Modular synthesizer hardware is a wonderful thing. I love being able to reach out and grab a knob. I love being able to plug in a cable and hear the results instantly. And some of the newer modules in the euro-rack world are amazing! Even so, there are certain things you might want that no hardware module will do, as far as I’m aware. Other things are available in hardware, but only in relatively primitive form or at considerable expense.
On the other side of the coin, Csound is an immensely powerful music synthesis language. There just about isn’t anything it can’t do sonically. But — no knobs and no patch cords. Making music with Csound is about as spontaneous as cooking breakfast. First you get the eggs out of the refrigerator….
But what if there was a module that would run Csound code? Then you could have both the spontaneity and tactile gratification of hardware and the versatility of software. A new company called QuBit is releasing such a module. It’s called Nebulae. But while I wish them great success, Nebulae is really quite limited in some important ways. It only has one output jack, for instance.
After mulling it over for a week or two, I broke down and ordered the Expert Sleepers ES-3 and ES-6 modules. Also a PreSonus 1818VSL audio interface. Why the combination? Because the ES-3 and ES-6 connect to a computer using ADAT lightpipe cables. The 1818VSL is one of several relatively affordable interfaces that has ADAT connectors. I had to order a pair of cables too; the cost adds up. But I already have a perfectly nice 5-year-old Macbook Pro that runs Csound, and I know how to program Csound.
After a few hours of testing and troubleshooting, I seem to have a stable system. My MacBook is now, functionally, a euro-rack module (though not, of course, mounted in the rack) that will do absolutely anything. It has eight outputs to the modular synth (via the ES-3), which can carry either audio or control voltages. It has six inputs back from the modular (via the ES-6) that can, again, carry either audio or CVs. The 1818VSL is sitting at the upper left; the ES modules are installed at the lower right corner.
Why integrate the MacBook with the modular? Here are a few modest ideas, most of which I haven’t yet tried out.
- Sample playback and sample manipulation as sound sources.
- A step sequencer with dozens of steps, capable of chaining patterns and switching patterns under voltage control.
- Multi-stage envelopes.
- A counter that can send a fixed number of gate signals and then stop.
- A multi-tap delay line or a reverb.
- Equal-power panning (a sore spot in most of the modules I own).
Troubleshooting such a system requires a bit of care. I discovered, for instance, that the ES-6 does not send precisely zero volts back to the computer; its lightpipe outputs have a bit of DC offset. (Adding trimpots to the circuit board for calibration would add considerably to the expense.) So when sending an audio signal over to Csound, I have to apply a basic highpass filter to it, to get rid of the DC. With a gate signal, I’ve found that downsampling it from audio rate to control rate and then using if/then logic to drop low-level values to zero is effective. If I want to center an LFO signal arriving from the hardware, I’ll probably have to figure out manually what small value to add or subtract.
The same thing turns out to be true running the other way. Not only do the ES-3’s output have a small positive or negative voltage, which is different from one output to the next, they also don’t seem to transmit exactly the same voltage in response to the same incoming digital audio value. When I set up a step sequencer in Csound and ask it to play three oscillators on different ES-3 outputs, I find that each output may need a slightly different step size. With Csound set up so that 0dbfs=1, the step size for an equal-tempered (12EDO) half-step is about 0.00795, plus or minus 0.00003.
This stuff is not especially hard to figure out, but it would probably baffle someone who was new to the technology.
Of course, there’s plenty to explore with the modular synth itself. I may not need to play the Csound card too often. But it’s good to know that when I need it, it will be ready. At this point, I’m not even sure what kind of music I’ll be making with this system; I’m still just getting my feet wet. What I’m telling myself is, you don’t buy a violin and expect that in a couple of weeks you’ll be making wonderful violin music. It takes time to learn an instrument.
Beyond that, in the case of the violin the musical goals tend to be fairly clear. An aspiring violinist doesn’t have to compose sonatas for her instrument, nor does she need to discover a style in which to compose the sonatas! There are some known idioms for modular synth music — it’s not a blank slate. But the idioms are by no means settled or concrete. We’re making it up as we go along. So the scope for fresh discoveries is quite broad.