Coatlicue — a cosmic feedback-drone synth

Coatlicue is the proposed name for a theoretical hyper-giant star whose violent winds compressed the gas surrounding it and gave rise to the Sun and our solar system. Then, after burning bright, its short life ended in a catastropic supernova. (The name is taken from the Aztec name for the mother of the Sun. It has some accents marks I do not know how to reproduce.)

Coatlicue the patch is an noise/feedback/drone synth modeled after amp feedback. A limiter keeps it from runaway feedback, so it plays “nice” (or as nicely as feedback can) — producing a range of cacophonous timbres without blowing out your ears/speakers.

There are two voices. Each begins with a comb filter (formed from a short delay line), fed by a small amount of noise. The output of this comb filter is split into two feedback paths. One path contains a resonant low-pass filter, whose frequency, in conjunction with the length of the comb filter, determines the frequency of the voice (more or less; other elements, including the resonance of the low-pass filter, influence this). The other feedback path passes through a fuzz module, then into another, long delay line. Both feedback paths are then routed into a compressor, acting as a limiter to control the amplitude of the feedback, before being fed into the input of the comb filter.

But although there are two voices, the most interesting results occur when the voices cross-modulate, with their outputs feeding into one another’s comb filters. On their own, the voices are fairly simple, producing quasi-sine waves, but when they feed into one another, the timbres become much more complex, as do the effects of variations on their parameters.

The output of these voices goes into a low-pass filter, acting as a tone control, and then a reverb lite. The output is — due to CPU constraints — dual mono; the same signal is present at both outputs. (The voice is, I believe, a good foundation for subsequent stereo effects, if you want a more expansive sound.)

Along the left- and rightmost columns are buttons used to play the voices. These control envelopes of varying length, from SHORTEST, to LONGEST, with the length of the envelopes becoming longer as you ascend the column. At the bottom of each column is an ON button. The voices can be set to be closed — in which case, the envelopes will open the voices — or open — which will cause the voices to drone, with the envelopes in this case serving to silence or gate the voice. The ON button behavior is mirrored in the corresponding stompswitches, left and right. When the voices are closed, opening them with the envelope will cause the UI buttons in that column to increase in brightness in relation to the envelope. When the voices are open, the same behavior will occur with the UI buttons dimming in accordance with the envelopes.

There are also overall LEVEL controls for each voice. The cross-mod is taken post-VCA, so it will be affected by changes to these levels.

Each voice has a FREQUENCY control, which can go into low/sub-audio ranges using negative values. These low/sub-audio ranges produces interesting gating patterns as the voices cross-modulate, although they are limited — below a certain frequency, the voice will simply go silent. Also, I cannot really figure out why, but the voices seem prone to becoming silent between ~80 Hz and 180 Hz. (My theory is some sort of phase cancellation occurs between the comb filters and the low-pass filters in this range, but I don’t exactly know why. Honestly, this patch is a bit of a mystery to me in how all of the controls interact, which makes it a lot of fun to play around with, but a real pain to try to explain.)

Keep in mind, however, that these frequencies will be affected by the most essential control for the patch, which is the CROSS-MOD control. This control determines how much of the output of each voice is fed into the other and can really transform the timbre as well as producing a frequency output which is the product of both voices.

The low-pass filter’s RESONANCE affects both timbre, and to some degree, frequency. As the resonance gets lower, the voice becomes more unstable; at very low resonances, the voice may disappear. (More mysteries! I don’t know why it occurs; I have only observations to report from my travels to this doomed star.)

The feedback path involving the low-pass filter has the most immediate impact on the overall output of the patch, but the other feedback path can also contribute more subtly to its timbre.

The other feedback path passes through a fuzz (which also serves as the output) and a delay. (Then both feedback paths are fed into a limiter.)

Changing the DELAY TIME can be subtle, but at shorter delays it affects the timbre more (as well as the frequency), and with longer delays, it functions more as a modulation.

Most controls are individual to voices, but the FUZZ control affects the input gain of both voices’ fuzzes equally. Changing the fuzz amount also… changes the timbre. I’m writing this a lot, but the truth is all of these controls are very interactive in producing different “flavors.”

In many of those interactions, the product can be rather understated… until they’re not. The voices can cut each other out at certain sentences, or produce strange alchemy in their cross-modulation. I can’t explain every interaction; I don’t understand every interaction. That said, some of the great sounds of the patch emerge as you nudge them out of a condition where they’re gating one another… odd sputters and crackles. I love these moments at the margins between silence and cacophony.

There is a low-pass filter, with a FILTER FREQUENCY control, before the reverb, which acts as a sort of primitive tone control, shaping the high harmonics. When fully open, there are a lot of “spitty” sounds produced. Closed too much, and it gets quite muddy. I like to keep it between 1.4k and 3k.

There are REVERB DECAY controls and REVERB MIX controls. The reverb adds a great deal to the character of the patch, but I would add that the patch also responds well to being used with an external reverb, especially if you want to expand the stereo field.

The patch accepts CV/expression via the Cport. Each voice has a bipolar CV AMOUNT control, so you can modulate in two directions from one source. You could also use an expression pedal to sweep the frequencies. I wanted to add an internal modulation bus, but there wasn’t enough CPU for it.

I plan to create another version of this that omits the reverb, to provide more CPU headroom, with the trade-off being an expansion in the CV capabilities (and possibly an internal modulation bus; we’ll see) for Euroburo users and efforts to make the patch truly stereo.

Leave a Reply

  • Platform:
  • Category: Sound Synthesizer
  • Revision: 0.1
  • License: Creative Commons Attribution Share Alike 4.0
  • Modified: 5 months ago
  • Views: 1116
    Likes: 15
    Downloads: 268