus project that isn’t very complex that would be good for me to study to see how to accomplish linear FM?
I am thinking that I might want to make both linear and exponential FM available in the Audulus Easel that I am trying to put together.
You can peek inside the VCOs to see the linear modulation section. The index value you might want to play with - make it larger and you get wilder tones, but unless you’re using a TZFM oscillator, you can cause the oscillator to stop if during the negative swing of the FM it pushes the oscillator down to 0Hz.
For example, if you’re FMing something at 440Hz with a 500Hz swing, for a portion of the swing, the primary oscillator will be stopped. With a TZFM oscillator, the wave actually reverses direction.
Different types of FM.audulus (263.1 KB)
1 Like
The FM on the Music Easel is Linear, but not through zero. @biminiroad’s explanation above should help you out.
The FM in the WCO is the same as the Buchla, it has an index of 800Hz, or a “swing” of 1600Hz since you send a bipolar signal signal into it which can be positive or negative. I follow up this expression with a clamp expression to keep the oscillator from creating negative frequencies.
Hope that helps.
1 Like
Thanks. Very helpful. I am curious. Any idea how they accomplished linear FM back then? Something that I read implied that linear FM wasn’t practically viable for sound synthesis until the late 70s as it was difficult to accomplish in analog. Of course, the person could have been totally mistaken.
1 Like
So I think the actual area of linearity wasn’t so good until the late 1970’s.
From wikipedia:
“Nonlinear elements such as transistors tend to behave linearly when small AC signals are applied to them. So in analyzing many circuits where the signal levels are small, for example those in TV and radio receivers, nonlinear elements can be replaced with a linear small-signal model, allowing linear analysis techniques to be used.”
Buchla was a NASA engineer for years before he quit his job to make synthesizers and be the sound engineer for the Grateful Dead. He was already familiar with Linear FM from radio broadcast and transmission that he would have had to work on at NASA. It’s likely that he just reused the circuits in the new context of audio since the earliest Buchla FM oscillators have been linear FM.
What you may want to do to make the FM sound more “authentic” is to make your input less linear. That would mimic the components available in the 1960’s and early 70’s which would distort at the extremes.
3 Likes
Thank you so much. I really appreciate these insights.
1 Like