July 25-26, EEGsynth HQ
Samon Takahashi and I have been working on a second EEGsynth musical set, to accompany Steady-State. For a set to work, a lot has to be in place.
Sound preparations
For months, I’ve been preparing the sonic and control elements of the patch. By now I have a number of Eurorack modules to choose from, and over the years I’ve been buying and selling quite a bit, so I also know pretty well what’s out there. And yes, one oscillator doesn’t sound like the other, at least not in Eurorack, and certainly not once you combine it with modulation: AM and FM, wavefolding and ring modulation. Then there is the form factor: more functionality typically means more knobs, which gets fidgety and isn’t ideal when playing live. Larger, more comfortable modules take up space in your rack, which is very limited, so you have to sacrifice extra filters, modulation and control functions. And then there is of course the composition: melodies or drones, minimal analog tones or digital samples, rhythms or random events? You end up making a lot of decisions, and for every new feature or sound you have to subtract another. Most importantly, you have to enjoy the result. And it has to be optimal for creating interesting, complex and responsive brain-music feedback loops. Developing a patch that can draw a person in, respond to their brain state, and create an immersive experience for the participant and the audience, well, that takes practice, and a focus on poetry: making the aesthetic decisions that let a piece work as a musical performance.
Two sides – two stories
unSteady is planned as the B-side to Steady-State on a yet-to-be-recorded LP. Steady-State is explicitly political, aggressive, uncompromising and forceful, while unSteady is intended as its counterpart: a wide, embracing horizon, seductive and psychedelic. Where Steady-State is a claustrophobic wall of noise, unSteady opens up the space with binaural movements and flow. Where in Steady-State the operator (me) tortures the subject (Samon) into oblivion, in unSteady the operator supports the subject, guiding them towards a new unity. From destruction and dissociation to a new connection and wholeness.
The composition rests on two minimal sound sources (voices) moving through space, independently or in phase. Over time sonic textures are built up, and the two voices are increasingly blended, resonating with and distorting each other, so they feel like they are playing off one another. Increasingly the space between the voices takes over, with quick stereo filter modulations and tube distortions stretching and breaking the space into fragments that come together and disperse again. The characteristic sound of Steady-State is largely repeated by using similar components, except for swapping in a clean stereo filter, but the stereo elements and dual voices make it distinct. All of this is of course just theory until you try it out and start to identify which modulations work, and in which order, to create a ~15 minute composition. That was the aim of the weekend session.
Mentalab Explore Pro
It was also the opportunity to work on the technical side: the EEGsynth patch and the new EEG hardware, the Mentalab Explore Pro. I was excited to set up the system on my new MacBook, since the Explore Pro is now supported on macOS, something I discovered after the Hackathon last April. After a quick fresh install of EEGsynth and DesktopExplore we were ready to go, or so I thought. For some reason I was not able to push or receive an LSL stream on the MacBook. Long story short, I ended up using my old Windows laptop for EEG acquisition with the Mentalab DesktopExplore, and from there pushed LSL over my home network to the MacBook running EEGsynth. And even then the stream was unstable and needed restarting every so often. Of course, this could just as well be an unstable home network or the less-then-actively maintained LSL library. Robert previously had to make a few fixes to get it working for us, so I wouldn’t be too surprised if the issue had to do with the LSL implemention. When I explained my problems to Mentalab, they responded quickly (as always) and assured me LSL should work absolutely fine on the MacBook. I’ll be looking into it with them to find out where I’m going wrong.
I also mentioned to Mentalab that I find the way of pairing the device a bit awkward: you have to enter the number on the back of the device directly, without ever seeing it show up among the system’s Bluetooth devices. When you’re out and about in a dark studio, bar or venue, or stressing before a performance or recording, I easily go in autopilot. Even if you remember to look at the device for the code, you might not be able to read it (it’s very small print). Mentalab explained that they’d like people to be able to pair without diving into any system settings. Fair enough, they have a point. So here is my solution: a large printed code. And on top of that, a clear indication of which way the electrode plug has to be connected. Of course, this is only an issue when you re-install on a new device, but in public performances there is no way to be too prepared…
Recording
As always, it took two days of work to get a couple of minutes of good recordings. We went back and forth over the composition, discussing, designing, repatching before diving back in, Samon’s brain synchronizing with the sound, me dealing with the chaos of the patch, the EEG calibration, and the technical challenges. I think it is going to be a great recording. Here is a teaser: