ABLETON 12 SOUND DESIGN ATLAS

Famous Dub Records — How They Were Made

Thirteen roots dub records, from Keith Hudson’s Pick a Dub (1974) to Mad Professor’s No Protection (1995), chosen because something public says how they were made. For each: what the sources say, what is only heard, and how to rebuild the method in Live 12.

Difficulty
Intermediate
Live edition
Suite
Max for Live
Not needed

Dub was made by engineers, on equipment that was usually second-hand, modified or built at home: a discarded broadcast-era desk with a stepped high-pass filter at Tubby’s, a quarter-inch four-track at the Black Ark, transformers wound by the same apprentices who later took over the desk. This page analyses thirteen records from 1974 to 1995, one per engineer or lineage, from Kingston to South London. They were chosen for their documentation, not their fame. That documentation is uneven: sleeve credits and Discogs entries are solid, a few interviews (Tape Op, Sound On Sound, Red Bull Music Academy, United Reggae) are rich, and for several records nobody ever wrote down which unit did what. Where gear is missing, the page teaches the process facts that are public: mix live in one pass, strip a rhythm to drum and bass, version an album you already made, bounce and print effects to free tracks, feed effects back into themselves. Every claim is labelled Documented (linked inline to the interview, sleeve credit, label note or feature it comes from) or By ear (listening analysis or scene knowledge, which is not fact). The Live 12 recipes rebuild a method, never a record. No melodies are transcribed. The ≤2-bar rolls are generic shapes that show an idea.

Treat every tempo on this page as By ear. None of these sleeves gives a BPM. Roots dub mostly sits roughly between 60 and 85 BPM if you count the snare or rim on beat 3 (one drop), and many engineers and DJs count it double. Before you build, drop the record into Live and check auto-warp’s result against your foot. At 75 BPM a bar is 3200 ms, a 1/4 is 800 ms, a dotted 1/8 is 600 ms and a 1/8 is 400 ms. At 70 BPM a bar is 3429 ms and a dotted 1/8 is 643 ms. See BPM to ms.
This page assumes the site’s core dub chapters, and links to them rather than repeating them: the one-drop and steppers kits, dub bass, skank and spring in Dub Reggae; tape echo, spring IRs, sirens, throws, runaway feedback and the four-channel desk performance in Dub FX & Delays; and the macro-mapped return in the Dub Delay Send Rack. On-U Sound and Adrian Sherwood are on a separate page, and dub techno is on Famous Dub Techno Records.

Overview: the thirteen records

RecordYearEngineer / artistWhereSignature technique
Pick a Dub1974Keith HudsonKingston (mixed at Harry J’s)A themed dub LP built from recycled rhythms and new overdubs
Dub from the Roots1974–75King TubbyTubby’s, WaterhouseThe stepped high-pass filter played as an instrument
Dub Wise (Morris On Dub)1975Sylvan MorrisHarry J’sDubbing a finished instrumental album, session for session
King Tubbys Meets Rockers Uptown1976Augustus Pablo, King Tubby, Errol ThompsonRandy’s, then Tubby’sMelodica as the voice, a mix made in another studio
Super Ape1976Lee “Scratch” PerryBlack ArkBouncing and printing effects on a four-track
African Dub All-Mighty Chapter 31977–78Errol Thompson (Joe Gibbs)Joe Gibbs, Retirement CrescentSound effects as arrangement on a 16-track rockers rhythm
Strictly Dub Wize1978Dennis Bovell (as Blackbeard)Gooseberry, LondonA UK band’s own tracks, mixed the engineer’s way
LKJ in Dub1980Linton Kwesi Johnson, mixed by Dennis BovellLondonDubbing a poetry record: the band answers a missing voice
Scientist Rids the World of the Evil Curse of the Vampires1981ScientistChannel One, then Tubby’sOne-pass live mix: high-pass, not phaser
Prince Jammy Destroys the Invaders1982Prince JammyChannel One, then Tubby’sClean separation, effects washing over drum and bass
The Commandments of Dub1982Jah Shaka (remix Mad Professor)Ariwa, LondonMixing for one sound system: frequency over power
The Dub Factor1983Sly & Robbie / Black Uhuru (remix Paul “Groucho” Smykle)Channel One, Compass Point, Fallout ShelterSyndrums and synths in a dub of a major-label album
No Protection1995Massive Attack v Mad ProfessorAriwa, LondonFeedback loops on aux sends over 48 tracks

The records

Keith Hudson — “Pick a Dub” (1974): the dub album as a themed record

1974 · Mamba (MAM 003, UK) · reissued by Blood and Fire · tempo By ear (slow one drop) · key undocumented · Tools: mixed at Harry J’s Recording Studio (Discogs credit); desk and effects undocumented

Quick start

Load
Glue Compressor, Limiter, Echo
Set
  • Version duplicate a finished set, delete the vocal and lead, rename it
  • Overdub one new part over the old rhythm, such as a melodica or a horn line
  • Mutes mute more than you keep; drum and bass carry whole sections alone
  • Master Glue Compressor 2:1 with 1–2 dB of gain reduction, then Limiter ceiling −1 dB
  • Returns the same Echo and spring returns on every track
You should hear
Austere drum-and-bass versions of old rhythms, each with one new overdub, sounding like one room and one hand.
Live edition
Suite

Why it matters. Documented: The Wire called it one of the first dub albums. Lloyd Bradley’s This Is Reggae Music ranks it with King Tubbys Meets Rockers Uptown, Super Ape and African Dub Chapter Three, and Jon Savage voted it the greatest dub album ever (all cited in Wikipedia: Pick a Dub). The Keith Hudson article describes it as widely considered the first deliberately thematic dub album, and as the first dub album to get a UK release.

Documented tools and method: the album reworks earlier rhythms. “Declaration of Rights” becomes “Black Right”, “Satta Massagana” becomes “Satia”, and the title track is a dub of the rhythm behind Big Youth’s “S.90 Skank”, with a new melodica part by Augustus Pablo. Aston and Carlton Barrett play bass and drums, and there are vocal fragments from Hudson, Horace Andy and Big Youth (Wikipedia). The original UK pressing credits mixing at Harry J’s Recording Studio and names the Soul Syndicate as backing band (Discogs: MAM 003). Discogs also lists Sylvan Morris, then Harry J’s engineer, as mixer on a later edition (Discogs: Sylvan Morris). That credit is not confirmed here for the 1974 pressing.

Signature moves → Live 12

  1. Build an album from rhythms you already own (Documented method). Make a folder of your finished tracks’ stems. For each one, duplicate the set, delete the vocal and lead, and rename the result. The new title is part of the work: Hudson’s titles turn a song into a new idea.
  2. One new overdub per version (Documented: Pablo’s melodica). Record or program a single new part over the old rhythm, such as a melodica, a horn line or a percussion layer, and keep everything else as it was. The overdub is what makes the version a new record rather than an instrumental.
  3. Make it austere (By ear). Mute more than you keep. The Barretts’ drum and bass carry whole sections alone. Keep the master clean and low: Glue Compressor at 2:1 with 1–2 dB of gain reduction, then Limiter with its ceiling at −1 dB, and no more than that.
  4. Sequence for a theme (Documented aim). In Arrangement, put the versions in one long set with consistent returns. Use the same Echo return and spring return on every track, so the album sounds like one room and one hand.
Recycling a rhythm only works with your own material or material you have cleared. Hudson owned his productions. Re-voicing somebody else’s rhythm without permission is a different thing.

What to steal: the dub album as a concept, not a B-side dump. Rename, re-voice and sequence old rhythms until they say something new.

King Tubby — “Dub from the Roots” (1974–75): the stepped high-pass as an instrument

1974 per Wikipedia (the Jamaican pressing on Discogs is dated 1975) · Total Sounds (Bunny Lee) · no catalogue number on the Discogs entry checked · tempo By ear · key undocumented · Tools: MCI desk with an Altec high-pass filter (Documented for the studio; per-track use By ear)

Quick start

Load
Audio Effect Rack, EQ Eight, Echo
Set
  • Audio Effect Rack ten chains, each one EQ Eight with only band 1 on as a 48 dB/oct low cut at a fixed frequency
  • Chain Select zones of about 13 values with no fades, Chain Selector on Macro 1
  • Steps for example 30 Hz, 70 Hz, 120 Hz, 200 Hz, 350 Hz, 600 Hz
  • Move jump to a high step for one bar, back to the bottom step on the downbeat
  • Echo Repitch on, Wobble and Noise low; only a snare or one vocal word sent to it
You should hear
A filter that jumps like a thrown switch, turning the kit into hats and snare crack for a bar before the full mix slams back.
Live edition
Suite

Why it matters. It is one of the Bunny Lee / Aggrovators dub sets that defined the Waterhouse sound. Documented: Bunny Lee produced it, the Aggrovators backed it (with Robbie Shakespeare, Carlton “Santa” Davis, Earl “Chinna” Smith and Augustus Pablo among the players), Tubby mixed it at his studio and Philip Smart is credited as editor (Discogs: Dub from the Roots). Wikipedia lists it with Total Sounds, 1974.

Documented tools (for the studio, not this record): in 1971 Tubby acquired an MCI desk from Dynamic Sounds and installed it at his home (Ableton blog: Studio as an instrument, part 3). Its high-pass filter, an Altec 9069B, was the big red knob at the top right of the console. Its discrete steps, phasing and clicks made it a hidden weapon in his remixes (Sound On Sound, on a modern replica). The Quietus describes ten notched frequency steps, a homemade tape delay, and reverb, delay and phasing as his other hallmarks. Wikipedia describes the filter thinning a horn down to a narrow squeal. Not every Tubby-credited mix was his: Jammy and Scientist also mixed there (Quietus).

Signature moves → Live 12

  1. A stepped high-pass, not a sweep (Documented hardware behaviour). Make an Audio Effect Rack with ten chains. Each chain holds one EQ Eight with only band 1 on, as a 48 dB/oct low cut at a fixed frequency. Open the Chain list, show the Chain Select Editor, and give each chain a zone of about 13 values with no fades. Map the Chain Selector to Macro 1. Turning the macro now jumps between frequencies, with a small click at each step, which is the point. The frequencies in the diagram below are this page’s choice: the sources here do not list the Altec’s values.
  2. The filter on the whole rhythm, at the desk (By ear). Put the rack on a group containing everything except the bass, or on individual channels. Use the step move on a drum bar: jump to a high step for one bar so the kit becomes all hats and snare crack, then drop back to the bottom step on the downbeat.
  3. Strip to drum and bass (Documented hallmark). The core arrangement move is muting everything except drums and bass, then letting the other parts return one at a time. Use the four-channel desk method on Dub FX & Delays for the mutes.
  4. Homemade tape delay (Documented that it was homemade; settings By ear). Use Echo with Repitch on, the Wobble and Noise controls in the Character tab low, and HP and LP filters inside the loop. The full recipe is here. Send only a snare or one vocal word to it.
A 48 dB/oct cut that jumps instead of sweeping sounds like a switch being thrown, not a filter being opened. The ear hears a new mix, not a moving filter. That is why the move reads as composition and not as an effect.

What to steal: quantise your filter. A control with ten fixed positions forces bold decisions and makes every change audible.

Sylvan Morris — “Dub Wise (Morris On Dub)” (1975): versioning a finished album

1975 · Jaywax (HJ 113, Jamaica), produced by Harry J · tempo By ear · key undocumented · Tools: Harry J’s Recording Studio (Documented); desk and effects undocumented

Quick start

Load
Utility, Drum Buss, Echo, Hybrid Reverb
Set
  • Dub set the finished instrumental saved under a new name, same tracks and returns, new mutes and sends
  • Drum group dry and narrow: Utility Width at 60–80%, Drum Buss Drive around 10%, Crunch low, Boom off
  • Echo send, 1/4 dotted, FB 45%
  • Hybrid Reverb send, spring IR, 1.5 s
  • Lead organ or melodica in for two bars at a time, always through the echo, never dry
You should hear
A tight, dry, close rhythm section with the lead instrument appearing only as a ghost through the echo.
Live edition
Suite

Why it matters. Morris is the engineer whom some artists and experts credit with creating the Studio One sound, though he was never credited there. He left Studio One around 1974 for Harry J’s. Documented in the Discogs artist profile, which cites Michael Veal’s Dub (Discogs: Sylvan Morris). This is the album that carries his own name.

Documented tools and method: the Jaywax LP credits Morris as engineer, Harry J as producer and arranger, and a band of George Fullwood (bass), Carlton “Santa” Davis (drums), Earl “Chinna” Smith and Tony Chin (guitars), Joe White (melodica) and Leslie Butler (piano, organ), recorded at Harry J’s (Discogs: HJ 113). The same musicians, producer, engineer and studio, and mostly the same track titles (“Expression In Dub”, “Seven Heaven Rock”, “Naked Spectrum”), appear on Leslie Butler’s instrumental album Ja-Gan, issued by Trojan the same year (Discogs: Ja-Gan, TRLS 112). The page infers that the dub album is a mix of the same sessions. That is a reading of the credits, not a stated fact. Documentation on how Morris mixed it is thin.

Signature moves → Live 12

  1. One session, two albums (Documented by the credits). Finish the instrumental mix first. Then save the set under a new name and mix it again as dub: same tracks, same returns, new mutes and sends. In Live, use Collect All and Save so the dub set never depends on the original.
  2. Keep the lead instrument as a ghost (By ear). On a keys-led instrumental, the organ or melodica is the “vocal”. Let it in for two bars at a time, always through the echo return, never dry.
  3. The tight, dry rhythm section (By ear). The Soul Syndicate players give a compact, close-miked feel. Keep the drum group dry and narrow: Utility Width at 60–80%, and Drum Buss with Drive around 10%, Crunch low and Boom off. The space comes only from the sends.
Do not take a dub of a finished mix by rendering stems with effects printed. A dub needs dry stems and live sends, or the echoes from the instrumental version will fight the new ones.

What to steal: treat every instrumental session as two records. The dub is a second performance of the same tape.

Augustus Pablo — “King Tubbys Meets Rockers Uptown” (1976): melodica as the voice

1976 · Yard Music (Jamaica) / Clocktower (LPCT 0085, US) · title track first issued as a single in 1974–75 · tempo By ear · minor feel, root undocumented · Tools: recorded at Randy’s, mixed at King Tubby’s by Tubby and Errol Thompson (Documented)

Quick start

Load
Analog, Saturator, EQ Eight, Echo, Hybrid Reverb
Set
  • Analog Osc 1 Rectangle PW around 30%, Osc 2 Saw −12 cents mixed lower; Filter 1 LP12 around 2.5 kHz
  • Amp envelope Attack 30–50 ms, Sustain 0.9, Release 120 ms; velocity to filter cutoff
  • Saturator Analog Clip, 3 dB; EQ Eight HP 250 Hz
  • Echo send, 1/4 dotted, FB 40%, Repitch
  • Hybrid Reverb send, spring IR, 1.4 s
You should hear
A breathy, reedy melodica playing short minor phrases that end on a long note while the echo answers.
Live edition
Suite

Why it matters. It is the dub album most often named as the best. Documented: Steve Barrow said it showed a studio engineer “could be considered as creative as the singers”, and Mojo placed it second on its 2024 list of reggae albums (Wikipedia).

Documented tools and method: the musicians were recorded at Randy’s between 1972 and 1975. Pablo then took the tapes to King Tubby’s for mixing, with King Tubby and Errol Thompson as mixing engineers. Pablo plays melodica, piano, clavinet and organ, over Carlton Barrett, Aston Barrett and Robbie Shakespeare (Wikipedia; credits and “Mixed At King Tubby’s Studio” on Discogs: Clocktower LPCT 0085). The title track is a dub of Jacob Miller’s “Baby I Love You So” (Ableton blog; The Quietus).

Signature moves → Live 12

  1. A melodica you can play (Documented instrument; synthesis By ear). Use Analog with Osc 1 Rectangle at PW around 30% and Osc 2 Saw at −12 cents, mixed lower. Set Filter 1 to LP12 at around 2.5 kHz with a little resonance. Amp envelope: Attack 30–50 ms (a reed needs breath to start), Sustain 0.9, Release 120 ms. Map velocity to filter cutoff for breath pressure. For a real melodica, record it close and put Saturator on it (Analog Clip, 3 dB).
  2. Minor, modal phrases that leave space (By ear). Play short phrases that end on a long note, and let the echo answer them. The roll below is a generic minor shape, not Pablo’s line.
  3. Record in one room, mix in another (Documented). Freeze and flatten your parts, or export dry stems, then open them in a new set with only the dub returns. Mixing away from the session makes you less precious about what was played.
  4. Dub the song, keep a ghost of it (Documented: a Jacob Miller vocal rhythm). Leave one or two vocal words in, sent to the spring return, so the song is still there under the dub.
A melodica patch with a fast attack sounds like an organ. Keep the attack at 30 ms or more, and add a little pitch drift (an LFO at 4–5 Hz on pitch, very low amount), or it will sound like a preset.

What to steal: a fragile lead over a heavy rhythm, mixed by someone else. The contrast between a small instrument and a massive drum and bass is the record.

Lee “Scratch” Perry & The Upsetters — “Super Ape” (1976): bounce, print, repeat

1976 · Island (ILPS 9417, UK) / Upsetter (Jamaica, as Scratch the Super Ape) · tempo By ear · key undocumented · Tools: four-track TEAC, Grampian spring reverb, Roland Space Echo, Mu-Tron Bi-Phase (Documented for the Black Ark)

Quick start

Load
Audio Effect Rack, Hybrid Reverb, EQ Eight, Saturator, Phaser-Flanger
Set
  • Drum bus rack chain 1 dry; chain 2 Hybrid Reverb spring IR, 100% wet, Decay 1–1.5 s
  • EQ Eight in chain 2, 48 dB/oct low cut at 250–400 Hz
  • Saturator Soft Sine, 6 dB; chain 2 raised until the snare gets metallic but the kick stays tight
  • Phaser-Flanger on a return, Phaser, slow LFO around 0.1–0.3 Hz, Feedback 50–70%, recorded to its own track
  • Bounce four audio tracks; when full, Bounce to New Track and delete the sources
You should hear
A metallic spring-soaked snare over a tight kick, and a dense mix built from generations of printed effects.
Live edition
Suite

Why it matters. It is the most famous Black Ark album and the clearest statement of Perry’s dense, smoky dub. Documented: Perry produced, engineered and arranged it at the Black Ark (Discogs: ILPS 9417). It was recorded on four-track equipment by repeated overdubbing, combining tracks to free space, and often arranged with later versions in mind. It uses vocal fragments by Perry, Prince Jazzbo, George Faith and Max Romeo (Wikipedia: Super Ape).

Documented studio: the Black Ark had a four-track TEAC, an Alice broadcast-style desk and later a Soundcraft, a Grampian spring reverb, an Echoplex, and later a Roland Space Echo and a Mu-Tron phaser (Ableton blog; Wikipedia: Black Ark). Gearnews names the TEAC A3340 quarter-inch four-track, with a second TEAC two-track for mixdown, and a Grampian 636 spring. Daniel Boyle, who rebuilt the setup, told Sound On Sound that Perry fed drums through the Grampian, rolled off its bottom end and blended it back in, much like parallel compression. He used the spring as a fuzz on guitar chords, then through the Bi-Phase and Space Echo (echo off). He layered many bounces on top of each other, and printed the Bi-Phase back to its own channel. Boyle calls the Bi-Phase “just the Lee Perry sound”. Perry used an AKG D12 on the kick (SOS).

Signature moves → Live 12

  1. Parallel spring on the drum bus (Documented technique). On the drum group, add an Audio Effect Rack with two chains. Chain 1 is dry. Chain 2 is Hybrid Reverb (spring IR, 100% wet, Decay 1–1.5 s), then EQ Eight with a 48 dB/oct low cut at 250–400 Hz, then Saturator (Soft Sine, 6 dB). Bring Chain 2 up under the dry kit until the snare gets metallic but the kick stays tight.
  2. Spring as fuzz on the guitar (Documented). Guitar → Hybrid Reverb (spring IR, 100% wet, short) → Roar or Saturator driven hard → Phaser-Flanger in Phaser mode → Echo with Feedback at 0 (the Space Echo’s spring and preamp, with no repeats). It is a distortion pedal built from a reverb.
  3. Print the phaser to its own track (Documented). Send a subgroup to a return with Phaser-Flanger (Phaser, slow LFO around 0.1–0.3 Hz, Feedback 50–70%). Record the return’s output onto a new audio track (set the track’s input to the return), then mute the return. The phase movement is now a part you can edit and mute.
  4. Bounce to free tracks (Documented). Allow yourself four audio tracks. When they are full, select two, use Bounce to New Track (Live 12.2) or freeze and flatten, and delete the sources. Every bounce bakes in the current effects and balance. That is the Black Ark density: generations of printed decisions.
  5. One mic on the kick (Documented). If you record drums, use one dynamic mic on the kick and accept its colour. In a Drum Rack, pick one kick sample and don’t layer it.
Bouncing destroys options. Save a copy of the set before every bounce generation (Save Live Set As …-gen2), or you will not be able to undo a balance you regret.

What to steal: print your effects and use the freed tracks for more. Density can come from generations of committed bounces, not from more tracks.

Joe Gibbs & The Professionals — “African Dub All-Mighty Chapter 3” (1977–78): effects as arrangement

Joe Gibbs Record Globe (Jamaica, dated 1977 on Discogs) / Lightning (LIP 12, UK, 1978) · Wikipedia dates it 1978 · tempo By ear (rockers) · key undocumented · Tools: Joe Gibbs’ 16-track studio at Retirement Crescent (Documented); effect units undocumented

Quick start

Load
Drum Rack, Collision, Operator, Roar, Spectral Time
Set
  • Drum Rack one non-musical hit per pad, played on phrase ends
  • Gong Collision, Mallet on Resonator 1 set to Plate, long decay, low brightness
  • Depth-charge Operator Osc A Sine, large negative Pitch Envelope, decay 400–800 ms, into Roar
  • Spectral Time on a return; one word sent, Freezer on just after the word, released on the next downbeat
  • Use one effect per 8 or 16 bars, where a fill would go
You should hear
Gongs, depth-charges and a frozen vocal moment dropped in where a drummer would fill, over rockers drums.
Live edition
Suite

Why it matters. Documented: FACT (David Katz) says Errol Thompson continued mixing the African Dub series at Gibbs’ studio, “of which Chapter 3 had the most lasting impact”. It cites the dub of Jacob Miller’s “Keep On Knocking”, retitled “The Entebbe Affair”, and describes a dub with gongs, synthesizer depth-charges and other dramatic effects. Bradley ranks it with the canon (Wikipedia: Pick a Dub).

Documented tools and method: Gibbs set up a 16-track studio and pressing plant at Retirement Crescent in 1975. Thompson, who had engineered at Randy’s, was the other half of “The Mighty Two” (Wikipedia: Joe Gibbs; Wikipedia: Errol Thompson). The Jamaican LP credits Thompson as engineer, remix engineer (with Gibbs) and producer, with Sly Dunbar on drums, Robbie Shakespeare and Lloyd Parks on bass, Tommy McCook, Bobby Ellis and Vin Gordon on horns, and Sticky Thompson on percussion (Discogs).

Signature moves → Live 12

  1. A sound-effects Drum Rack (Documented: gongs, depth-charges). Build a Drum Rack of non-musical hits, one per pad. Gong: Collision with the Mallet on Resonator 1 set to Plate, a long decay and low brightness. Depth-charge: Operator Osc A Sine with a large negative Pitch Envelope amount and a decay of 400–800 ms, into Roar. Add a door slam or a bell from your own recordings. Play them like a percussionist, on phrase ends.
  2. Effects on the 16-track grid (Documented: 16-track). With more tracks than Tubby’s four, Thompson could keep effects on their own channels. Put each effect hit on its own track with its own send amounts, so a gong can go to the spring and a depth-charge to the echo.
  3. Rockers drums under the drama (By ear; Sly Dunbar documented as drummer). Use the steppers/rockers kit from Dub Reggae: kick on every beat, rim on 3, fast offbeat hats.
  4. Freeze one vocal moment (Documented, on another Thompson dub). FACT describes Thompson freezing a snippet of reverberating vocal. In Live: put Spectral Time on a return, send one word to it, and switch the Freezer on just after the word (automate it, or map it to a pad). Release it on the next downbeat.
Sound effects become comedy if they land on every phrase. Use one per 8 or 16 bars, and always put it where a fill would go.

What to steal: use non-musical sound as part of the arrangement, on its own tracks and sends, played at phrase ends where a drummer would fill.

Blackbeard (Dennis Bovell) — “Strictly Dub Wize” (1978): the British engineer’s own mix

1978 · Tempus (TEMLP 001, UK) · tempo By ear · key undocumented · Tools: recorded at Gooseberry Sound Studios, London (Documented); effect units undocumented

Quick start

Load
EQ Eight, Electric, Chorus-Ensemble, Echo, Hybrid Reverb
Set
  • Bass mono below 150 Hz, EQ Eight shaping a 60–90 Hz bump rather than a deep sub
  • Electric clean keys let through the drop-outs
  • Chorus-Ensemble Chorus mode, low amount
  • Echo 1/8 dotted, FB 35%
  • Hybrid Reverb send, spring IR, 1.2 s
You should hear
Sound-system bass weight with warm, chorused keys sweetening the drop-outs.
Live edition
Suite

Why it matters. It is the dub album of the man who shaped British reggae from Matumbi to lovers rock. Documented: Bovell ran the Jah Sufferer sound system in South London, engineered at Dip Records (precursor of the Lovers Rock label), formed Matumbi and released dub albums as Blackbeard: Strictly Dub Wize (Tempus, 1978) and I Wah Dub (More Cut/EMI, 1980) (Wikipedia: Dennis Bovell).

Documented tools and method: the sleeve carries Bovell’s note that it is his own music, “mixed the way that I prefer”. It was recorded at Gooseberry Sound Studios, with Bovell, Dave Hunt and Mark Angelo Lusardi credited as engineers and John Hassell as mastering engineer (Discogs: TEMLP 001). Desk and effects are not documented in the sources checked here.

Signature moves → Live 12

  1. Write it, play it, mix it (Documented: his own music). Bovell is a multi-instrumentalist. Give yourself every role: program or play each part in your own set, then switch to the mixer and stop being the player. Lock the clips (no more editing) before the dub pass.
  2. Sound-system low end (Documented background; recipe By ear). UK sound-system culture prizes bass weight. Use the round dub bass from Dub Reggae, mono below 150 Hz, with EQ Eight shaping a 60–90 Hz bump rather than a deep sub. See Mixing for sound systems.
  3. Lovers-rock sweetness in the dub (By ear). Let a clean, chorused guitar or electric piano through the drop-outs: Electric or a Rhodes sample, then Chorus-Ensemble in Chorus mode at a low amount, then Echo at 1/8 dotted. That warmth is the British twist on the Jamaican drop-out.
British dub mixes were cut for big systems. A mix that sounds full on headphones can be thin on a rig. Check the 60–110 Hz region on a sub or in a spectrum analyser, not only by ear on small speakers.

What to steal: mix your own music as a different person from the one who wrote it, “the way that I prefer”.

Linton Kwesi Johnson — “LKJ in Dub” (1980): dubbing a poetry record

1980 · Island (ILPS 9650, UK) · tempo By ear · key undocumented · Tools: mixed by Dennis Bovell (Documented); studio and gear undocumented in the sources checked

Quick start

Load
EQ Eight, Echo, Hybrid Reverb
Set
  • Arrangement written around a spoken guide vocal, then the voice muted for the dub
  • Horns one group, Echo send at 1/4 dotted, FB 40%
  • Harmonica EQ Eight HP 400 Hz, LP 5 kHz, Echo at 1/8, FB 25%
  • Spring Hybrid Reverb, spring IR, 1.2 s
  • Bass moving with the speech rhythm, never muted for long
You should hear
Horns and harmonica answering lines that are no longer there, so the gaps become the space.
Live edition
Suite

Why it matters. Dub poetry places the voice at the centre, so the dub of a dub-poetry album is a strange object: the poem is removed and the band arrangement that was written around it is left. Documented: it consists of dub versions of tracks from Johnson’s Forces of Victory (1979) and Bass Culture (1980), produced by Johnson and Bovell (as Blackbeard) and mixed by Bovell (Wikipedia: LKJ in Dub). Johnson’s work is usually written in collaboration with Bovell, and he mostly recorded with the Dub Band (Wikipedia: Linton Kwesi Johnson).

Documented personnel: Floyd Lawson and Vivian Weathers (bass), Jah Bunny and Winston Curniffe (drums), John Kpiaye (guitar), Julio Finn (harmonica), Bovell and Webster Johnson (keyboards), Rico Rodriguez (trombone), Dick Cuthell (trumpet, flugelhorn), Henry “Buttons” Tenyue, and James Danton (alto). Every track title is a song title plus “Dub” (Discogs: ILPS 9650).

Signature moves → Live 12

  1. Arrange around a voice, then remove it (Documented source albums). Write the arrangement with a spoken guide vocal so the horns and harmonica answer between lines. Mute the voice for the dub. The answers are now calls with nothing to answer, and the gaps become the space.
  2. Horns and harmonica as the lead (Documented players; treatment By ear). Put the horn section on one group with an Echo send at 1/4 dotted. Give the harmonica a narrow band (EQ Eight HP 400 Hz, LP 5 kHz) and a short spring. They take the voice’s role in turn.
  3. Keep the poem’s rhythm in the bass (By ear). Dub poetry bass lines often move with the speech rhythm. In the dub, the bass is the last trace of the words: never mute it for long.
The words of a poem are its content. Do not transcribe or sample a poet’s voice without permission. For practice, record your own spoken guide and delete it.

What to steal: write the band around a voice, then take the voice away. The arrangement keeps its shape and the dub writes itself.

Scientist — “Scientist Rids the World of the Evil Curse of the Vampires” (1981): one pass, high-pass, no phaser

1981 · Greensleeves (GREL 25, UK) · produced by Henry “Junjo” Lawes · tempo By ear · key undocumented · Tools: rhythms at Channel One, mixed at King Tubby’s (Documented); high-pass filter (Documented as his sound, not per track)

Quick start

Load
Auto Filter, Echo, Hybrid Reverb, Limiter
Set
  • Auto Filter on the drum or skank group: High-pass, 24 dB slope, Resonance around 40–60%, Frequency moved by hand in small steps
  • Echo send, 1/8 dotted, FB 50%
  • Hybrid Reverb send, spring IR, 1.5 s
  • Mix Automation Arm on, the whole mix performed in one take
  • Changes every 4 or 8 bars change one thing; Limiter on the group
You should hear
A swirl like phasing from a hand-moved resonant high-pass over hats and cymbals, in an unpredictable one-pass mix.
Live edition
Suite

Why it matters. The Greensleeves Scientist albums brought dub to a new audience with comic-book themes on the sleeves, over the hardest early-80s rhythms from the Roots Radics. Documented: the sleeve says the rhythm tracks were laid at Channel One and the album was mixed at King Tubby’s at midnight on Friday 13 June 1981. It credits Scientist as mixer, Lawes as producer, and the Roots Radics, with Errol “Flabba” Holt on bass and Style Scott and Carlton “Santa” Davis on drums (Discogs: GREL 25; Wikipedia). The material comes from Michael Prophet, Wailing Souls, Johnny Osbourne and Wayne Jarrett rhythms (Wikipedia).

Documented method: Scientist started at Tubby’s winding transformers and learned from books and electronics experiments. He worked on a four-track there, and later on a 16-track desk at Channel One (Wikipedia: Scientist). Asked about reverb and tape echo, he told Tape Op “No phasers. High-pass filter.”, crediting the filter for the sound people mistook for phasing. He describes dub as recomposing a recorded performance with surprise and pattern changes. Early sessions were usually one take, recorded and mixed straight away in a short booking.

Signature moves → Live 12

  1. The “phaser” that is a filter (Documented). Use a resonant high-pass moved by hand, not an LFO phaser. Auto Filter on the drum or skank group: High-pass, 24 dB slope, Resonance around 40–60%. Map Frequency to a knob and move it in small steps by hand. The resonant peak moving over hats and cymbals gives a swirl that sounds like phasing. Or use the stepped rack from Dub from the Roots, since he mixed on Tubby’s board.
  2. Mix in one pass (Documented attitude). Arm automation, press record in Arrangement and perform the whole mix with mutes, sends and the filter from start to finish. Keep the take. Edit only clear mistakes. If you need four hours, you are overthinking.
  3. Surprise through pattern change (Documented aim). Plan nothing. Every 4 or 8 bars, change one thing: mute the bass, open the filter, throw the snare to the echo, bring in a percussion part that was absent. The listener should not be able to predict the next 8 bars.
  4. Radics weight (By ear). A dry, loud snare with a short room, a deep round bass and very little else. The Radics’ slower, heavier rockers leaves more space than the mid-70s Aggrovators. See dub bass.
High resonance on a high-pass at low frequencies can produce a loud peak in the bass. Keep the filter on the drum and skank groups, never on the bass, and put a Limiter on the group.

What to steal: mix fast and live, and let a hand-moved filter do what you would otherwise reach a phaser for.

Prince Jammy — “Prince Jammy Destroys the Invaders…” (1982): drum and bass with the effects washing over

1982 · Greensleeves (GREL 29, UK) · tempo By ear · key undocumented · Tools: recorded at Channel One, mixed and produced at King Tubby’s (Documented); gear undocumented

Quick start

Load
Echo, Hybrid Reverb, EQ Eight, Utility
Set
  • Echo return, 1/4, Feedback 60–70%, LP 2.5 kHz in the loop
  • Hybrid Reverb spring IR into a Hall tail via Blend, 3–4 s
  • Bands skank high-passed at 250 Hz, keys and returns at 300 Hz
  • Utility Width 130% on the returns only; bass in mono
  • Throws whole bars of skank and keys sent, then the dry sources muted
You should hear
A loud, dry, clear drum and bass with huge, band-limited washes rising over it.
Live edition
Suite

Why it matters. It is the sister album to Scientist’s Greensleeves run, with the same Roots Radics and the same studios, mixed by Tubby’s other great apprentice a few years before his digital revolution. Documented: the sleeve credits Jammy as producer and mixer and the Roots Radics as backing band, recorded at Channel One and mixed at King Tubby’s (Discogs: GREL 29). His dubs are known for their clear sound and their effects (Wikipedia: King Jammy).

Documented method: Jammy told United Reggae that he learned Tubby’s mix style mainly by watching him work. He describes those dubs as “drum and bass and echo reverb and delay washing all over”. Rhythms were built at Channel One, voiced at King Tubby’s and cut to dubplate. He also says Tubby built his own delay for the sound system.

Signature moves → Live 12

  1. Drum and bass as the floor (Documented). Build the mix with drums and bass loud, clear and dry, then add nothing permanent above them. Everything else comes and goes through the returns.
  2. Effects washing over (Documented; settings By ear). Use two returns that run long: Echo at 1/4 with Feedback 60–70% and the LP inside the loop at 2.5 kHz, and Hybrid Reverb (spring IR into an algorithmic Hall tail via Blend, 3–4 s). Throw whole bars of skank and keys into them, then mute the dry sources so only the wash is left over the drum and bass.
  3. Clarity through separation (Documented reputation; method By ear). Give every source its band: high-pass the skank at 250 Hz, the keys at 300 Hz and the returns at 300 Hz. Keep the bass in mono and the returns wide (Utility Width 130% on the returns only). The effects can then be huge without blurring the low end.
  4. Think in dubplates (Documented workflow). Export each finished mix as a single file and play it on a big system, not just monitors, before you call it done.
Long returns on a dry mix pile up. Put Echo’s Ducking on (Character tab), or a Compressor sidechained from the drum group on the reverb return, so the wash rises in the gaps and sinks under the hits.

What to steal: a dry, clear drum and bass with huge, band-limited washes above it. Clarity lets the effects be bigger.

Jah Shaka — “The Commandments of Dub” (1982): mixing for one sound system

1982 · Jah Shaka Music (SHAKA 824, UK) · tempo By ear (steppers feel) · key undocumented · Tools: recorded at Ariwa Studios, mixed by Jah Shaka and Neil Fraser (Mad Professor), Shaka credited for bass and effects (Documented)

Quick start

Load
Audio Effect Rack, EQ Eight, Operator, Echo
Set
  • EQ Eight in an Audio Effect Rack, two bell bands only: 60–120 Hz and 2–6 kHz, gains and frequencies on four macros
  • Echo after the siren, Feedback 50–60%, Wobble up, Repitch on
  • Syndrum Operator Osc A Sine or Triangle, fast-falling Pitch Envelope, decay 300–600 ms, a little noise from Osc B
  • Master quiet, 6 dB of headroom until the final render
You should hear
Bass weight from frequency and Q rather than loudness, a swooping siren echo and a live syndrum over the dub.
Live edition
Suite

Why it matters. Shaka ran a South London roots sound system from 1970 and became its spiritual figurehead. He released dub albums, often under the Commandments of Dub banner, on his own label from 1980 (Wikipedia: Jah Shaka). The record joins two lineages on this page: it was made at Mad Professor’s Ariwa.

Documented tools and method: the sleeve credits Jah Shaka as producer, mixer, bassist (on most tracks) and effects, with Neil Fraser (Mad Professor’s real name) also credited for the mix and for remix. Drums are by Wazair and Erroll Drummie, guitar and melodica by Hughie Izachaar, and it was recorded at Ariwa Studios. The tracks are titled “Verse 1” to “Verse 10” (Discogs: SHAKA 824). In his RBMA lecture Shaka said: “We depended more on frequency than the power.” He explained that his pre-amp has a parametric on the bass and one on the top, with the crossover built in, and that he adjusts it per record. His siren box dates from around 1969 and feeds a single delay, an H&H tape-loop unit, and he plays a syndrum live. Vivien Goldman’s 1981 NME report describes him playing syndrum over the record and riding the amp’s knobs alone at the controls (NME via uncarved.org).

Signature moves → Live 12

  1. A two-band parametric as the master (Documented). On the master or a playback group, set EQ Eight with only two bell bands: one in the bass (60–120 Hz) and one in the top (2–6 kHz). Map both gains and frequencies to four macros in an Audio Effect Rack. Set them per track and leave them, as he describes.
  2. Siren into one delay (Documented). Use the siren from Dub FX & Delays. Insert one Echo after it, with no chain of effects: Feedback 50–60%, Wobble up for a tape-loop feel, and Repitch on so time changes swoop.
  3. Play a syndrum over the dub (Documented). Build a syndrum voice: Operator Osc A Sine or Triangle with a fast-falling Pitch Envelope, decay 300–600 ms, and a little noise from Osc B. Put it on a pad controller and play it live over the record, recording the performance.
  4. Frequency, not loudness (Documented principle). Mix with the master quiet. Get the weight from the bass band’s frequency and Q, not from limiting. Keep 6 dB of headroom on the master until the final render.
A per-record EQ is for playback, not mastering. If you print the bass boost into your file, it will be wrong on every other system. Keep the playback rack on a separate group and bypass it for the export.

What to steal: mix for the system you will play on. Use a few fixed controls and set them per record, rather than a loudness chain.

Black Uhuru (Sly & Robbie) — “The Dub Factor” (1983): Syndrums in a major-label dub

1983 · Mango / Island (MLPS 9756) · tempo By ear · key undocumented · Tools: Syndrums (Documented credit), recorded at Channel One and Compass Point, remixed at the Fallout Shelter by Paul “Groucho” Smykle (Documented)

Quick start

Load
Operator, Echo, Vocoder
Set
  • Syndrum Operator Osc A Sine, fast-falling Pitch Env, Decay 400 ms, Osc B Noise at low level
  • Fills Osc A Coarse on a macro, fills falling in pitch across 3–4 hits
  • Echo send, 1/8 dotted, FB 50%
  • Vocoder on a synth pad, a vocal fragment as the modulator, thrown to the echo
  • Use three or four syndrum fills a minute at most
You should hear
Electronic tom fills falling in pitch and trailing into echo over a dry rockers rhythm, with half-voice vocoder throws.
Live edition
Suite

Why it matters. It is Sly & Robbie’s rhythm section at its most electronic, dubbed for an international label, and it pointed dub toward the 80s digital era. Documented: the US LP credits Sly & Robbie as producers, Sly Dunbar on drums and Syndrums, Robbie Shakespeare on bass, Wally Badarou on synth and vocoder, and Paul “Groucho” Smykle for the remix, with assistant remix engineer Stephen Street. It was recorded at Channel One and Compass Point, Nassau, and remixed at the Fallout Shelter. Discogs notes that the back cover says “Remix” and the labels “Dub Mix” (Discogs: MLPS 9756).

Documented method: in their RBMA lecture, Sly says “It was the Syndrum first”: he bought a pair with tour savings, and later a Simmons for Black Uhuru’s Anthem. He describes Channel One as a studio whose owner loved bass and drums. In the 1976 Rockers beat they introduced a harder pattern that replaced the one drop (Wikipedia: Sly and Robbie). Smykle’s console moves are not documented in the sources checked here.

Signature moves → Live 12

  1. Syndrum fills over a real kit (Documented instrument). Put an Operator syndrum (as in the Shaka entry) on its own track, but tune it: map Osc A Coarse to a macro and play fills that fall in pitch across 3–4 hits (high to low pads). Send it to an Echo at 1/8 so each hit trails.
  2. Rockers drive (Documented style). A kick on every beat or a heavy two-kick pattern, the rim on 3, and a tight, dry, gated snare. Gate on the snare with a short Hold, or Drum Buss Transients up. See one drop and steppers.
  3. Synths and vocoder as dub material (Documented: Badarou on synth and vocoder). Use Live’s Vocoder on a synth pad with a vocal fragment as the modulator, then throw it to the echo. In the dub, a vocoded word is half instrument, half voice.
  4. A dub mix done in a different country (Documented). The rhythms were recorded in Kingston and Nassau and remixed in London. Export dry stems and dub them in a fresh set, with returns built for this mix only.
Syndrum sounds date fast if they are everywhere. Use them as fills and accents, three or four times a minute at most, over a real kit or a sampled one.

What to steal: new electronic drums used as fills over a played rhythm section. The new sound is an accent, not a replacement.

Massive Attack v Mad Professor — “No Protection” (1995): feedback loops over 48 tracks

1995 · Wild Bunch / Circa (WBRCD 3, UK & Europe) · tempo By ear (follows Protection) · key undocumented · Tools: SSL desk, two locked 24-tracks, Roland SDE-3000 delays, Lexicon 480L (Documented for Ariwa)

Quick start

Load
Echo, Hybrid Reverb, Limiter
Set
  • Return A Echo, Wobble 0, Noise 0, Repitch off, 1/4 dotted, FB 45%
  • Return B Hybrid Reverb, Hall 3 s
  • Cross-sends A → B at −12 dB, B → A at −18 dB, ridden by hand
  • Limiter on each return, −1 dB
  • Source stems cut into 2–4 bar blocks and rearranged before the dub
You should hear
Echo and reverb feeding each other into loops that build and collapse, over a song edited until its structure is gone.
Live edition
Suite

Why it matters. It closes this page’s window and carries the roots-dub method into the trip-hop era. Documented: after Protection, Massive Attack asked Mad Professor to remix one song, then more, until he had remixed nearly the whole album. He heavily edited the original material into a slow, beat-heavy mix with a lot of reverb and vocals that fade in and out. Every track was retitled, for example “Protection” as “Radiation Ruling the Nation” (Wikipedia: No Protection). The CD credits Mad Professor for the remix, mixed at Ariwa Studio (Discogs: WBRCD 3). The originals on Protection are analysed on another page. This entry is only about the dub.

Documented tools and method: Mad Professor told Sound On Sound (2007) that No Protection was worked from two 24-track tapes locked together, 48 tracks at 30 ips, with Lynx synchronisers. Ariwa’s core is an SSL desk. A Lexicon 480L is always on the first stereo aux, MXR phasers are wired in, and he keeps Roland SDE-3000 delays in every room, calling the first one “a magic delay”. His basic technique is feedback loops, built on the aux sends or through the SSL’s in-line channels, and he sets effect lengths by feel. On NiceUp he says he would rather go inside the effects than play an instrument, and pushes them to the edge of distortion and beyond.

Signature moves → Live 12

  1. Feedback loops on the sends (Documented). Make Return A an Echo and Return B a Hybrid Reverb. In the mixer, turn up Return A’s send to Return B and Return B’s send to Return A a little. That is a loop: the echo feeds the reverb, which feeds the echo. Put a Limiter at the end of each return. Ride the cross-sends to build and collapse the loop. The safe runaway method is on Dub FX & Delays.
  2. A reverb that is always on (Documented: 480L on aux 1). Every track gets a small amount of Return C (Hybrid Reverb, a Hall at 2–3 s, HP 250 Hz). It is the room the whole mix lives in. The dub throws happen on A and B above it.
  3. Clean digital delay, not just tape (Documented: SDE-3000). For the main delay, use Echo with Wobble and Noise at 0, Repitch off and the filter wide open, or the Delay device with Sync on. Clean repeats keep their pitch and level, which is what lets them loop without turning to mush.
  4. Edit the source before you dub it (Documented: heavily edited). Cut the original stems into 2–4 bar blocks in Arrangement and rearrange them until the song’s structure is gone. Then dub the new structure.
Return-to-return sends can run away fast and stay at full level. Always have a Limiter on both returns, and keep a hand on the cross-send macro. Do not leave the loop running while you adjust something else.

What to steal: make effects feed each other, and edit the source until it is yours before you start the dub.

Patterns across the canon

Read across all thirteen and the same decisions keep coming up, from a four-track in Waterhouse to an SSL in South London.

  1. The engineer is the artist. It is Documented in Barrow on Rockers Uptown, Scientist describing dub as recomposing a performance (Tape Op), Mad Professor going inside the effects (SOS, NiceUp), and the sleeve credits throughout. The performance is the mix, not the playing.
  2. Strip to drum and bass, then let things back in. It is Documented for Tubby (Quietus) and Jammy (United Reggae), and By ear on every record here. The drum and bass is the floor, and everything else visits.
  3. A filter you play, not a modulator. Tubby’s stepped Altec (Documented), Scientist’s high-pass instead of a phaser (Documented), and Perry’s phaser printed to its own track (Documented) all put movement under the hand.
  4. Recycle, retitle, re-voice. Hudson’s themed LP, Morris’ dub of the Ja-Gan sessions, LKJ in Dub, The Dub Factor and No Protection are all Documented as new versions of existing recordings, usually with new titles.
  5. Limits make density. Four tracks forced Perry to bounce and print (Documented). Sixteen let Thompson put effects on their own channels (Documented 16-track; the per-channel method is By ear). Forty-eight gave Mad Professor room for loops (Documented). Each step changed the sound.
  6. Mix for a system. Shaka’s two parametrics (Documented), Jammy’s dubplates (Documented), Bovell’s sound-system background (Documented): the target is a speaker stack, not a pair of monitors.
LineageRecords hereDocumented key toolLive 12 core
Waterhouse (Tubby’s)Dub from the Roots, Rockers Uptown, Vampires, Destroys the InvadersMCI desk, Altec stepped high-pass, homemade delayChain-selector HP rack, Echo with Repitch, one-pass automation
Black ArkSuper ApeTEAC four-track, Grampian spring, Bi-Phase, Space EchoParallel spring rack, Phaser-Flanger printed, Bounce to New Track
Joe Gibbs / Randy’s / Harry J’sAfrican Dub 3, Pick a Dub, Morris On Dub16-track (Gibbs), effects as arrangement, versioningFX Drum Rack, Save As for the dub set
UK sound systemsStrictly Dub Wize, LKJ in Dub, Commandments of DubSound-system playback EQ, siren into one delay, syndrumTwo-bell EQ Eight macros, Operator siren and syndrum
International / 80s–90sThe Dub Factor, No ProtectionSyndrums and vocoder; SSL, 48 tracks, SDE-3000, 480LSyndrum rack, Vocoder, return-to-return feedback loops

Related