How Daniel Avery builds blown-out techno textures on Ultra Truth

Put on "Lone Swordsman" or "Chaos Energy" and the first thing that registers is that the synths sound like they're being eaten by their own reverb. The pads have a strain to them, the drums don't sit on a grid so much as drift through fog, and a steady noise floor binds the whole thing together. None of that is an accident. Avery's Ultra Truth workflow is built on a small handful of moves he's been refining since Drone Logic, and most of them are reproducible in the box if you understand what each one is doing to the signal.

Tape saturation on every channel, pushed harder than you'd expect

The grainy, slightly worn quality across Ultra Truth's synth bus and drums isn't a single hardware tape pass. It's a plugin run aggressively across many tracks. Tape saturation does three things at once: it gently compresses transients, adds even-order harmonics that read as warmth, and rolls a little air off the top so the sound feels less digital at the edges. Pushed lightly, you barely hear it. Pushed harder, it gives that pre-distortion sense of material strain - the sound feeling like it's about to break without actually breaking.

Avery's specific tool for this is MeldaProduction's M Saturation, which he told MusicTech "gives a pretty satisfying replication of tape saturation, something I push pretty hard on everything." The useful detail in that quote is "everything." It's running on individual channels, accumulating across the mix - not saved for a master bus afterthought.

Practically: insert M Saturation (or Saturn 2 in Tape mode) on drum subs, synth busses, and any sample that sounds too clean. Drive between 30 and 70 percent, mix between 30 and 50 percent so transients keep their shape. On sub-bass, skip it or use a high-pass on the saturator's sidechain - harmonic addition down there muddies the low end fast.

The reverb is three reverbs

The cavernous depth on the record comes from layers interacting, not one big plate or one big hall. Avery's reverb chain, as he described it to Future Music, runs the Eventide Space hardware pedal alongside Valhalla VintageVerb, with the two "throw[ing] up interesting results" together. The technique generalises beyond those two specific units.

Set up three sends. Send one is a short, dense room or plate at 0.4 to 0.8 seconds, almost no pre-delay, used to glue transients. Send two is your medium hall or chamber, 1.5 to 3 seconds, pre-delay around 20 to 40ms to let the dry signal speak first. Send three is the long wash - 4 seconds to effectively infinite, heavily filtered, sitting low in the mix. VintageVerb's Chaotic Hall and Dirty Hall modes add tape-style modulation inside the decay, which is why the tails on Ultra Truth don't fall in a clean line. They wobble.

High-pass every return. The longest send wants a filter around 400 to 600Hz; shorter sends can sit lower at 200 to 300Hz. Without that filter, low-frequency energy from three reverbs at once stacks into a low-mid wash that swallows the kick.

Synths through pedals, not just plugins

Avery's signature texture work started when he was sharing studio space with S.C.U.M. and putting drum machines through their guitar pedals, an approach he described to MusicRadar as "techno with a shoegaze aesthetic." That instinct is all over Ultra Truth: hardware synths feeding into hardware effects, with the impedance mismatch and circuit colouration doing work that plugins approximate but don't fully replicate.

The Roland JX-3P is central. Avery told Reverb the JX has a "floatier, spacier sound" than a Juno-60, which makes it less useful for cutting leads and more useful as a bed - exactly the kind of source that benefits from being drowned in reverb. His phrase for the combination of JX-3P and VintageVerb is direct: "That alone has been such an enormous part of my sound for several years now."

Working in the box, the closest equivalent is to chain a slightly dull-sounding analogue emulation (Diva on a darker patch, Repro-1 with the filter pulled back) into a multiband distortion before the reverb sends. Saturn 2 with three bands works well: lows at 5 to 15 percent drive, mids around 30 to 50 percent in a tape or tube style, highs at 10 to 25 percent through a transformer model. Parallel the wet at 40 to 70 percent so you keep the original note's definition under the grain. That gets you closer to "blown speaker" territory than a single-band distortion will.

Daniel Avery - "Chaos Energy" (2022), lead single from Ultra Truth

Found sound as percussion

A lot of the mid-frequency percussion on Ultra Truth doesn't come from a drum machine. Avery told XLR8R he's been "taking a contact microphone out with me around London and running the results through layers of distortion pedals." Those recordings end up in his percussion layers, sitting alongside or instead of 808/909 hits.

The reason this works: a contact mic pickup of a metal pipe being struck has timing and tonal variation that sequenced samples can't fake. Each hit is slightly different. Pitch-shift a few of them into a kit, distort the midrange to give them body that contact mics naturally lack, and you get percussion that feels like it exists in a physical space rather than a session.

A gentle low-shelf boost (2 to 4 dB around 100 to 200Hz) before any distortion stops contact-mic samples from sounding thin once they're driven. Then multiband saturation on the midrange adds the grain without further sharpening the top.

Keep the noise

The single most counter-intuitive move on the record is that Avery doesn't gate the hardware. He amplifies the hiss. He credits this directly to working with Alessandro Cortini, telling DJ Mag that Cortini "leaves a lot of the noise and hiss and quite often exaggerates that on his own music." On Ultra Truth, that translated to a deliberate philosophy: "Instead of hiding things that might feel like mistakes, I push those to the front, pushing the hiss, pushing the distortion, pushing the atmosphere that an old bit of equipment can throw up."

Photo by Dmitry Demidov on Pexels

There's even a story about a broken cassette machine that wouldn't record properly but produced "a static atmosphere which sits over an entire track like a beautiful fog." He kept it.

In a DAW, this means committing. Either gate nothing and let every channel breathe its own noise, or print a single low-level noise bed across the whole record so the texture is consistent. Inconsistent hiss - some tracks gated, some open - is the version that sounds like a mistake. Continuous noise reads as room tone, and room tone is what tells the ear the music is in a place rather than in a grid.

Print one of your synths with the noise floor up. Run it through M Saturation, a multiband distortion, and three reverb sends with different tails. Leave the hiss in the breakdowns. That's the recipe, scaled down to whatever room you're working in.