bad hair day, good synth day

The Medusa arrived yesterday — a bit late, box damaged, but synth intact.

It’s the sort of synth where, if I tried it for five minutes at Knobcon or something using its factory presets, I’d almost certainly have walked away and not looked back. Probably in under five minutes. The presets are, for the most part, awful.

But once I figured out what I was doing with both the synth section and the grid, and started exploring a little… this is one extremely cool piece of gear.

I understand why so many of the demos and several presets are bland: the analog oscillators only do the classic simple shapes, the digital ones are raw and basic but not in a particularly gnarly or exciting way, and the filter doesn’t have much character of its own. It’s a clean slate and you have to do your own color mixing, to mangle some metaphors. With judicious tuning, use of the oscillator sync, FM, copious modulation sources and all sorts of clever tricks available with the grid sequencer/performance controller, if it sounds boring it’s the musician’s fault. Synthesis means the assembly of smaller parts into a whole, and that’s what has to happen here. It can certainly do beautiful, brutal, quirky or scary, and it’s not even difficult to get there. It can’t do everything, but that’s why I have more than one synth… 🙂

As with most synths it really benefits from some reverb or delay. Every other Dreadbox synth has one of those built in; this doesn’t. This is not a problem for me at all, but it does contribute to some of those lackluster demos since people seem to think it’s “cheating” to add external FX.

The grid is brilliant. Note mode lets you select (or create) a scale, and also set the layout to the interval of your choice (or a “guitar” mode); it’s a combination that is really great for jamming, thinking differently about intervals and chords, or playing runs that would be (almost?) impossible on a piano-style keyboard. The velocity setting doesn’t work very well with the onboard synth and honestly is best turned off, but pressure and X/Y movement can be assigned to various parameters. Grid mode is a sequencer that’s mostly intuitive (with a couple of odd quirks) and lets you assign not just notes and chords to any individual step, but ties, steps that randomize on each playthrough, and parameter settings. Between that and the paraphonic modes — and the way those modes interact when you’re using FM or sync — it’s definitely got some unique capabilities.

The Lyra-8 was inspiring and influential, leading me toward the 0-Ctrl and certain techniques. But I do think the Medusa is the better choice for me now and for the future.


Last night I read another article that speculated on plausible scenarios involving the November election. Overall, it seems clear the Trump campaign strategy has been to cast doubt on the legitimacy of the election — an old GOP favorite, where “voter fraud” is used as the excuse to disenfranchise voters in Democratic-leaning areas (who are proportionally more black and poor than not) but cranked up to 11. And of course this fits right in with the cultish, Orwellian strategy of calling the news media and science into question so that the Regime can lay sole claim to the truth. Trump, unless defeated in an absolutely crushing landslide, is almost certainly not going to concede but will attempt to bend space and time around the black hole of his ego.

Reading the article was a mistake.

I dreamed last night of being at a college for some kind of artist residency thing in electronic music — COVID wasn’t a concern in the dream world — when it was announced that the school was closing due to widespread outbreaks of election-related violence. It was perpetrated by right-wing militias and by police forces (which are not that far from being the same thing), but of course blamed on antifa and George Soros…

The scene changed to the school’s auditorium, and the woman in the seat in front of me pulled a small pistol from her purse and started pointing it at people and waving it vaguely around while spouting off some QAnon garbage and accusing me of being an anarchist. And that’s when I woke up. (I realized later it was Patricia McCloskey, of the couple in St. Louis who threatened protesters by waving guns at them illegally.)

Sigh. 40 days until Election Day. 118 days until the next Presidential term begins.

I went ahead and traded my Lyra-8 to someone directly for a Medusa. After both were en route, I got a really weird set of messages from USPS about how the Medusa’s address had been changed and it was being held for pickup in the post office in the sender’s city. So that was another recent worry: had I been scammed or something? After a few moments of thinking about it, it didn’t really make sense. I figured it was a mistake, and yes: the sender had accidentally reversed the To and From addresses when buying the label, someone from USPS figured it out and fixed it, and that looked really bizarre to their tracking system. I’ll still be relieved when it arrives, and also happy I hope. I think Medusa is going to work out very well.

I have also gone for a DAFM Synth Genesis, which sports the Yamaha YM2612 “OPN2” FM chip — as found in the Sega Genesis, and also that MegaFM synth I had been thinking about. This is smaller and significantly cheaper, and though it doesn’t have a dedicated slider for every parameter, I’ve heard it’s pretty friendly to work with. I’ve already familiarized myself a bit with how the synth engine works, since Plogue Chipsynth MD is an emulation of it. It’ll be nice to have the real deal and all its wonky fuzzy noise for drones and textures. If I seriously fall in love with the sound, it might be worth reconsidering the MegaFM (and possibly setting Akemie’s Castle aside).

don’t look now

Good thing I didn’t jump immediately for the Quadrantid Swarm, because I have another thought.

Polyend Medusa Black Analog and Wavetable Synthesizer with Sequencer  (Limited Edition) | Sweetwater

The Polyend + Dreadbox Medusa is another one of those synths I didn’t pay that much attention to upon first release, because of its price and my assumption that I didn’t want a grid-based sequencer.

The price of a new Medusa now is almost down to half what it was and the “special edition” panel shown here also looks better than the original. And as it turns out, that grid-based sequencer is also a touch controller, and it is mighty.

In “note mode”, you can select a musical scale as well as the interval between rows — so you can set it up for handy chord fingering patterns or octaves or whatever makes sense to you. Each of the 64 pads also sends X, Y and pressure modulation signals which can be assigned to synth parameters.

In sequencer mode, each pad not only stores a note value but parameters. If you want, you can have each pad play a completely different sound, set up like a drum machine. Or you can use a row or column to sweep across parameters. Or assign the notes you want to use in your composition to a few pads, and simultaneously play those and parameter pads. You can copy/paste to arrange assignments into a sequence, or just step sequence more normally. I’m not sure whether you can sequence the sequencer’s own clock rate from there, and the sequencer’s motion seems to be strictly linear, but it’s still a pretty great concept.

The synth itself is more conventional in some ways, with 3 analog oscillators (with sync and expo FM) and 3 basic digital oscillators (with non-morphing wavetables). But you can assign them to play in unison, 3-voice mode (each voice is 1 analog + 1 digital) or 6-voice mode — note rotation as you play can lead to some interesting variations. The synth has a single multimode filter, making it paraphonic like the Microfreak. There are no built-in effects, which is a difference from other Dreadbox synths.

There are, unfortunately, some really bland-sounding demos out there which have turned a lot of people away from the synth. But there are also some stunning counterexamples — this is not a boring synth, but rather one that requires and rewards some sound design effort, and at least a touch of reverb or delay. Among actual users, opinions seem to be split — some absolutely love it, some just never quite get along with it.

Many of those best demos are using it exactly as I would, for drones and ambient stuff, and I kind of suspect I’m more likely to be in the “love it” group.

Quadrantid Swarm is a synth made to give instant gratification with weird, angry, dark and/or metallic sounds. It does have semi-modularity and that sweet spring reverb going for it. But overall it is much less expressive and flexible, and I think in the long term, the Medusa would likely serve better.

So as soon as either the Lyra or Reface sells or I encounter an extra-good deal, I expect I will go for a Medusa.

something completely different

I don’t write that much about the games I play, partly because there’s just not a lot of change there. I tend to grab onto a game, or at least a genre, and stick with it for a good long while if I don’t burn out and get bored quickly.

What I’ve been playing for the past several months is Dirt Rally 2.0, Noita, Bejeweled 3, and (in the past month-ish) Guild Wars 2.

Dirt Rally 2.0 is currently doing its second World Series event, in both rally (on trails) and rallycross (on a mixed-surface gravel/asphalt track) categories. The finalists will compete for a $20,000 prize pool, and the rallycross winner also gets to test-drive a real-world electric car that’ll be used in the new eRX2 rally category. But everyone who participates will unlock a car in the game, so that’s nice.

I’m not a serious racing simmer, I’m a filthy casual who plays with an XBox controller, “bonnet view” (outside the windshield in the center of the car), and usually with automatic transmission enabled. I had low expectations going into this. But I’m very happy with coming in just below the top 10% overall, so I guess this post is a brag of sorts.

That still puts me a few light-years away from any sort of prize, as each qualifying round will take only 6 drivers in the final and send two of them to the semi-final. But hey, I don’t completely suck at this game after putting in (according to Steam) 261 hours in it 🙂

I did much less well in the Rallycross qualifying, squeaking into the top third. But then I don’t find rallycross as much fun to drive as roads and trails that go somewhere, which is a bit reminiscent of the off-roading I used to do with my dad in the 80s, with (at various times) a rail buggy, go-kart and motorcycles. Just a lot faster than we ever used to go 🙂

still not an exact science

My ethnicity estimate has been updated again, as it was in mid-November last year. Here’s what it said before:

And here’s the new version:

However, in the details it does say that Scotland could range from 7%-43%. Most of the rest are within a few percent one way or the other.

DNA and statistics both being what they are, adding more data to the pile and re-evaluating everyone doesn’t necessarily make for a more accurate assessment for any specific individual. Mostly this is just a bit of fun.

3/4 gaffer tape covers the ugly graphic strip on the Microfreak just fine. The downside is, it looks like there’s a piece of tape on the synth. It’d be nice if there was something a little more professional looking. But really this is just splitting hairs.

I’ve decided to let go of the Lyra-8. I like the way it plays, but it is fairly rigid about the way it sounds, and I’m just not into it anymore.

The most likely replacement is the eowave Quadrantid Swarm. Like the Lyra, it has 8 touchplates for oscillators that can be individually tuned. But instead of the organ oscillators, it has digital oscillators with different shapes and a unison mode, two analog filters in series, a spring reverb and a sequencer. It’s semi-modular and can also accept MIDI input as well as its touchplates. From the few demos I’ve seen on YouTube, it has a really nice wide range of sound character and should fit right in.

I’ve put some thought into other options, such as the Moog Subharmonicon or Dreadbox Nyx, but I think I’ve convinced myself that the QS is what I want to run with. I do want to sell some gear before moving it in, though.

lil’ oddball

The Microfreak showed up this afternoon and I’ve spent enough time with it to get solid first and second impressions.

While many of the oscillator modes are directly taken from the open-source code for Mutable Instruments Plaits, for the most part it doesn’t sound like “Plaits in a keyboard.” It’s its own weird little thing.

A lot of people think it’s kind of ugly, but some have said to reserve judgement until you see it in person. I think the only real sin is the graphic strip with the Arturia logo, which is a needlessly busy pattern right next to the already necessarily busy-looking textured touch keyboard. I’m just going to cover that with some gaffer tape.

The touch keyboard itself felt a bit awkward at first but I got used to it relatively quickly. It’s very sensitive, in a good and expressive way once you get to know it. There is a useful Hold button which works for drones as well as the arpeggiator, and the arp can be saved into the sequencer for further editing and manipulation. The “spice and dice” buttons can trigger either permanent or temporary random variations to the sequence which always seem to be musically compatible, if not necessarily brilliant.

There’s a modulation matrix that is cleverly implemented, and modulation amounts can be modulated by other modulators. As in, the touch pressure can increase the amount of vibrato, or an LFO can affect how much the cycling envelope affects the timbre. Circular and self-modulation patches are also possible and quite spiffy.

It’s paraphonic, in that there can be multiple oscillator voices playing simultaneously (with their digital envelope and VCAs) but there is a single analog filter, and a single cycling envelope. Many of the oscillator models are pretty complete without the filter anyway. One particular strength of this setup is that drones or arpeggios going through a single resonant filter can use it to glue the sound together and to emphasize particular harmonics. In fact, the moment I discovered this is when the last doubt about keeping this synth evaporated — I expect this to be one of the major ways in which this synth will fit into my music. I’ve been enjoying drones for hours today already 🙂

Like its predecessor, I have it running through the Elektron Analog Drive. It can add a nice bit of warmth and solidity, or dirty it up beautifully.

flying onto the shelves

Mutable Instruments Shelves arrived over the weekend to fill out the biggest remaining gap in the modular. It’s a shelving and peaking EQ, where each band has a full frequency range available and its frequency, gain and Q can be voltage controlled. The two peak EQ bands are implemented using filters, and have lowpass, bandpass and highpass outputs as well. So it’s quite a versatile piece of gear. Some of the things it’s good for:

  • Basic EQ tasks. However, the available boost and cut isn’t as strong as I might want for some purposes, and it would make more sense to just use CraveEQ or Toneboosters EQ4 for that.
  • Feedback loops. A little Q and positive gain, and it will resonate nicely inside of a feedback loop. It’s particularly great with delays, and with MSLR to maintain the tone while keeping it from getting out of hand.
  • CV-controlled EQ can be awfully fun. Taking a notch filter (or two) with a gentle slope and sweeping its frequency around with an LFO is a nice bit of (possibly) subtle modulation. With a bit more Q it’ll do filter sweep-like sounds without actually cutting anything completely; to me the character is more interesting than most filter sweeps tend to be.
  • Anything a normal filter can do. It sounds pretty respectable to just use one of the LP, BP, or HP outputs with it.
  • Parallel filter trickery. It does only have one input, but with two independent filters, each of which has three outputs, plus the main EQ output, I can turn a mono signal into mid-side stereo, process different bands with different effects and mix them afterward, or split bands to go into Planar.

I don’t see much reason to hang onto FXDf, or even Ripples now. Once I figure out what I will replace them with, I’ll pull them out. I have a few candidates in mind but none are really pulling on the strings strong enough yet to jump on it. Loquelic Iteritas or Telharmonic would certainly be interesting but do I need another oscillator? Intellijel Bifold seems like a quite good wavefolder, but between the folder on Shapeshifter, FM Aid’s folding abilities. the folder built into ENOSC and rather capable wavefolding available in both Disting and Bitwig, I’m not convinced I need it.

The other toy I’ve taken delivery of is… well, it is kind of a toy, but it’s also got some surprising instrument-like features, and the combination is a bit frustrating. The Artiphon Orba, which I Kickstarted back last December, is a little handheld musical gizmo. It has 8 pads that respond to velocity as well as movement on two axes (for pitch bend or dynamics), and an accelerometer to detect tilting, shaking, and slapping the side. It has four “modes” — drums, bass, chord and lead — and a record mode that records sequences on each channel, with overdubbing. It can also act as a MIDI controller (with MPE!) over USB or Bluetooth. It’s got a pretty decent built-in speaker with ports both on the bottom and sides so it’s not muffled no matter how you position it.

On the down side: onboard, it has only one sound per mode; you can select from about 7 different sounds using an app that updates the device over its connection. Many of the sounds are just not great, but it’s got a few relative winners. Sometimes there are unpleasant pops at the start of notes, and switching modes while trying to record parts can pause and silence the whole thing for a second or two. The recorder doesn’t seem to record tilt gestures. Octave switching is a pain, and often results in accidentally switching modes instead; you can forget doing it smoothly while playing to extend the usable range. The built-in speaker is plenty loud (a little too much with some of the lead sounds in their higher octave) but the output jack is so woefully low level that I boosted it with both Gozinta and MSCL. The PC app is as buggy as Florida in July, and the Android version is nonexistent (typical of every Kickstarter device I’ve seen that promised both Android and iOS support).

So, as an instrument in its own right, it’s mostly a fun toy — for a beginner, generally easier to jam on and get something that sounds moderately okay than a Casio keyboard or the like. But how about as a MIDI controller for more serious efforts? Let’s start with Bluetooth MIDI being unsupported in Windows, which they could have fixed with the app or a driver but didn’t, so it’s tethered with a USB cable. You can’t change the scales assigned to each mode — so the bass is always major, the lead is always pentatonic major, and the chords are fixed (depending on which sound you’ve loaded). As far as I can tell so far, it sends velocity, mod wheel on tilt, and it probably sends MPE on the two axes per pad. There is apparently a third-party semi-unofficial app for MIDI configuration, which I hope can unlock the thing’s potential a bit more. But it won’t surprise me if it remains mostly a toy, which I might sample occasionally with Phonogene or freeze in Clouds or Mimeophon, after a copious gain boost that is.

But hey, it lets me have the fun of doing some idle finger drumming, which I haven’t done since switching from Maschine to Bitwig, or playing some honestly kind of nifty growly drones with the “Ohm” bass sound.

The next bit of gear should arrive Friday: I picked up an Arturia Microfreak in a Labor Day sale. I disdained it when it was released, and wasn’t sure I liked the touch keyboard when I very briefly stabbed at it at Knobcon last year — but I have heard so many good things about it, and good things people have done with it, that I’m willing to give it a solid try. If it works out, it will replace my Reface CS, which has been seeing less and less action.

The CS is fun to jam with but it can be a little difficult to make it relevant to what I want to do. The wee Freak though is mostly a polyphonic version of Plaits, plus an SEM filter and a touchplate pressure keyboard — and I know from the 0-Ctrl that I do like a touchplate pressure keyboard when it works well — and a neat little arpeggiator/sequencer that can dynamically randomize its content in a way that tickles my generative sequencing funny bone, and also record parameter automation. So that could potentially be quite nifty. Given that used prices are hovering very close to new prices, it is practically zero risk if I don’t like the thing after all.

Something else I have been thinking about is the Elektron Model:Cycles. I ignored it when it was announced, because it’s an Elektron groovebox, kind of a glorified drum machine, and I don’t do the sort of music that uses drum machines. Except… it’s really an inexpensive little FM synth, and there’s apparently a small community that swears it was secretly made for ambient music. Its FM is definitely more on the pure and simple side, being modern and clean and 2-operator based — but you layer that a bit and it really can sound gorgeous. I’ll have room for it in my setup if the Freak takes over from the CS, but it would eat into my available inputs. Also I can’t help but think it won’t really do anything I can’t do with Bitwig’s FM-4 instrument, aside from maybe inspire me to use it more often. (The Microfreak itself also has an FM mode…) So I’m gonna have to think on that more. I might be better off with something that has some crust on it, like an old FB-01 (cheap, but requires software to edit it without going crazy) or a MegaFM (less cheap, would probably have to take over from the Lyra-8) or a Volca FM (cheap, can also control and sequence itself, and can play 6-op DX7 sounds yet has a few hands-on controls to tweak it). Or just not make any of those moves at all.

I had a thought about what I could do for the next album, but it’s a little out of my comfort zone: record patch-from-scratch sessions and see if I can make it an actual performance. It will probably require a lot of trying and there will probably be many times where I’ve got a boring and ugly first minute, then some really sweet sounding awesomeness, and then it kind of falls apart or gets ruined with a sudden blast of noise as I try to add another voice. I really admire the people who can do this as a show, but I’m sure there is some preparation and practice involved. Still, it could be fun and rewarding to try it…