Overmono made a dance album that doesn't constantly beg for the drop
Pure Devotion arrived August 7 — eleven tracks written in the Icelandic mountains with two synthesisers and an FX unit. What happens to club production when you stop optimising every arrangement for a viral clip?

Overmono released Pure Devotion on Friday, August 7, 2026 — eleven tracks, the Welsh brothers' second album after 2023's Good Lies.
The detail that explains the record: Ed and Tom Russell wrote much of it in isolation in the Icelandic mountains with, by their own account, little more than two synthesisers and an FX unit. XL's release page carries the duo's own account of the process: "we wanted to try make something that captured the physical space of our studios. setting up loads of ways of processing, recording, re amping sounds… 'the vocal isn't sitting quite right? let's track it through an old tannoy speaker in the basement. that synth needs to twang harder? let's play it through an old train station announcement speaker from the 1930's.'" They also describe running tape over magnets, blowing up speakers, leaning on a cheap overclocked FX unit, and oven-baking a cymbal covered in coffee grounds, crisps and vinegar — a trick they credit to Sylvia Massy.
UK garage, two-step and rave remain the core reference points. But the kicks hit harder, the tempos run quicker, and the moodier sections lean into downtempo and trip-hop. It is tighter and more focused than the sprawl of the debut.
The reason this is worth a producer's attention is not the gear list. It is what a small, fixed setup does to arrangement.
Constraint as an arrangement tool
Two synths and one effects unit removes the single biggest time-sink in modern electronic production: choosing.
When your palette is fixed, you cannot solve a boring section by adding a new sound. You have to solve it by changing what the existing sounds are doing — filtering, re-voicing, muting, re-sequencing, or processing the same source three different ways. That produces records where sections feel related rather than stacked, because they are literally made of the same material.
It also produces a different relationship with effects. One FX unit used on everything becomes a glue layer: a shared reverb tail or saturation character across every element, which is the analogue equivalent of committed bus processing.
What "not begging for the drop" actually means
"Drop-optimised" arrangement has a recognisable anatomy:
| Section | Function in a drop-first record | Function in a texture-first record |
|---|---|---|
| Intro | Minimum viable groove to beatmatch | Establishes tone and space |
| Build | Filter sweep, snare roll, rising noise | Often absent entirely |
| Drop | Maximum energy, all elements, sub arrives | Energy arrives gradually or sideways |
| Breakdown | Removal of drums to reset tension | A change of texture, not a removal |
| Outro | DJ tool tail | Resolution — the track ends because it's finished |
A record that relaxes that structure is not "less energetic." It redistributes energy. Instead of a single peak the arrangement is engineered around, you get several smaller shifts — a filter opening two bars early, a vocal chop dropping an octave, a sub arriving under a section that was already playing.
The practical trade-off: these tracks are harder to clip. There is no eight-second moment that works out of context on a phone. That is a deliberate cost.
Techniques worth stealing
- Reuse one source three ways. Take a single synth line; use it clean as the lead, pitched down and saturated as a bass, and heavily reverbed and reversed as a pad. The track gains layers without gaining new sounds.
- Let the sub arrive late and quietly. Instead of the sub landing at the drop, fade it in under an existing section. The listener feels the change before they identify it.
- Record through something imperfect. Playing a stem out through an old PA, a cheap speaker or a guitar amp and re-recording it adds frequency loss and distortion that no plugin preset quite reproduces — and it commits you to the result.
- Cut the build, keep the tension. Try removing the snare roll and the noise riser entirely. If the section still lands, the arrangement was doing the work. If it doesn't, the riser was hiding a weak transition.
- End tracks instead of tailing them. A written ending changes how the whole arrangement reads.
The broader shift
Club music spent several years being shaped by short-form video, where the incentive is a single identifiable moment. There is now a visible counter-move: records built for the length of a room rather than the length of a clip.
Neither is more legitimate. But if every arrangement decision is being made in service of a fifteen-second excerpt, the other ninety percent of the track becomes packaging. Pure Devotion is a good argument for the ninety percent.