I’m currently reading Designing Sound by Andy Farnell (a recommendation I came across here recently - though I forget by who :/), which explores synthesis (largely non-musical) using Pure Data. He has a section that discusses the waveshaping function tanh being used to squash a sine wave and introduce harmonics. Because tanh is nearly linear close to zero, you introduce fewer harmonics when your input signal is quiet – and more harmonics as your input grows louder. I am guessing that this is not a novel idea to many of you.
In pd (and in many other programs, I’m sure), you could draw a waveshaping function that exaggerates this effect much more than tanh – something that resembles y=x around zero and is very different from y=x for larger values. A great example of this is discussed in this video, starting around 9:20: https://youtu.be/sg2TpxQ9CbU
This was sort of a light bulb moment for me, as I’ve mostly thought about distortion – at least in the context of electronic music – as a static effect instead of what it actually is, which is an effect that reacts to dynamics! This was obvious when thinking about, say, playing an electric guitar: your amp or overdrive pedal will react to how much you dig into the strings with your pick – more amplitude leads to more harmonics.
For some reason, I haven’t been thinking about this same interplay between dynamics and overdrive/distortion in the electronic music realm. On my Digitakt, I crank up the overdrive to make a particular sound or sample a bit crunchier or grittier. But then I set that sound’s level to whatever I want in the mix, and it is somewhat statically used from there. Boring! Maybe I will program some hits to be louder or quieter, but I’m not really thinking about modulating the sound with respect to amplitude. I think of amplitude as one parameter I can modulate, but I don’t generally think of amplitude as a way to modulate other parameters (like timbre, for example).
Does this make any sense?! Does the overdrive on my Digitakt respond to velocity / sample volume in the same way a Tube Screamer reacts to how hard you play the guitar running through it? How do you use amplitude dependent harmonics in your music? I’m curious about gear used to do this, techniques employed, and your thoughts in general on the wonderful world of amplitude dependent timbre!!
Not sure about Digitone, but Syntakt’s analog drive absolutely responds better to louder signals (and can then be filtered or mixed down via the amp), and there are often complaints that the drive doesn’t do much or is “too tame,” which is true when not enough signal hits it. Interesting read, I’ve been looking forward to the chance to read that PD book, saw the same recommendation I think.
Interesting. I’ll have to do some experiments on the DT. I am not really crazy about the DT’s overdrive, but maybe I’m not hitting it hard enough.
Another interesting thing about a custom waveshaping function is you could make it do the opposite of what I mentioned above: more harmonics when quieter and fewer harmonics when louder. Or a sort of unipolar behavior: more harmonics for the wave in [0,+1] and fewer in [-1, 0], for example.
I suppose some of this can be achieved by matching your amp envelope with your filter envelope, but the waveshaping approach feels different and easier to program in some ways.
The book is very cool. It is also very technical. The beginning is all about the physics of sound and is a bit over my head. But the book lends itself to jumping around if you like – reading the bits you find interesting or useful and skipping topics as you see fit. I think it could be an in depth study or just a reference, depending on what you want from it.
I’m just starting out with building my own overdrives/distortions/waveshapers using Max. I don’t really have much to add to the conversation right now, but I’ve always enjoyed messing about with the relationship between amplitude and harmonics.
I do think, having recently gone back to using computers, that there is a lot of laziness around when it comes to distortion. Effects just seem to be something that is added on top of a mix or track for a lot of people, whereas I’ve always made it a part of the sound design, a byproduct of live recording I guess, but yeah, it probably explains that divide between people who think the Syntakt has a shitty distortion and people who know how to get the best out of it.
Distortion isn’t the icing on the cake, distortion is the cake.
Before the existence of distortion effects, distortion for electric guitar was generated by simply increasing volume. People would drive the volume so hard the wave would clip and the sound would distort, which adds overtones (harmonics) – literally overdrive. So, yeah, amplitude definitely plays a part.
Great book, by the way. I loved his quick review of physics in the beginning.