The Trench — a drone synth for textured, rhythmic atmospheres

If the layout of the Trench looks similar to the Canyon patch (https://patchstorage.com/the-canyon-a-drone-synth-for-dense-distorted-atmospheres/), it is because I initially developed them at the same time. Where the Canyon utilized offset oscillators to create rich harmonic textures, the Trench initially used pinged filters to create bell-like tones — and I hated it. It just didn’t do anything for me. But then, the other day, I randomly loaded it and started messing with it — what if it droned by using looped envelopes, what if the sound source could mix between a sawtooth oscillator and noise, what if we gave that sawtooth a relatively wide frequency range…. and I started to like it. Like, really like it. It’s not going to replace a lead synth, but hot dang does it do textures pretty great.

The basic idea of a pinged filter still represents the core of the Trench, but where a pinged filter is achieved by sending a brief burst of sound (a ping as it were) to a resonant filter, the envelope used for the Trench’s sound source can be set to much longer envelope times, and that allows it to produce sounds much more diverse than the usual pinged filter.

Each color-coded row of the front panel is a keyboard module — tuned to E minor pentatonic over four octaves) — controlling an individual voice of the synth. The source for the resonant filter is a mix of noise and sawtooth oscillator (or one or the other; you control the mix). The sawtooth’s frequency is free-running, relative to the filters and can be tuned over six octaves, from -A3 to +A3 — with sub-audio frequencies, the oscillator produces trills, their distance dependent on the frequency. In the audio range, the frequency of the oscillator will interact differently, depending on how it is processed by the filter (and ring modulator).

Each synth can be set to drone. But the drone is achieved by a looped envelope, which gives the droning voices a rhythmic element. Although all the envelopes share an attack control, each one’s decay is individual, so as they drone, dense, rhythmic textures form. Noise modulating the filter can add even more character to the sound.

The voices are joined by a ring modulator, tied to the frequency of the resonant filter — with sub-audio frequencies, this can introduce an additional layer of rhythmic play; at audio frequencies, it can contribute to the timbre of the sounds, harmoniously or in dissonance. Then everything is washed with modulated delay and plate reverb.

The patch is stereo and passes stereo signals at its input.

===== The Trench M patch notes =====

‘M’ means MIDI and replaces the keyboard module control via the front page with MIDI control. Additionally, pitchbend has been routed to the ring modulator frequency, and a control to set the range of the pitchbend is available via the front page. (I thought this was the more beneficial routing, but you can re-route pitchbend as you see fit.)

For the most part, this adaptation preserves the original design. The combination of free-running and drone voices made possible through individual keyboard module control is not as simple when operating over MIDI. The most ‘accurate’ design would have been to employ each voice on an individual MIDI channel and treat the patch as what it sort of really is: multi-timbral instances of similar voices rather than a polyphonic synth sharing the same means of articulation.

In this case, I chose to preserve the individual voice drone switches, but that does mean as you play through, the pitch of the droned voices will change.

The MIDI channel is set to 1 by default, but this can be changed. The MIDI note in module and the pitchbend in module are found on the same page.

Additionally, the non-M, old-school (by a few days) version of the Trench included in this download makes one revision: the level of the synth set via the control page will not affect the level of the audio passed through the patch (which will maintain unity gain).

A brief video on the MIDI version can be found here: https://youtu.be/5UjQ6CMjJXE

===== Back to original patch notes =====

Controls:

The four keyboard modules correspond to different voices. By default, they are tuned to an E minor pentatonic scale, over four octaves (from E2 to E6), but you can retune them as you see fit.

At the end of each keyboard module are a “Decay” control and a “Drone” pushbutton. The “Decay” control determines the decay of the exciter’s envelope, from 1.33ms to 60s (exponential). The “Drone” button causes that voice’s envelope to loop.

Along the bottom are global controls for the patch.

The voices all share an “Attack” control for their envelopes, once again 1.33ms to 60s (exponential).

The “Source Balance” controls a mix for the exciter, between a sawtooth oscillator (0) and a noise module (1), with points in between being mixes of both. The “Source osc freq” controls the frequency of the sawtooth oscillator from -A3 (-1) to +A3 (1). In sub-audio frequencies, the oscillator produces rhythmic trills as the wave re-excites the filter. In audio frequencies, the character of the sawtooth changes as it interacts more harmonically (and dissonantly) with the filters’ frequencies.

“Ring mod offset” determines the frequency of the ring modulator, relative to the filter frequency for a given voice. Sub-audio frequencies introduce tremolos with a slight phasing quality, while audio frequencies introduce the harmonically rich sidebands of classic ring modulation.

“Decay-Resonance” controls the resonance of the filters. Since resonance can ring out, it also affects how sounds decay.

“Noise mod” applies noise (Gaussian modulation achieved by multiplying random modules) to the filter frequency. This can produce everything from a faded tape sound to more aggressive distortion.

“Ambience” is a macro that controls the mix of both reverb and delay, as well as the delay time and feedback, the decay of the reverb, and the amount of modulation sent to the delay.

Because the various elements of the patch produce different gain amounts, the “Level” control is linked to the level of the output module, offering +/- 20 dB gain.