Saturday, September 24, 2016
Adventures in Synthesis: Spring Reverb Feedback, Sequential Switch, and Wild Modulations
Published on Sep 24, 2016 Chris Beckstrom
"The other day I was experimenting with a sequential switch, using it to cycle through various waveforms coming from a single oscillator. It's a bit like the classic dubstep wobble bass, where the divisions of the wobbling change every bit. Of course since this is a modular synth, things can- and do- get much crazier.
There are three four voices in this patch: the bass drum, the 'hihat'-sounding things, the the wild/glitchy oscillator, and spring reverb feedback. The master clock is coming from my computer, which is sending MIDI notes to my Arduino oscillator. I use the gate output of the Arduino to clock the whole patch. I was using a regular analog oscillator for that, but something weird was happening and its pulses were not even (it sounded drunk).
The bass drum is just a twin-T (filter!) module pinged with a trigger from a gate to trigger module which in turn receives the master clock. The bass drum goes into a channel on my outboard mixer.
The "hihat" sounds are two different modules: my PT2399 delay and my CMOS noise module. Both go into separate VCAs and are triggered/modulated by different rhythms. The CMOS noise goes into a VCA which is turned on/triggered by the output of an envelope follower's gate output. The envelope follower is listening to the output of the crazy saw wave, so the more active the saw wave is, the more active the CMOS noise is.
The delay unit, when nothing is plugged in, spits out a lot of hissy noise (probably due to a lack of low pass filters in the feedback circuit, oh well). I find this very useful, especially since I don't have a dedicated "noise" module. The delay noise goes into another VCA which is triggered by an AND logic, comparing the clock and a random square wave oscillator and only pulsing when both are HIGH. A feature, not a bug, of my vactrol-controlled double VCA is there is bleed between the two (because my electrical tape does not keep out all the light). This means that sometimes the CMOS noise is also triggered by the VCA controlling the delay noise, and vice versa. This keeps things interesting and a little random! Both noises are mixed together and sent out to a channel on my outboard mixer.
Four waves from one oscillator are patched into the four inputs of a sequential switch, and one of those waves (the pulse) clocks the switch. This means that for every cycle of the oscillator, the sequential switch moves forward one step, allowing whatever wave is plugged into that step to come out the output. This has the effect of, basically, a different wave each beat. A 10-step sequencer, clocked by the master clock, has its CV sent out to a passive matrix mixer, which is then sent to the CV input of the oscillator sending waves to the switch. This means that not only does the frequency of the oscillator change every beat, I can also change how much the sequencer affects this change.
The output of the switch is patched into the CV input of a saw wave. Changing the amount of CV the (what is basically an) LFO receives changes the wildness of the saw wave. The saw comes out, goes into a VCA whose amplitude is modulated by an envelope generator (which is pinged by the master clock), and then finally goes into the outboard mixer.
The last voice is spring reverb feedback. The output of the spring reverb goes into a crunchy amp, then into another less crunchy amp, then back into the input of the reverb. When both amps are turned down, we just hear the reverb. When both are turned up a bit, the reverb feeds back and generates a tone. Depending on the gains of each amp, this tone can have different qualities and pitches. It's quite difficult to control, but I think that just means I need to practice more. This fourth voice is alligator clipped to a 1/4" jack (because I ran out of 1/4" outputs on the modular) and sent into the mixer.
There is also a fifth voice, I suppose, a sort of drone. I have some sort of grounding issue with my modular (maybe because it's plugged into an extension cord with only two plugs) which is causing a low frequency oscillation. For this patch I consider it a feature, not a bug, and just keep in there. A lot easier than tearing down my patch and trying to find the source.
I start this video scratching the reveb springs with a screw, which I think is a pretty cool sound. At one point I accidentally toss it aside, so I just use my fingers to play it.
When I play the patch, I'm mostly changing a few key things: The CV amount of the LFO, the pitch of the LFO and the main oscillator, the gain of the reverb feedback amps, the delay time of the delay noise, tone settings on the CMOS noise module, and the input level of the envelope follower (lower level = less activity).
At the very end I smack the outboard mixer to make its internal spring reverb boing a little bit."