Concatenatin’ Keys

Thanks to software-based synthesis, composing and recording microtonal music is easier today than it has ever been. Ever. Like, in the history of the human race. Playing microtonal music remains a great deal more difficult.

The tricky bit is, you need a keyboard to send MIDI notes to the computer. I mean, unless you’re planning to type notes into a Csound score, which is even more laborious. The moment you stray outside of the standard 12-note-per-octave keyboard, your options become almost insanely restricted. A few tiny companies build alternative keyboards, but they’re both expensive and, in most cases, designed in odd ways, because they’re being built not by people who hope to make money but by people who have a Grand Vision of the Future of Music. And their grand vision is not, in general, widely shared.

Up to now, I’ve been struggling along with a standard keyboard (an M-Audio Axiom 61, not that that matters). I have pretty much taught my brain to find chord and scale shapes on this keyboard, in tunings with 17 or 19 or 22 or 31 equal-tempered notes per octave. It’s a brain-twister, but not impossible.

This week I hit the wall. I noticed that a piece I’m working on in 31ET had too many melodic phrases that meandered up and down in single scale steps. I more or less improvise these melodies while the sequencer plays, and I can glue my fingers to about six scale steps (three with each hand) and play a line that’s at least modestly interesting. Figuring out which key to play in order to do a larger melodic leap is not very practical in real time, especially when you consider that the pattern with which the scale lays out on the keyboard will change depending on what key you’re in.

A related problem is figuring out what notes are in the chords I recorded yesterday. I spend a lot of time hunting and pecking, trying to figure out which note is the right one for the bass part I’m editing.

Starr Labs makes a swell keyboard, the U-648, that has a uniform grid of hexagonal keys, which means the chord and scale shapes are the same no matter what key you’re in. But the U-648 costs $4,000, and spending that kind of money would make me a little nervous.

C-Thru Music makes a smaller controller called the Axis-64. Actually, it comes in two models. The Axis-49 is quite stripped-down, with only 98 hexagonal keys and no LCD at all. (Also no pitch or mod wheel and no rear-panel inputs for pedals.) I have one sitting at my elbow right now. In order to assign the Axis-49’s keys to the MIDI note numbers that I need, I’ve written a little patch in Pd. The Axis sends its MIDI notes (#1 through #98) to the computer. Pd receives them, remaps them to create a useful keyboard layout, and transmits the revised note on to MIDI Yoke, which hands them off to FL Studio. It’s a bit convoluted, but it works.

The velocity response of the Axis-49 is more than a bit funky, and the pattern of colors on the key tops has nothing to do with the scales I’m using, but it’s a quasi-viable input device. Still, 98 keys aren’t enough. With my preferred layout, I have exactly two octaves of a 31-note scale. I’d like three.

The Axis-64 has a lot more keys. It hard-wires them to MIDI note numbers, however, in order to provide what C-Thru hopes will be a sensible isometric note layout. Trouble is, their layout only makes sense if you’re playing music with 12 notes per octave. I have long felt this was a fatal flaw.

But now I’ve learned that the Axis-64 has a three-channel split keyboard mode, in which three groups of 63 keys each transmit on three separate MIDI channels. Within each group, there is no duplication of notes: Each key plays a different note. In split mode, then, the channel/note number data transmitted by each key is unique to that key.

This is good news, because it means I could write a Pd patch that would allow me to play the full Axis-64 keyboard using my own preferred note layout. Unfortunately, I’ve also learned that the Axis-64 is selling (direct from the manufacturer) for $2,500. Ouch! For that princely sum, I would expect a lot better-feeling keyboard.

The key layout I use, in case you’re curious — well, we need a new term in order for me to explain it. I call the smallest possible step in an equal-tempered microtonal scale a chroma-step. In 12ET, a chroma-step is the same as a half-step. But in 31ET, a diatonic or melodic half-step is usually three chroma-steps (sometimes only two). It would make no sense to call the smallest interval in 31ET a half-step, because then a diatonic whole-step would be five half-steps wide.

Various players might have their own preferences, but I find that a good keyboard layout uses an interval of three ascending chroma-steps as you move up a vertical column of keys (“up” being away from the player) and five ascending chroma-steps on the hex diagonal as you ascend to the right.

Finding chord voicings is still a bit hunt-and-peck, especially as I’m not always using the same scale. But once I’ve sound a shape for a voicing, I can move it around chromatically with the greatest of ease. The same shape always produces the same chord voicing.

Also, because the keys on the Axis-49 are somewhat smaller (only a bit larger in diameter than your fingertips), reaching large intervals with one hand is easy. With a 17-note scale, I can stretch three octaves. Try that on a standard keyboard!

I’m finding that it’s quite easy to discover chord voicings with this system that I would never be likely to stumble onto while using a standard keyboard. On a standard keyboard, my brain is too busy thinking, “Okay, a minor seventh plus a minor sixth equals a perfect fifth…”, or whatever.

To answer your next question, no, I’m not planning to post any videos on YouTube showing this keyboard in action. There are already way too many dumb videos on YouTube in which people attempt to demonstrate microtonal music or alternative keyboards without doing anything that’s even remotely musical. My plan is to really learn the two-dimensional chord shapes on this keyboard and then let you hear the finished music.

The notion that a chord voicing is not a linear arrangement of intervals but rather a two-dimensional shape — it’s kind of a brain-buster. Fortunately, I love brain-busters!