catchlight
By wangus on January 25, 2026 11:59 pm
these ~16 chords took me many days to compose.
no idea what the volume on the track is like. i hastily normalized by eye and uploaded.
m8 6 tracks sequencing the chords, run through Mood mkii delay -> Lossy -> Analog Heat mkii (Saturation circuit) -> Bluesky -> TP-7 . live twiddling knobs plus M8 touchscreen.
bounced to M8, trimmed to even number of bars, sequenced, patched up some rough spots (m8 touchscreen was being kinda glitchy). layered a touch of bass and percussion. final pass through Analog Heat Enhancement circuit, a bit of live twiddling.
my fav chord is the Amin11 voiced as A-E-B-C-G-D . i don't think i've done that before.
also fun fact you can play this on piano. all the chord voicings are two hands' range.
——————
some more thoughts
I was imagining a piano sound for the main synth. At first I tracked it out in hypersynth, but it’s just not the right tone. Too rich timbre for the crunchy harmonies.
The new wavesynth wavetable options are like a cheat code, just fiddle around with them till you hear something interesting. Modulating the scan parameter is really finicky; the wave tables are pretty dense (large amplitude modulations just sound goofy). There’s one envelope moving scan slightly, two fluffing a filter cutoff (one fast, one slower pulse), one key tracking panning.
I wanted some randomness, so there’s an ERR command (is that what it’s called?), but more interestingly some velocity/RET/RNL nonsense in the table. I wanted like a random delay to triggering the note, but I don’t think there’s a clean way to do that. So I futzed till I got this sound, which kind of gets what I want, but also has some weird clicks. But that’s just extra character.
I need to toy with Mood more. I was a little disappointed that I couldn’t get the tape half doing something coherent realtime with the chords. Maybe it just doesn’t work that way, maybe I just don’t know it well enough. Or maybe the chord progression is so complex on its own that it’s hard for something to independently add variation sensibly.
The chords started from an unrelated (unshared) OP-1f sketch that has the same rhythm in the bassline (just octaves though). I had shower-thunk the concept of the chords building on that groove into that sketch, but then it evolved into this separate piece.
I haven’t written something “purely compositional” like this (idk off hand how else to describe it; like not built on sound design or beats or grooves or riffs) in a long time. While the harmonies might be more modern, I’m approaching with a classical western tonal music mindset. like Bach n shit.
Like the start is functionally like straight out of well-tempered clavier. Tonic (Dm), “””””two, five”””””, tonic, tonic with flat 7th harmonizing as dominant to the 4 (G). But the deceptive cadence to the 4, the 2 (E) is also an important center. And that’s essentially it, wandering through those keys.
The progression is four 4-step cycles. First, starts on Dm and moves to A11no3, dominant resolution to the second cycle that repeats Dm but starts to suggest Gm. It ends also with an A chord, but this one is a m13, losing the dominant pull back to D, allowing the plagal motion to E starting the third cycle. The third cycle is the most straightforward of all, resolving via Dmaj/F#dim to a Gmaj to start the fourth cycle. Then the fourth cycle, we have a hasty turnaround back toward Dm to restart. It’s like the first repeat of the first section of classic sonata form; we’ve established that Gmaj is the secondary tonal center of the piece, which would get developed further if this composition continued. But here we just get the repeat. The repeat is the same harmony, just dwindling voicing till the end where it’s reduced to just the D tonic.
But melodic features also strongly drove the composition. Only half-intentionally, the first half does a misleadingly basic sequential motion in the top voice (it just outlines Am, Gmaj triads, but it’s hard to hear it that way). And the bass note follows a motion that’s similarly simple if you pull it out of context. That really makes the chords more interesting to me, and probably is also a big reason I get them to work despite the big dissonance in some.
there’s also some places where you can make more sense of the harmonies as a sort of smearing/anticipation. Like in the first cycle, the last chord is an A chord with dominant function to D tonic, but it doesn’t have the characteristic C# leading tone, because that’s just a thing that works with dominant 11 chords. But the chord before _does_ have the C#, though it’s more a predominant function. So I kinda think of those two chords as having together a blended predominant/dominant function. Similar perspectives to be had on some other motions.
one last unrelated thought: the ZDF filters are rad. Can really add ringing fuzz to any bass waveform with sharp edges. I modulate the wavesynth SIZE parameter together with the filter cutoff to sort of animate the overtone.
thanks for coming to my improvised TED talk
ps another random note: the only drum sample is a 707 ride obliterated lol
Audio works licensed by author under:
CC Attribution Noncommercial No Derivative Works (BY-NC-ND)
those transitionary sounds hit me right in the face only to bring me right back down with the soft chord hits :')