I have an enduring fondness for analog modular synthesis. With its beckoning cornucopia of knobs, patch cords, and blinky lights, a big modular system is a kind of playground for grownups. And the modular synthesizer scene is booming right now.
But do I really need to rush out and spend thousands of dollars on a modular instrument? Or am I just indulging in gear porn?
To challenge myself on this question, I sat down this morning and wrote (if “wrote” is the right word) a modest little piece using Csound. Csound is software, and it’s a free download — cost, $0. Three minutes of blissful minimalism are not part of my usual style, but minimalist drones are arguably the most prominent feature of music made with modulars, so let’s see what we can put together without benefit of knobs or patch cords.
The point of this piece, I suppose, is that it utilizes two different kinds of physical modeling synthesis. (The “bowed string” tones are a wave guide opcode.) Physical modeling is not something that a modular synth can do. This piece has upwards of a dozen random number generators crunching away. To duplicate that in hardware would require, at the very least, three separate random generator modules, and they’d have to be able to generate randomized spline curves. Oh, and two of the three instruments are engaged in somewhat randomized stereo panning. Would you like to buy two separate stereo panner modules in order to be able to do that?
Then we come to the reverb. Not a great reverb, to be sure, just Csound’s stock reverbsc opcode, but an equivalent hardware reverb module would cost you several hundred dollars. And can we talk about the pitches? Pure just intonation. I have the vague impression that there’s a quantizer module out there that will do non-equal temperaments, but you’d need four of them to do this piece, and that would set you back $1,200 or so.
This particular piece happens to use only a three-row, 8-step sequence, which cycles three times. You just about can’t buy a hardware step sequencer with more than 16 steps … but of course if you happen to need a 29-step sequence, or one with 129 steps, Csound is happy to oblige. It’s all just data and variables.
The accusation that might be leveled against a Csound “modular” in software is that using it is not intuitive. But I’m starting to be skeptical of that assertion. A hardware modular is anything but intuitive, until you learn how to use it. On the contrary: It’s a bewildering maze of esoteric functionality.
Once you devote some time to learning how Csound works, is it any less intuitive? I’m not convinced of that.
Oh, and by the way — if you’re interested in exploring Csound but not sure where to start, I wrote a book about it. The book is on Amazon.