ABLETON 12 SOUND DESIGN ATLAS

Deep Dubstep Track — Roots & Meditational Deconstruction

Deconstruction of an archetypal 140 BPM roots and meditational deep dubstep tune for a sound system: a 32-bar DJ intro, a sine sub that plays the tune, horns, melodica and a vocal sample as the spine, skanks and dub throws for space, and a drop that is pressure rather than noise. Every element as a Live 12 recipe, a headroom map, a 10-step rebuild and a live dub version.

Difficulty
Intermediate
Live edition
Suite
Max for Live
In some recipes

The blueprint: 140 BPM, 176 bars (≈ 5:02), E phrygian, half-step. This is the deep counterpart of the dubstep track deconstruction. That page’s energy comes from a wobble riff, a build and a growl layer. This one has no riser, no wobble and no impact. Its weight comes from three decisions. (1) The sine sub is the lead instrument. It plays a phrased 2-bar tune on E0 (41.2 Hz), and everything above it is there to frame it. (2) The track is written for a DJ and a rig. It opens with 32 bars with no sub so the outgoing tune can be blended on its bass, and it closes with 24 bars that mirror the intro. The mix leaves 6 dB of headroom and avoids brickwall limiting, because a big system plays dynamics and punishes density. (3) The drop is subtraction. Bar 47 empties to a siren and an echo tail. Bar 48 is a kick and a sustained E0 arriving together, and for 8 bars nothing else enters. Labels used on this page: Documented means the choice follows a named, linked source on how this music is made or played. Recipe means it is this page’s own construction. The track itself is invented. It transcribes no record.

BPM
140, felt at 70 (bar = 1714 ms, 1/2 = 857 ms, 1/4 = 429 ms, 1/4 dotted = 643 ms, 1/8 = 214 ms, 1/8 dotted = 321 ms, 1/4 triplet = 286 ms, 1/16 = 107 ms, 1/16 triplet = 71 ms). Documented: Mala describes working at about 140 and changing perceived speed through space and time rather than tempo (RBMA lecture).
Key
E phrygian (E F G A B C D). Sub root E0 = 41.2 Hz, answer note B0 = 61.7 Hz, b2 F0 = 43.7 Hz, one drop note D0 = 36.7 Hz. The kick is tuned to B0. The reasons for the choice are in sound system tuning.
Length
176 bars: 16 intro A (atmos and vocal), 16 intro B (drums, no sub), 16 pre-drop, 32 drop 1, 32 “space”, 32 drop 2, 32 outro. The phrases run 8 bars inside 16- and 32-bar sections.
Mood
Devotional, heavy and patient. The dread comes from the b2 and the gaps. The peak is drop 2 bars 128–136, where a steppers kick runs under the horns.
Stereo image
Sub, kick, wub, rim and snare body are mono. The skank is about 120% wide. Horns are a mono pair panned ±15. The returns run 130–150% wide and the atmos 160%. Utility Bass Mono sits on the master at 150 Hz as a safety net. Nothing below 150 Hz should change when you fold the mix to mono.
Delivery
A club master (about −10 LUFS integrated, limiter catching 1–2 dB) and a “dub” for your own sets with no limiter and peaks at −1 dBFS. The dubplate background is in Mix decisions.

Arrangement

The structure comes from sound-system dub and the dubplate DJ, not from EDM. The intro is a DJ tool. Intro A (bars 1–16) is a field-recording bed, vinyl crackle, a spoken or chanted vocal phrase and a rim on beat 3 from bar 9. Intro B (17–32) adds the half-step kick, filtered shakers and a high-passed skank, and still has no sub. These 32 bars are where the previous tune’s bassline plays out. The pre-drop (33–48) brings the horn motif, vocal throws and, from bar 41, two short sub “hints” per 4 bars at −6 dB. A siren swells at 45–47. Drop 1 (49–80) is drums and sub alone for 8 bars. The wub enters at 57, skank and hand drums at 65, horn stabs at 73. “Space” (81–112) drops the kick. Melodica answers and dub throws carry 16 bars, then the sub returns solo over rims for 8 bars before a 1-bar silence. Drop 2 (113–144) changes the bassline (cells B and C, octave jumps) instead of stacking layers, and adds a 4/4 steppers kick for 8 bars at 129. The outro (145–176) mirrors the intro. The sub leaves at 152, so the next DJ gets 24 bars without low end. Documented: in Lauren Martin’s VICE oral history, Artwork recalls Skream and Benga writing simple loops with a 16-bar intro before the drop, because DJs played only about the first minute of each tune (VICE, “An Oral History of Dubstep”, 2015). The 32-bar intro here is this page’s recipe. It doubles that window so a deep tune can be blended on its bassline.

Energy curve. Energy here means sub pressure plus density, and it moves in steps. Bars 1–32 sit at about 2/10: texture, rim and filtered drums, and the sound system’s scoops do nothing. The horn motif at 33 sets the melodic spine at about 3/10. The sub hints at 41 are short notes at −6 dB that tell the room the bass exists without delivering it. Bar 47 falls to near zero. Bar 48 jumps to 7/10 with only three elements playing. Each 8-bar block of drop 1 adds one layer: wub, then skank and perc, then horn stabs. The space section falls to 3/10 for 16 bars, rises to 5/10 when the sub plays solo at 97, and goes back to zero at 111. Drop 2 starts at 7/10 and peaks at 9/10 in bars 129–136, with the steppers kick, the octave-jump cell C and the horns together. It steps back to 8/10 until 144 and falls away through the outro. The loudest moment of the track carries about the same RMS as drop 1. What changes is the bass writing and the kick density. Documented: stripping back over building up is Mala’s stated preference, and he described “something quite beautiful in taking things away” (Zweikommasieben interview).

Transitions. No sweeps, no white noise and no snare roll. Into drop 1: a Drift siren swells over bars 45–47 and is cut at bar 47 beat 3. The last rim of bar 47 goes to the Echo return with feedback at 85%. Beats 3–4 of bar 47 hold only that echo tail. Bar 48 beat 1 is kick plus sub. Into the space section: the sub stops a bar early (bar 79), a tom answer on the last two beats mirrors the sub cell’s last notes, and the kick does not return. Into drop 2: the solo sub over rims (bars 97–110), then a full bar of silence at 111 broken only by a vocal word thrown to the Dark Hall return. Within drops, every 8th bar gets a snare throw to the Dark Hall, and every 16th bar removes the sub. Documented: the deep dubstep reference pages here call this “tension through space” (sub-only drops). The recipe below shows how it is spread across a whole track.

Frequency layering

On a sound system the 36–65 Hz band is the scoops’ job, and it holds only the sine. The kick is tuned to B0 (61.7 Hz) and decays in 160 ms. It sits slightly above the root, and its short body hands the bar to the sub. Documented: choosing a kick that hits higher than the sub is standard dubstep advice (Sound On Sound, Dubstep Basics, 2010). The parallel harmonic chain (80–200 Hz) exists for smaller rigs and phones. On a big rig you barely notice it, and on a laptop it carries the tune. The wub is high-passed at 120 Hz and low-passed near 1.4 kHz, so it stays warm and never competes with the spine. The spine is time-shared rather than EQ-carved. The horns play in the pre-drop and in the last 8 bars of each drop, the melodica only in the space section, and the vocal only where neither plays. That is how a three-lead track stays clear with no ducking between leads.

Element by element

Kick — short, tuned to B0, half-step, with ghosts and an 8-bar steppers run

Instrument: Drum Rack, Drum Sampler, Operator · Level: essential · Genres: deep dubstep, halfstep, steppers

Quick start

Load
Drum Rack, Drum Sampler, Operator, Tuner
Set
  • Pad C1: Drum Sampler dry, short kick tuned to B0 with the Tuner; Decay 160 ms, Punch 25%
  • EQ Eight HP 45 Hz 48 dB/oct, −3 dB dip at 300 Hz
  • Body: Operator Osc A sine at B0, pitch envelope +12 st over 25 ms, amp decay 140 ms; 5 dB under the knock
  • Ghost C#1 same sample, Decay 90 ms, LP 1.5 kHz, −8 dB
  • Velocities downbeat 120, pickup 96, ghosts 35; no groove on the kick clip
You should hear
A short, dry knock above the sub’s root that frames it instead of competing, gone within 160 ms of each hit.
Live edition
Suite

The kick frames the sub. It does not compete with it. It plays one hit per bar in half-step, a pickup in bar 2, ghosts at velocity 35, and a 4/4 steppers pattern for 8 bars of drop 2. The sound is a dry sampled knock tuned to B0, with an Operator sine body that is gone within 160 ms. The single-sample knock and the kick/snare programming are already taught in spacious half-time drums. This rack only adds the steppers chain and the tuning to this key.

  1. Drum Rack pad C1: Drum Sampler with a dry, short kick. Tune its fundamental to B0 (61.7 Hz) with Live’s Tuner on the pad output. Decay 160 ms, Punch 25%. EQ Eight HP 45 Hz 48 dB/oct and a −3 dB dip at 300 Hz.
  2. A second chain on the same note: Operator, Osc A sine at B0, pitch envelope +12 st over 25 ms, amp decay 140 ms, Sustain −inf. Mix it 5 dB under the knock. It adds roundness, not length.
  3. Ghost on C#1 (the same sample, Decay 90 ms, LP 1.5 kHz, −8 dB) and steppers on D1 (a softer kick, Decay 120 ms). Programming the steppers run on its own note lets you mute it per scene in the live version.
  4. Velocities: downbeat 120, pickup 96, ghosts 35, steppers hits 2–4 at 90. Do not groove the kick clip. Documented: the Attack Magazine breakdown of a deep 140 beat runs the whole kit at 50% swing (straight) and puts the shuffle in the hats and percussion (Attack, “Mystik Dubstep”, 2012).
B0 is 20 Hz above the E0 root. If you tune the kick onto E0 instead, its body and the sub’s first cycle sum on every downbeat. You then hear a 3–6 dB bump on the meter, and on a stacked rig a hollow spot when the phases disagree. Above the root, and short, is the sidechain-free way to keep them apart.

Rim and snare — rim for seven bars, the full snare and a Dark Hall throw on the eighth

Instrument: Drum Sampler, Simpler · Level: intermediate · Genres: deep dubstep, roots, dub-influenced

Quick start

Load
Drum Rack, Drum Sampler, Simpler, EQ Eight
Set
  • Pad D1 rim and snap 1–127, snare body 110–127; bars 1–7 at velocity 100, bar 8 at 120
  • Rim: Drum Sampler rimshot, Decay 140 ms, tuned to B2
  • Snap: Simpler finger snap, HP 800 Hz, −6 dB
  • Snare body: Drum Sampler dark snare, Decay 230 ms, LP 6 kHz
  • Send B Dark Hall, −inf except +3 dB for a single 1/16 on each bar-8 hit
You should hear
A dry rim on beat 3 for seven bars, then a darker full snare with a hall tail on the eighth, gone by the next downbeat.
Live edition
Standard and up

Beat 3 always gets a hit, but the sound changes with the phrase. Bars 1–7 of each 8 use a dry rimshot plus a high-passed finger snap. Bar 8 adds a dark snare body (180–260 Hz) and sends that one hit to the Dark Hall return at +3 dB. The return is gated by the kick (Flip on), so the tail is gone by the next downbeat. The reverb design and the gate are in reverb-drenched snare. Here the point is the 7+1 rhythm of the sound across the track.

  1. Pad D1: three chains with velocity zones. Rim 1–127, snap 1–127, snare body 110–127 with a fade of about 5. Program bars 1–7 at velocity 100 and bar 8 at 120.
  2. Tune the rim to B2 (the 5th) with Drum Sampler’s Transpose and the Tuner. The Dark Hall tail then rings inside the harmony.
  3. Send B automation: −inf throughout, +3 dB for a single 1/16 on each bar-8 hit. Draw it once, then copy the 8-bar clip envelope.
  4. Toms: three Drum Sampler pads tuned to E2, B1 and G1 (82.4, 61.7 and 49.0 Hz) with Decay 380 ms. Use them only as the 2-hit answers shown. G1 is near the sub, so keep its velocity at 70.
Documented: the Attack breakdown layers a weighty electronic snare (energy at 160–200 Hz) with a high-passed, reverbed clap, and feeds one acoustic snare per two-bar loop into a delay (Attack). The 7+1 throw here is the same idea stretched to an 8-bar phrase.

Shaker, hats and hand drums — the only swung parts, with a pause on beat 3

Instrument: Drum Rack, Groove Pool · Level: intermediate · Genres: deep dubstep, roots, meditational

Quick start

Load
Drum Rack, Drum Sampler, EQ Eight, Glue Compressor
Set
  • Drum Rack shaker, closed hat, woodblock and hand drum on Drum Sampler, own clip
  • Groove Pool MPC 16 Swing-62, Timing 100%, Random 6%, Velocity 15%; commit before arranging
  • EQ Eight shaker HP 4 kHz, hat HP 6 kHz, woodblock HP 400 Hz, hand drum HP 100 Hz
  • Pans shaker 20 R, hat 10 L, woodblock 30 L, hand drum 15 R; tops pause on beat 3
  • Glue Compressor 2:1, attack 10 ms, 1 dB GR
You should hear
The only swung parts of the track, wide from position alone, stepping aside on beat 3 so the rim has the air to itself.
Live edition
Standard and up

The swing lives in the tops: a shaker, a thin closed hat, a woodblock and a hand drum (cajón or bongo), all on a Groove Pool swing. Two rules come straight from documented practice. The tops step out of the way on beat 3, so the rim has the air to itself. The hand drums are low-cut at 100 Hz so they never touch the kick or the sub. The general triplet and swing method is on the triplet hats technique.

  1. Put the tops on their own Drum Rack and clip. Groove Pool: MPC 16 Swing-62, Timing 100%, Random 6%, Velocity 15%. Commit the groove before arranging, so later edits keep the feel.
  2. Shaker HP 4 kHz, closed hat HP 6 kHz, woodblock HP 400 Hz, hand drum HP 100 Hz. Documented: the cajón samples in the Attack breakdown are low-cut at about 100 Hz, and its hats and shaker pause on beat 3 (Attack).
  3. Pans: shaker 20 R, hat 10 L, woodblock 30 L, hand drum 15 R. Width comes from position, not wideners.
  4. Intro B: run the whole tops bus through Auto Filter HP at 1.2 kHz (MS2 circuit), opening to 20 Hz over bars 41–47. This is the only filter sweep in the track.

Sub — an Operator sine as the lead: cells A, B and C, glide 70 ms, no ducking

Instrument: Operator · Level: essential · Genres: deep dubstep, roots, meditational

Quick start

Load
Operator, EQ Eight, Utility
Set
  • Operator Osc A sine only, Voices 1, Glide 70 ms, amp A 10 ms R 80 ms
  • Velocity → Volume 0
  • EQ Eight HP 24 Hz 48 dB, LP 90 Hz 48 dB
  • Utility Mono; track outside every group, no sidechain
  • Cells A A A B in drop 1; delete the sub in the last bar of every 16, no fade
You should hear
A pure mono sine that is the tune of the track, every note at the same weight, with clean cuts as releases of pressure.
Live edition
Suite

The tune of the track. It is a mono sine on its own track, outside every group, with no sidechain. The patch, the note range and the writing method are covered in pure sine sub as the lead and writing the sine sub, so this section shows only what this track does with them. There are three 2-bar cells: A (statement), B (response) and C (octave answer, drop 2 only). They are arranged as A A A B in drop 1 and A B A C in drop 2, with a full bar of silence every 16.

  1. Build the patch from the genre page (Operator sine, Voices 1, Glide 70 ms). Set Velocity → Volume to 0, so every note has the same weight however the clip was played.
  2. Write cells A, B and C as separate 2-bar clips in Session view, then record them into the arrangement in the cell order shown in the drop 1 zoom. Keep the A A A B grammar. The listener learns A, and B is the event.
  3. Silence: delete the sub in the last bar of every 16-bar block (bars 63, 79, 127 and 143 in Live’s 1-based bar count), and in bar 47 and bar 111. Do not fade. A cut is a release of pressure. A fade just sounds like the level going down.
  4. Sub hints (bars 41–46): play only the B0 answer of cell A, at −6 dB clip gain. That is two notes per 4 bars, enough to show the room what key it is in.
  5. Level: the sub track runs at −15 dBFS RMS (−12 dBFS peak, since a sine’s crest factor is 3 dB). Every other element is mixed against it. Check each note on Spectrum with block size 8192. If one note measures 3 dB or more above its neighbours, fix it with clip gain, not EQ (see tuning).
Documented: Mala has described building basslines from a plain sine in Reason’s Subtractor, then filtering and low-cutting, because sub energy you cannot hear still moves the cones and makes a line sound “flabby” (RBMA lecture). The 24 Hz 48 dB high-pass and the absence of any saturation on the main sub track are this page’s Live version of that control. Documented: in the VICE oral history, Mala describes stripping tunes down to “just straight bones and a three-note melody”. He also credits Jason Goz with teaching him that some frequencies do not translate to vinyl, and that cutting bottom end can make a track sound heavier (VICE oral history). Cells A and B use three or four pitches each for that reason.
Do not sidechain the sub to the kick in this style. The kick sits on B0 and decays in 160 ms, and the sub attack is 10 ms. Ducking would tear a 70–100 ms hole in the lead instrument on every downbeat, which on a big rig is felt as the tune stuttering. Sound On Sound’s general dubstep primer does suggest sub sidechaining. Deep sub-led writing replaces it with tuning and arrangement, as the genre page argues.

Wub — Operator FM an octave up, LFO at 1/2, a second LFO reshaping it over 4 bars

Instrument: Operator, Auto Filter · Level: intermediate · Genres: deep dubstep, DMZ-era, roots

Quick start

Load
Pitch, Operator, Auto Filter, Saturator, EQ Eight
Set
  • MIDI from the Sub track, Post FX; Pitch +12 st
  • Operator B→A, B ratio 1 at −14 dB, LFO 1/2 sine → B + filter, retriggered per note
  • Auto Filter MS2 LP 12, 420 Hz, Res 25%
  • Saturator Analog Clip 3 dB
  • EQ Eight HP 120 Hz 48 dB, LP 1.4 kHz
You should hear
The sub’s cells an octave up, breathing twice per bar, so the sub becomes visible on small speakers.
Live edition
Suite

The wub is the sub made visible. It plays the same cells one octave up (E1, 82 Hz), high-passed at 120 Hz, breathing on a synced 1/2 LFO. This track adds one thing to the Operator wub recipe. A second, slower movement reshapes the wub across a 4-bar pattern, so two passes of cell A never sound identical. Documented: Coki describes his basses as an LFO-moved sound with a midrange layer moved by a second LFO, and says he changes the midrange over a four-bar pattern (The Arts Desk Q&A). The Live 12 build below is this page’s own.

  1. MIDI: route the sub’s cell clips to the wub track (MIDI From: Sub track, Post FX), with a Pitch device at +12 st. One edit to the sub writing then updates both layers, and they can never drift apart.
  2. Operator: build the wub from the genre page. Retrigger the LFO per note on, so each cell starts its breath on beat 1. The LFO at 1/2 gives two swells per bar. On cell A’s long E0 anchor (11 steps, about 1.2 s), that is just over one and a half breaths.
  3. Add the LFO (Max for Live) device after the rack. Map it to the Mouth macro, triangle shape, rate 4 bars, depth 35%. Over 4 bars the wub moves from closed and woody to open and vocal and back.
  4. Entry: the wub enters 8 bars after the sub (bar 57, then bar 121) and leaves with the sub for the 1-bar silences. Clip volume runs at −20 dBFS RMS, so it reads on a phone and stays well under the sub on a rig.
The slow-wobble divisions and Shaper curves on the bass page and its LFO shapes both plug straight into LFO 1 here if you want the wub to lean or inhale differently.

Skank — Electric offbeats on beats 2 and 4 of the half-step bar, ducked by its own echo

Instrument: Electric · Level: intermediate · Genres: roots, deep dubstep, dub-influenced

Quick start

Load
Electric, EQ Eight, Saturator, Echo, Hybrid Reverb
Set
  • Electric short tines, Damper on, release 40 ms
  • Notes Em (E3 G3 B3) then F (F3 A3 C4), 1/16 stabs on steps 5 and 13
  • EQ Eight HP 280 Hz 48 dB, LP 5 kHz; Saturator Medium Curve 2 dB
  • Echo 1/8 dotted, FB 35%, Ducking on, threshold −24 dB
  • Hybrid Reverb Convolution, spring IR, 1.4 s, 25%
You should hear
Short roots chord stabs in the half-step backbeat gaps, moving Em to F, staying out of the wub below.
Live edition
Suite

The roots signal: short Electric chord stabs on the half-step backbeat gaps (steps 5 and 13), moving i (Em) to bII (F) in phrygian. The full patch is on skanking offbeat chords. In this track it has three jobs. It is a filtered, high-passed DJ cue in intro B. It enters at bar 65 of drop 1, where it is the only harmony. It plays an 8th-note variant against the steppers kick in drop 2.

The skank and the horns share 300 Hz–3 kHz. The arrangement never plays the skank under the horn motif, only under the horn stabs, and the stabs land on beat 1 while the skank is on the offbeats.

Horns — an Analog brass pair for the motif and the stabs

Instrument: Analog · Level: intermediate · Genres: roots, deep dubstep, dub

Quick start

Load
Analog, Chorus-Ensemble, Saturator, EQ Eight
Set
  • Analog Osc 1 saw, Osc 2 saw +7 cents, both 0 dB; PEG +0.4 st, 60 ms
  • Fil 1 LP12 900 Hz, Resonance 10%; filter env A 70 ms, D 400 ms, S 30%, amount +45%
  • Amp A 25 ms, D 300 ms, S 70%, R 220 ms; Glide off
  • Chorus-Ensemble Chorus 15%; Saturator Analog Clip 2 dB
  • EQ Eight HP 220 Hz, bell −2 dB 800 Hz, LP 7 kHz
You should hear
A two-saw brass sound with a lip scoop and a blow behind the attack, a two-player section when doubled and panned.
Live edition
Standard and up

The horns carry the roots identity: a 2-bar phrygian motif in the pre-drop and before drop 2, and single stabs on beat 1 in the last 8 bars of each drop. Stock Live has no dedicated brass synth, so this is a subtractive brass sound from Analog. Two detuned saws, a filter envelope that opens a little behind the amp, and a small upward pitch scoop give the lip. Doubled a 4th or a 5th below and panned ±15, it sounds like a two-player section.

  1. Analog: Osc 1 saw, Osc 2 saw detuned +7 cents, both at 0 dB. Noise off. Use Osc 1’s pitch envelope (PEG) at +0.4 st with a 60 ms time to scoop up into each note.
  2. Fil 1 LP12 at 900 Hz, Resonance 10%. Filter envelope: A 70 ms (later than the amp, which is where the “blow” lives), D 400 ms, S 30%, amount +45%. Key tracking 50%.
  3. Amp: A 25 ms, D 300 ms, S 70%, R 220 ms. Glide off: horns tongue every note.
  4. Velocity → filter envelope amount +30%, so harder notes are brighter, like a harder breath. Play the stabs at 115 and the motif at 85–100.
  5. On the last note of each motif, open Send A (Echo) for that note only. The F3 repeats into the gap where the sub returns.
If you have Meld, an alternative horn uses its Swarm or Bubble-free saw engine with Meld’s own filter, but the Analog route above is the one this page’s levels were built with. For horn character from samples, see strings, brass and physical models.

Melodica — the space section’s answering voice, phrased against Echo

Instrument: Meld (or Analog) · Level: intermediate · Genres: roots, dub, meditational

Quick start

Load
Meld, Chorus-Ensemble, EQ Eight, Echo
Set
  • Meld reed-type engine + saw at −12 dB, LP 2.5 kHz, A 40 ms, Glide 40 ms Legato
  • Chorus-Ensemble Chorus 20%
  • EQ Eight HP 300 Hz, LP 6 kHz
  • Echo insert, L 1/4 dotted, R 1/4, FB 45%, Dry/Wet 28%, Ducking on
  • Playing octave 4, every other 2 bars; at most 1/16 Q at 50%
You should hear
A reedy answering voice that resolves the horns’ F, leaving every other bar to the echo’s repeats.
Live edition
Suite

In the 32-bar space section the kick is gone and the melodica answers the horn motif’s unresolved F. The patch is the reed sound on dub siren and melodica. Here it sits in octave 4, plays every other 2 bars, and leaves the bar in between to the Echo’s repeats. That alternation is the call and response of the section.

  1. Play it live from a keyboard or Push with the Set scale at E phrygian, and record 16 bars. Keep a take that leaves gaps. Do not quantise beyond 1/16 Q at 50%.
  2. Mod wheel → Meld filter cutoff (or pressure → a breath parameter) for swells on long notes.
  3. From bar 97 (sub solo) the melodica plays one last answer, then stops at 104. The sub has the room.

Vocal sample — one spoken or chanted phrase as a signpost, not a song

Instrument: Simpler, Auto Shift · Level: intermediate · Genres: roots, meditational, deep dubstep

Quick start

Load
Simpler, Auto Shift, EQ Eight, Compressor
Set
  • Simpler Classic, Warp off, transposed so the vowel sits on E or B, Filter HP 200 Hz
  • Auto Shift E phrygian, Correction 25%, Formant Shift 0%
  • EQ Eight HP 250 Hz, bell −3 dB 3 kHz, LP 8 kHz
  • Compressor 3:1, attack 5 ms, 3 dB GR
  • Sends Echo for single-word throws, Dark Hall for the last word
You should hear
One whole phrase as a signpost, then single words thrown into the returns, never chopped into a riff.
Live edition
Standard and up

A short phrase marks the intro, the pre-drop, the space section and the outro: a spoken line, a chant, or a toaster’s call from a session you recorded or a sample you have cleared. It is never chopped into a riff. It appears whole, then as single words thrown to the returns. Documented: in Red Bull Music Academy’s account of one DMZ landmark, the bassline and the vocal came first and the beat arrived months later (RBMA Daily, “Anti War Dub”). The lesson for this archetype: write the sub and the vocal together, before the drums.

  1. Load the phrase into Simpler (Classic, Warp off) so it can be triggered from MIDI. A one-shot keeps its timing intact. Transpose so its sustained vowel sits on E or B. Auto Shift at 25% correction only nudges it.
  2. Intro A (bars 5–12): the phrase whole, dry centre, Dark Hall at 15%.
  3. Pre-drop (bars 37–44): single words on beat 1 of every 2nd bar, each one thrown to the Echo for one beat.
  4. Bar 111 (the silence before drop 2): the last word alone into the Dark Hall at 100% send. This is the only sound in the bar.
The vocal and the horns share 250 Hz–4 kHz. They never play together in this arrangement, which avoids any need for ducking between them. If you must overlap them, duck the horns 3 dB from the vocal with a Compressor sidechain, release 250 ms.

Siren — one gesture per drop, cut on beat 3

Instrument: Drift · Level: beginner · Genres: roots, dub, deep dubstep

Quick start

Load
Drift, Echo, Utility
Set
  • Drift Osc1 tri, Pitch Mod from the LFO (free, 3 → 9 Hz, about ±5 st by ear) and Env2 (about +12 st over 3 bars)
  • Echo 1/4 dotted, FB 60 → 85%, HP 400 Hz
  • Utility Width 140%, −14 dB
  • Cut clip volume to zero at bar 47 beat 3, not faded
You should hear
A siren rising in rate and pitch, cut dead so the echo tail is the last thing before the drop.
Live edition
Suite

The siren appears twice: bars 45–47 and 109–111. It rises in LFO rate and pitch and is cut, not faded, at bar 47 beat 3, so the Echo tail is the last thing before the drop. The patch and the performance gesture are on dub siren and melodica. This page fixes only the timing and the cut.

Atmos and crackle — the room before the system plays

Instrument: Simpler, Granulator III, Spectral Resonator, Vinyl Distortion · Level: intermediate · Genres: meditational, deep dubstep

Quick start

Load
Simpler, Spectral Resonator, EQ Eight, Utility, Vinyl Distortion
Set
  • Simpler field recording, Warp Texture, loop crossfade 800 ms
  • Spectral Resonator MIDI mode from E0 + B0 + E1, Decay 2 s, Mix 20%
  • EQ Eight HP 400 Hz, LP 12 kHz; Utility Width 160%, −22 dB
  • Vinyl Distortion separate return, Crackle only, Tracing off, −30 dB
  • Automation 0 dB in bars 1–32, falling to −inf by bar 40
You should hear
A field recording ringing in E with crackle beneath, the room before the system plays, fading as the drums arrive.
Live edition
Suite

A field recording (rain on a tin roof, a temple courtyard, a quiet street) runs through Spectral Resonator so it rings in E, with Vinyl Distortion crackle under the intro only. The build is the atmos bed technique. Its placement here: it opens the track alone, ducks out by automation in the drops, and returns for the space section and the outro.

  1. Automate the atmos track: 0 dB in bars 1–32, falling to −inf by bar 40. It returns at bar 81, is gone again by 104, and comes back from 145.
  2. The crackle return is only open for bars 1–32 and 161–176. Its disappearing as the drums arrive is a cue the dance floor reads without knowing why.

Returns — Echo for throws, Dark Hall for size, Spectral Time once per section

Effect: Echo, Hybrid Reverb, Spectral Time · Level: intermediate · Genres: deep dubstep, dub

Quick start

Load
Echo, Roar, Hybrid Reverb, Gate, Spectral Time
Set
  • Return A: Echo L 1/4 dotted, R 1/4, FB 60%, HP 400 Hz LP 3 kHz, Wobble 20%, Noise 8%
  • Return A: Roar Single, Soft Sine 12%, post LP 4 kHz
  • Return B: Hybrid Reverb Dark Hall, Decay 4.5 s, Pre-delay 30 ms, 100% wet
  • Return B: Gate SC from Kick, Flip on, Hold 240 ms, Release 180 ms
  • Return C: Spectral Time Delay on, Time 1/8, Feedback 70%, Tilt −40%, into Dark Hall
You should hear
Throws, size and spectral smear that arrange the space section, none of them adding energy where the sub lives.
Live edition
Suite

Three returns do most of the arranging in the space section. The throw method is on dub delay throws, and the send rack is in the dub delay send rack. Everything on the returns is high-passed at 300–400 Hz, so no return ever adds energy where the sub lives.

The drop — pressure, not noise: what bar 48 contains and what it withholds

Arrangement · Level: essential · Genres: deep dubstep, roots, meditational

The drop has three elements: kick, rim and sub. There is no impact, no crash and no reverse cymbal. Its force is contrast. The 32 bars before it had no sustained sub. Bar 47 had no sub at all, and its last two beats held only an echo tail. The first note of the drop is the longest note in cell A (11 steps, 1.18 s) at the track’s reference level. On a big rig the body feels that as weight arriving. Documented: Mala names the space between the high frequencies and the sub as the source of jungle’s weight, and describes testing tunes on the system at FWD>>, where he could hear instantly what was wrong (RBMA lecture).

  1. Bar 47: mute everything at beat 1 except the rim on beat 1, the siren until beat 3, and the returns. Do it with track activators in an arrangement automation lane, not clip edits, so the live version can copy it as a scene.
  2. Bar 48 beat 1: kick, sub and nothing else. The rim comes on beat 3. The tops stay out until bar 57, together with the wub.
  3. Level check: drop 1 bar 48 and drop 2 bar 129 should read within 1 dB on a short-term LUFS meter. If drop 2 is much louder, you have added layers instead of writing.
The same sine at the same level feels heavier after silence. This is partly the ear (loudness depends on what came before) and partly the room: on a large rig, pressure builds in the space over the first several hundred milliseconds of a sustained low note. Short notes never get there, and long notes after a gap do. That is why the first drop note is long. Documented: Pinch, in the VICE oral history, recalls a Bristol night where the room felt like a pressure chamber, and Loefah describes the gaps between drum hits as an instrument in their own right (VICE oral history).

Mix decisions

Gain staging and headroom. Mix with nothing on the master. The sub sets the reference: −15 dBFS RMS, −12 dBFS peak. The kick peaks at −12 dBFS. Worst case, the kick peak coincides with a sub peak: 0.251 + 0.251 = 0.502, which is −6.0 dBFS. The rim and snare body peak at −13 dBFS on beat 3 over a sustaining sub (0.224 + 0.251 = 0.475, −6.5 dBFS). With the wub, horns and returns in, drop 2 peaks between −6 and −5 dBFS on the master before any processing. That 5–6 dB is the headroom. The master chain then adds about +4 dB of make-up into a limiter that catches 1–2 dB on the kick-plus-sub peaks only. Documented: on level, Jason Goz, who cut dubplates for the early scene at Transition, advises “Don’t watch the meter”. He describes cutting-desk headroom as room to be used, not a target to hit (Spitfire Audio feature by Emma Warren). Mala has said Transition’s lathe-cut acetates were where his tunes were mastered and cut, and that cutting forced him to be fully sure of a mix first (RBMA lecture). A dubplate is a lacquer acetate made for exclusive sound-system play (Wikipedia: Dubplate).

ElementMix level (pre-master)WidthNotes
Sub (E0 sine)−15 dBFS RMS / −12 peakMonoThe reference. Direct to master, no group
Sub harmonics−33 RMS intro / −29 dropsMono80–200 Hz only; mute it and the big-rig sound should barely change
Kick−12 peakMonoB0, 160 ms
Rim / snare−13 peak (bar 8 accent −11)MonoDark Hall send only on bar 8
Tops bus−28 to −24 peakPanned ±10–30Groove only here
Wub (E1)−20 RMSMonoHP 120 Hz, LP 1.4 kHz
Skank−22 peak120%HP 280 Hz
Horns (pair)−18 peak each±15 panNever with the vocal
Melodica−18 peakCentre, echo wideSpace section only
Vocal−16 peakCentreWords thrown to the returns
Returns A/B/C−20 to −14 peak130–150%All HP ≥ 300 Hz
Atmos / crackle−30 RMS / −40 RMS160%Out by bar 40
Master, pre-chain−6 to −5 peak (drop 2)Bass Mono 150 HzThe headroom you mix into

Sidechain and ducking map. Sound On Sound’s general primer suggests sidechaining the sub to the kick. This archetype replaces that with tuning, note lengths and arrangement, and keeps the ducking that remains above 300 Hz.

SourceTargetAmount / shapeWhy
KickSubNoneKick on B0 and short; sub attack 10 ms. Ducking would break the lead line
KickReturn B (Dark Hall) GateFlip, Hold 240 ms, Release 180 msThe hall closes on each downbeat, so kick and sub land in clean air
Self (input)Echo on skank, melodica, Return AEcho Ducking on, threshold −24 dBRepeats fill only the rests
Snare (bar 8)Wub2 dB, attack 1 ms, release 60 msLets the throw snare’s body speak through the 220 Hz region
VocalHorns (only if overlapped)3 dB, release 250 msArrangement normally keeps them apart
ArrangementAtmos, crackle, skankAutomation, not compressionThey leave the drops entirely

EQ carving.

Bus structure.

Documented: Utility’s Bass Mono switch folds everything below its Bass Mono Frequency (50–500 Hz) to mono, and Bass Mono Audition lets you hear only that band (Live 12 manual, audio effect reference). The 150 Hz setting is this page’s choice. It is a safety net: if it audibly changes anything, a track upstream has stereo low end that needs fixing at its source. Documented: two points from the VICE oral history. Loefah recalls panning his bass until Goz told him the tune would have to be cut mono. Goz says he held levels back when cutting and kept the music dynamic, because loud, squashed records tire the listener (VICE oral history). Those are the reasons for mono low end at source and for the 5–6 dB of headroom above. The wider headroom-first approach for big rigs is the subject of sound system mixing.

Build it yourself in 10 steps

  1. New set at 140 BPM, Set scale E phrygian. Tracks: Sub and Sub Harmonics (outside all groups), a Drums group, a Tops bus, a Music group, Atmos. Returns: A Echo, B Dark Hall (gated by the kick, Flip), C Spectral Time into B, D crackle.
  2. Write the sub first, with nothing else playing. Cells A and B as 2-bar clips (Operator sine, Glide 70 ms, Velocity → Vol 0, HP 24 / LP 90, Utility Mono). Set it to −15 dBFS RMS. Record the vocal phrase or choose a cleared one now, and line its stressed syllable up with cell A’s downbeat.
  3. Kick rack (B0 knock and Operator body, 160 ms, plus ghost and steppers chains) and the rim/snap/snare-body pad with velocity zones. Program the 2-bar skeleton: kick 1, rim 3, pickup on bar 2. Kick at −12 peak, rim at −13.
  4. Sub harmonics: route the sub MIDI to a copy of the Operator, then Saturator Soft Sine 12 dB, band 80–200 Hz, −18 dB under the sub. Toggle it on a phone speaker and on monitors. It should rescue the tune on the phone and be nearly inaudible on the monitors.
  5. Wub: MIDI From the sub track, Pitch +12, Operator FM with a 1/2 LFO, Auto Filter MS2, HP 120 / LP 1.4 kHz, and the 4-bar LFO on the Mouth macro. −20 dBFS RMS.
  6. Tops on their own clip with MPC 16 Swing-62, hats and shaker silent on beat 3, hand drum HP 100 Hz. Skank (Electric, Em → F on steps 5 and 13, Echo Ducking).
  7. Spine: the Analog horn pair (motif and stabs), the Meld melodica answer, the Drift siren. Write each so it leaves the others’ bars empty: horns in the pre-drop and drop ends, melodica in the space section, vocal where neither plays.
  8. Arrange the 176 bars from the timeline. Critical details: no sub before bar 41 or after bar 152. Sub silent in bars 47, 63, 79, 111, 127 and 143. Bar 47 cut on beat 3. Drop 1 holds three elements for 8 bars. Wub enters at 57 and 121. Steppers kick in 129–136 only. Bar 111 is one thrown word.
  9. Mix to the levels table with the master chain bypassed. The master should peak at −6 to −5 dBFS. Engage Bass Mono 150 Hz and confirm nothing changes. Mono-check the whole track: the skank and returns get narrower, and nothing below 300 Hz moves.
  10. Bounce two versions: the club master (Glue ≤1 dB, Limiter 1–2 dB, ceiling −1 dB true peak, about −10 LUFS integrated) and a dub with no limiter, peaking at −1 dBFS. Play the dub on the biggest system you can reach before you change anything else. The rig is the final test.

Variations

Meditational / halfstep. One sub note per 2 bars from E Bhairav (E F G# A B C D#), a kick only on bar 1 of every 2, no skank and no horns. The spine becomes a bowl or bell through Spectral Resonator and a chant phrase. Dark Hall goes up to 7 s and the tops thin out to a shaker and a hand drum. Drops get even quieter in density. The structure stays, with the space section stretched to 48 bars. See bass line 8 on the bass page for a matching sub.

Steppers roots. The 4/4 steppers chain runs through both drops, the skank plays 8th offbeats throughout, and the sub writing moves to shorter reggae-shaped notes (2–3 steps, release 60 ms) in G minor. Horns lead every 8-bar phrase, and the siren appears at every 16. The wub is optional. A rolling sub at this density leaves little room for one. This is closest to the dub sound systems’ own steppers repertoire. Pair it with dub reggae.

Dark minimal (dungeon-leaning). No skank and no horns. The spine is a single detuned Operator FM bell and a cold field recording. The wub gets a lopsided 1/4-dotted LFO and the vocal becomes a pitched-down word. The kick moves to 2-step positions in bar 2 (beat 1 and the a-of-2). Swap the rim for two short rims tuned about −3 st that alternate as call and response. Documented: Attack’s 2-step groove tutorial uses that rim pairing and drops rims and percussion at the end of each 8 bars before the rim returns on the last beat (Attack, Beat Dissected, 2026). It is written at 93 BPM, so take the idea, not the grid. It keeps the same 32-bar DJ intro and the same headroom rules. Compare the darker variant on the dubstep track page’s “deep / dungeon 140” note, which starts from a wobble patch rather than a sine lead.

Live performance version

A deep dubstep live set is closer to a dub mix than to a DJ set. Stems play as Session clips, and the performer works the sends, the cuts and the sub mutes like a dub engineer on a desk. Every scene below keeps the track’s 8-bar grid, so you can loop a scene and still land the next drop on a phrase. Documented: Sleeper describes writing in Ableton Live with Operator for most of his sound design, and recording arrangements live through an Akai APC with effects automated by hand (Kmag, “Sleeper in the Studio”, 2014). The scene, rack and mapping layout below is this page’s own.

Scenes

SceneBlueprint barsPlayingManual move
1 Room1–16Atmos, crackle, vocal phrase, rim from 9Vocal word throws (Throw A)
2 DJ zone17–32+ kick, filtered tops, filtered skank; no subTops HP macro, loop as long as the blend needs
3 Horns33–40Horn motif, drums, atmos fadingEcho throw on the motif’s last note
4 Hints41–47+ sub hints at −6 dB, siren 45–47Siren rate on an encoder; Cut button on bar 47 beat 3
5 Drop A48–56Kick, rim, sub cell A A A BHands off. Let it weigh
6 Drop A+57–80+ wub, tops, skank, horn stabsSub Mute button for the 16th bar; snare throws
7 Space81–96Rims, hand drum, melodica, returns openEcho feedback ride; Spectral Time once
8 Sub solo97–111Sub over rims, last melodicaCut on 111, thrown word into the Dark Hall
9 Drop B112–128Cells A B A C, D0 slide entryWub Mouth macro
10 Steppers129–144+ steppers kick, horn stabs and motifSteppers level macro; tom answer pads
11 Outro145–176Drums, atmos, vocal; sub out at 152Into the next tune

Performance rack

Push mapping

Hardware integration

On why the rig matters at all: Coki put it bluntly. Without the right sound system, “you can’t really hear dubstep” (The Arts Desk). Pinch describes the music he values as “spacious, otherworldly mood spaces” and looks for surprise in sound-system music (Clash interview). This archetype is built around both ideas: room, and the moment it fills. The DMZ nights that shaped this lineage ran in a dark room with a heavy system and the DJs on the dance floor’s level (FACT, 10 years of DMZ), under a flyer line inviting people to meditate on bass weight (RBMA Daily).
Further reading. Lauren Martin’s Aftershock: The Seismic Impact of Dubstep (Velocity Press, 3 August 2026) is an oral history built from interviews with 28 scene figures. It covers FWD>>, DMZ, Tempa, Hyperdub, Skull Disco and the scene’s return to sound-system roots. According to Beatportal’s announcement, it follows the music through record shops, pirate radio, cutting houses, forums and sound systems, and expands Martin’s 2015 VICE oral history of dubstep. That free piece is cited above for Mala, Loefah, Goz, Pinch and Artwork. This page draws on no content from the book. It is listed as the long-form history behind the DMZ nights, dubplates and sound-system weight that the archetype reflects. For more Attack Magazine technique in this area, see its sub bass lines video tutorial.

Related