Leads

Operator FM lead (ratio 1:2 with filtered feedback)

Two sines in series, modulator at ratio 2 or 3, modulator level under an envelope so the note starts bright and settles. The DX-era "brass/bell" lead, and with a square carrier it becomes a hard digital techno lead.

1. Operator, algorithm B→A (modulator B into carrier A). Osc A Sine, level 0 dB, Coarse 1. Osc B Sine, Coarse 2, level −12 dB. Raise B until the sound gets a brassy edge; around −8 dB it is a trumpet, at −2 dB it is a buzzsaw.
2. Osc B envelope: A 0, D 400 ms, S −12 dB, R 200 ms. The FM amount now decays, so each note starts bright and settles, which is what the ear wants from a lead.
3. Osc A envelope: A 5 ms, D 300 ms, S −6 dB, R 250 ms.
4. Voices 1, Glide on, Glide time 50–80 ms. Operator’s glide is legato-only, so overlapping notes slide and separate notes do not.
5. Operator LFO: Sine, 5.5 Hz, destination pitch, amount 10–15 cents, with the LFO envelope attack around 300 ms so the vibrato comes in after the note settles.
6. Filter: Operator’s low-pass at 6 kHz, a little resonance, to round the top end. Tone at 60% if the FM aliasing gets harsh.
7. Saturator Soft Sine 3–5 dB, Chorus-Ensemble Classic 15%, Echo 1/8 dotted with 30% feedback and the high-pass at 400 Hz on the delay.

Variations: Osc A Square for a harder techno lead; add Osc C at ratio 1 modulating B (serial D→C→B→A with D off) at −20 dB for a growl under the tone; Operator Spread 30% for a stereo lead in the synth. See Operator.

Wavetable supersaw and unison modes

The seven-voice detuned saw stack. Wavetable’s Unison section does it in one device: Classic mode detunes a set of voices around the centre pitch; Shimmer adds a slowly moving pitch spread; Noise randomises per-voice position. The classic big-room supersaw is Classic mode, 5–8 voices, with the fundamental reinforced by a mono Sub.

1. Wavetable. Osc 1: Basics → Saw (or Vintage category for an analogue-flavoured saw). Osc 2: same table, Transpose +12 st or −12 st, level −8 dB, for octave thickness.
2. Unison: Classic, Voices 6 (4 for a thinner retro lead, 8 for a wall), Amount 30–45%. In Classic the amount is detune spread; above 50% the pitch centre gets vague and the lead stops being a note.
3. Sub oscillator on, level −6 to −10 dB, Tone 0 (sine). The Sub is not affected by unison, so it anchors the pitch in the centre of the stereo image.
4. Filter 1: OSR low-pass, cutoff 5–7 kHz, resonance 10%, Env 2 → cutoff +25% with a 300 ms decay so the attack sparkles.
5. Amp env (Env 1): A 10 ms, D 300 ms, S 80%, R 350 ms. Trance leads: A 0, longer release. Hyperpop: A 0, D 150, S 60%.
6. Mod matrix: LFO 1 (Sine 5.8 Hz, fade-in 400 ms) → Osc 1 and 2 Pitch at 10 cents for vibrato; velocity → Filter 1 cutoff +20%.
7. Post: EQ Eight HP 120 Hz (the Sub carries the bass, nothing else should), a −3 dB bell at 400 Hz to remove supersaw mud; Saturator Analog Clip 2–4 dB to glue the voices; Chorus-Ensemble Ensemble mode 15–25% (three-voice ensemble, the thing that turns a saw stack into "trance"); Hybrid Reverb Shimmer 15–20%, pre-delay 20 ms.

Variations: Shimmer unison with Osc 1 on a Vintage table for the synthwave pad-lead; Random Note unison at 3 voices with small amount for a hyperpop chord-lead; Filter 2 as a Notch swept by LFO 2 at 0.1 Hz for a phasey top. See Wavetable.

Analog mono lead with glide and filter drive

The monosynth lead: saw plus square an octave down, 24 dB low-pass with drive, legato glide, LFO vibrato. Analog is the closest thing in Live to a Moog-style lead and its Quick Routing makes the two-filter layout easy.

1. Analog. Osc 1 Saw, level 0 dB. Osc 2 Square, Octave −1, level −6 dB, Detune +6 cents. Quick Routing: both oscillators into Filter 1 (the first routing option).
2. Filter 1: LP 24, cutoff 1.2 kHz, resonance 20%, Drive on. Filter envelope: A 0, D 350 ms, S 30%, amount +40%.
3. Amp 1 envelope: A 2 ms, D 200 ms, S 80%, R 150 ms. Voices: set the polyphony to 1, enable Legato, Glide on, Glide time 60–100 ms. With Legato on, glide happens only when notes overlap, so write the MIDI with overlaps where you want slides.
4. LFO 1: Sine 5.5 Hz, Delay (the LFO’s fade-in) 250 ms, routed to Osc 1 and 2 pitch at 10–20 cents. Analog’s vibrato parameters in the global section do this too if you prefer one control.
5. Overdrive 15–25%, Tone 50%, Dynamics 50% after: Analog’s filter drive is subtle, Overdrive gives the fur.
6. Delay ping-pong 1/8 at 25% feedback, filter in the delay at 300 Hz–5 kHz, Dry/Wet 20%.

Variations: Osc 2 Sync on (Osc 2 synced to Osc 1, pitch envelope on Osc 2 +24 st) for the hard-sync tearing lead; Filter 2 as HP 12 at 200 Hz in series to clean the bottom. See Analog and Drift, acid 303.

Drift lead: voice modes, Shark Tooth, MPE

Drift is the fastest lead in Live. Its voice modes (Mono, Poly, Stereo, Unison) change the whole character with one menu: Mono for a glide lead, Stereo for a lush two-voice-per-note spread, Unison for a cheap supersaw. Drift amount adds analogue instability, and it is MPE-ready so a Push 3 pad slide becomes per-note pitch and pressure.

1. Drift. Osc 1 Shark Tooth (between saw and triangle, less fizz than a pure saw). Osc 2 Saw, Octave −1, level −8 dB, Detune +7 cents. Noise 0 (or 5% for air).
2. Filter MS2 low-pass, cutoff 1.5 kHz, resonance 15%, Drive 30%. Env 2 → filter cutoff +35%, Env 2: A 0, D 400 ms, S 20%.
3. Envelope 1 (amp): A 3 ms, D 250 ms, S 75%, R 200 ms.
4. Voice mode: Mono with Legato on and Glide 70 ms for the classic mono lead. Try Stereo: each note gets two voices panned apart with a slight detune, a wide lead with no chorus. Unison: a four-voice detuned stack with Unison amount as the spread. Poly for chord-leads.
5. Drift amount 20–35%: random slow pitch and filter wander per voice, the analogue "not quite in tune" that makes it feel like hardware.
6. LFO: Sine 5.5 Hz → Pitch 12 cents, with the LFO envelope-attack so vibrato fades in; or map the mod wheel to LFO amount in the matrix for played vibrato.
7. MPE: enable it in the device; in the mod matrix route Slide (per-note Y) → filter cutoff, Pressure → LFO amount (vibrato on press) and per-note pitch bend is automatic. On Push 3 this is a lead you can bend per finger.
8. Post: Pedal in Overdrive mode at 15% for amp grit, Chorus-Ensemble Vibrato mode at a low amount, Echo with left 1/4 and right 1/8 dotted at 35% feedback.

See MPE expressive sounds and tuning, scales and MPE.

Meld chip and swarm leads

Meld’s macro oscillators each have engine-specific controls, so a lead can be a Chip engine (pulse width, bit-reduced, instantly 8-bit) on one side and a Swarm Sine (a cloud of detuned sines, a soft chorus-like mass) on the other, blended. Per-oscillator filters and the mod matrix with two LFOs and two envelopes make it a complete lead synth, and the Scrambler (12.2) adds rhythmic chaos on demand.

1. Meld. Oscillator A: engine Chip. Its first macro control shapes the pulse width (narrow pulse = thin and NES-like, 50% = square), the second the bit reduction/quantisation. Set pulse width 25–35%.
2. Oscillator B: engine Swarm Sine. Controls: swarm density (number of sines) and spread (detune). 6–10 voices, spread 30–45%, Octave +1. This is the soft, wide halo under the hard chip voice.
3. Per-osc filters: A → low-pass 24 dB at 2 kHz with Envelope 2 → cutoff +40% (D 300 ms); B → high-pass at 300 Hz so the swarm does not clutter the low mids.
4. Envelope 1 (amp): A 0, D 200 ms, S 70%, R 150 ms. Voice mode mono with glide for a chip lead; poly for the swarm chord-lead.
5. Mod matrix: LFO 1 Square at 1/16 synced → Osc A pulse width at 20% for a PWM pulse; LFO 2 Sine 5.8 Hz → pitch 10 cents (vibrato). Velocity → A filter cutoff.
6. Scrambler (12.2): a pitch/rhythm scrambler applied to the oscillator with a rate and amount; turn it up on a build or map it to a Push knob for a burst of random octave jumps on a hyperpop lead.
7. Saturator Hard Curve 3–5 dB for a digital edge, Chorus-Ensemble Classic 20% on the whole thing, a short 1/16 Delay for the chip "stutter".

Variations: Osc A engine FM for a DX-style lead with the Swarm Sine as its chorus; Osc A Dual Saw with the second saw detuned for a two-voice saw lead inside one oscillator; Osc B Formant with the vowel control on an LFO for a talking lead. See Meld.

Simpler sampled-vocal lead

One sung vowel or a word, held in Simpler Classic mode with a loop, played as a melody. Warp in Complex Pro keeps the formants (so a high note is not a chipmunk), or turn Warp off and embrace the pitch-shift formants as the sound. Glide and vibrato make it a lead rather than a sample.

1. Drop a sustained vocal sample into Simpler. Classic mode. Set Start just after the consonant, Loop on, loop region over a steady part of the vowel, Loop crossfade 20–40 ms so it does not click.
2. Warp on, mode Complex Pro, Formants at 0 for natural pitch-shifting; move Formants with the note (map a MIDI note-to-formant via Expression Control or an LFO) for the stylised chipmunk/giant trick. Warp off for the classic UK garage pitched-vocal chop sound where high notes go thin and fast.
3. Simpler Filter: LP 24, 5 kHz, drive 10%. Amp env: A 10 ms, D 200 ms, S 90%, R 200 ms.
4. Voices 1, Glide on, 40–60 ms, so legato lines slide like a singer.
5. Simpler LFO: Sine 5.5 Hz → pitch at 15 cents with Attack 300 ms. Alternatively Auto Shift (12.1) after Simpler: pitch correction to the scale of the track, its Vibrato section at 20% and 5.5 Hz, and the Formant control to re-shape the voice.
6. Chorus-Ensemble 20%, Echo 1/8 with Reverb 30% inside Echo, EQ Eight HP 200 Hz and a −4 dB bell at 2.5 kHz if the vowel gets nasal.

See voice and vocals, vocal chop rack.

Vibrato: LFO, Expression Control (aftertouch, mod wheel) and Auto Shift

Vibrato is the thing that makes a lead sound played. Three implementations: the synth’s own LFO with a fade-in (automatic, same every note), Expression Control mapping mod wheel or aftertouch to the LFO amount (played, from Push or a keyboard), and Auto Shift’s vibrato on an audio chain (works on anything, including resampled leads).

1. Rate 5–6.5 Hz. Singers and string players sit around 5.5–6 Hz; slower reads as a wobble, faster as a trill. Depth 10–30 cents for a natural vibrato, 50+ cents for a stylised one (synthwave, trance).
2. Fade-in: 150–400 ms. Every Live synth LFO has an attack or fade-in: Operator (LFO envelope), Wavetable (LFO Attack), Analog (LFO Delay/Attack), Drift (LFO shape with an envelope target), Meld (LFO fade). Set it so the note starts straight and the vibrato blooms. Retrigger the LFO per note so the fade restarts.
3. Played vibrato: put the Expression Control MIDI effect in front of the synth. Source Aftertouch (or Mod Wheel), mapped to the synth’s LFO amount, Min 0 Max 100%, curve slightly exponential so a light touch adds a little. Push 3 pads send per-pad pressure; with MPE enabled this becomes per-note vibrato.
4. Mod wheel → filter cutoff or FM amount at the same time, so pushing the wheel adds vibrato and brightness together, like a player digging in.
5. Auto Shift on audio: vibrato Rate 5.5 Hz, Amount 20%, with the fade so it does the same thing to a resampled or sampled lead; add pitch correction to the scale and you can tune a sloppy sample at once.
6. Pitch bend range: set the synth’s bend range to ±2 st for subtle bends, ±12 for the dive-bomb; Push and most keyboards send the same bend, the range is a device setting.

See modulation and Max for Live for the LFO device as an alternative source.

Detune, width and effect staging: saturation → chorus → delay → reverb

Lead width comes from detune first (unison, two oscillators a few cents apart, Drift Stereo mode), then from stereo modulation (Chorus-Ensemble), then from a stereo delay; reverb comes last and is kept small. Saturation goes before the modulation so the distortion stays coherent and the stereo effects spread the result. The order is not arbitrary: reversing chorus and saturation collapses the width, reversing delay and reverb makes a wash.

1. Detune inside the synth: two oscillators at ±6–12 cents, or unison 3–6 voices. Beyond 20 cents the pitch centre gets lost on a mono line; wide detune is for pads and chord-leads.
2. Saturator before everything: Analog Clip or Soft Sine 2–5 dB. It compresses the beating of the detuned voices into a denser tone and adds the 2–4 kHz harmonics that cut.
3. EQ Eight HP 100–150 Hz. Leads do not need anything below the lowest written note.
4. Chorus-Ensemble: Classic (two-voice chorus, retro and obvious), Ensemble (three-voice, lush, the trance/synthwave widener), Vibrato (pitch only, no mix). Rate 0.2–0.5 Hz, Amount 20–35%, Width 100%, Warmth on for a darker chorus. Keep the Dry/Wet at 30–50% so the centre survives.
5. Echo: left 1/8, right 1/8 dotted (polyrhythmic stereo), feedback 30–45%, Stereo control 50–70% to cross-feed, the Echo’s filter HP 300 Hz LP 5 kHz so the repeats are darker than the dry, Dry/Wet 20–30%. Gate/Duck in Echo: enable Duck so the repeats dip while the lead plays and swell in the gaps.
6. Hybrid Reverb: Quartz or Prism, Decay 1–1.5 s, pre-delay 20–30 ms (keeps the attack dry), low-cut 300 Hz, Dry/Wet 10–20%. Long reverbs on a lead push it back; if you want a big space, send to a return and duck the return with a Compressor sidechained from the lead.
7. Utility last: Width 100–120%, Bass Mono 120–150 Hz so the fundamental is dead centre regardless of what the effects did.

See delay design, reverb and space, saturation and distortion.

Doubling and octave layering in an Instrument Rack

Big leads are two or three leads: the main voice, an octave-down double with a dark filter (adds weight without changing the melody), and an octave-up or fifth-up double at low level with a bright filter (adds cut). In a rack each chain gets its own Pitch MIDI effect, filter and level, and the macros balance them.

1. Instrument Rack. Chain 1: the main lead from any recipe above.
2. Chain 2 "Sub double": Pitch MIDI effect −12 st, a simple saw or square in Drift or Analog, low-passed at 500–800 Hz, mono, −6 to −10 dB. It adds the chest to the lead; keep its envelope slightly slower (A 10 ms) so the main voice’s attack leads.
3. Chain 3 "Air double": Pitch +12 st, a thin FM or sine source, high-passed at 2 kHz, −12 to −16 dB, Utility width 130%. It adds cut on small speakers.
4. Chain 4 "Fifth": Pitch +7 st, used only in the chorus (chain selector on a macro). A parallel fifth makes a mono lead sound like a power chord; it breaks on complex melodies, so use it on simple hooks.
5. Set Glide on all chains to the same time so slides stay together. Macro-map the filters and vibrato amounts together.
6. On the rack output: Glue Compressor 2:1, 10 ms, 200 ms, 2 dB GR to gel the layers; then the effect staging chain from the previous recipe.

See layering rules and lead rack.

Grime square lead

The classic eskibeat lead: a raw square wave, little or no filter, short notes, hard pitch stabs and slides, often doubled an octave up, bone dry or with a single tight delay. It is ugly on purpose: the square’s odd harmonics at 1–4 kHz cut through a sparse 140 BPM beat without any processing. Assume 140 BPM.

1. Operator, all oscillators as carriers (the rightmost algorithm, A+B+C+D in parallel). Osc A Square, 0 dB. Osc B Square, Coarse 2 (an octave up), −10 dB. C and D off.
2. Osc A env: A 0, D 150 ms, S −3 dB, R 40 ms. Short release; grime leads stop dead. Voices 1, Glide 20–40 ms (fast slides between the stabs).
3. No filter, or Operator’s LP at 8 kHz just to stop the top aliasing. Pitch envelope: +2 st, 15 ms decay, for a slight bite on the attack.
4. Redux (12 bit, Downsample light) for the PlayStation-era digital grain. Sample rate reduction around 20 kHz adds aliasing that suits it.
5. Delay 1/8, feedback 15–25%, Dry/Wet 10–15%, filter in the delay HP 500 Hz. That is the whole chain; resist adding reverb.
6. Pattern: stabs on the 1/16 grid with octave jumps, a lot of rests; the lead is a rhythmic element. Push the Pitch MIDI effect in the clip for the pitch-dive at the end of a phrase.

See grime,grime deconstruction, UK bass 140.

Four genre leads: trance, synthwave, techno bleep, hyperpop

Four archetypes, each a known starting point plus the three or four changes that make the genre.

See psytrance (for the trance side), synthwave, deep techno, hyperpop.

Making a lead cut through: EQ, transient, sidechain to vocals

A lead that is lost in the mix is almost never a level problem. It is masked at 300–1200 Hz by chords, at 2–5 kHz by hats and vocals, or its attack is smeared by chorus and reverb. The fixes are subtractive EQ on the other elements, a transient on the lead, and if there is a vocal, a sidechain so the lead ducks the vocal’s band when the vocal is present.

1. On the chords/pad: EQ Eight −3 to −5 dB bell at the lead’s fundamental range (often 400–900 Hz) and a gentle high shelf −2 dB at 4 kHz. The lead gets its space from what you remove elsewhere.
2. On the lead: HP 150 Hz; a −3 dB cut at 400–600 Hz (muddy harmonics); +2 to +3 dB bell at 2.5–3.5 kHz Q 1 (presence). Do not boost above 8 kHz on a lead; that is hat territory.
3. Transient: Drum Buss with Drive 0, Transients +20–30%, Dry/Wet 100%; or the Multiband Dynamics upward expansion on the mid band with a fast attack. The attack is what locates the lead; chorus and reverb smear it, so restore it after them, or keep the effects on a send with the dry lead centred and dry.
4. Vocal present: Multiband Dynamics on the lead, sidechain input from the vocal, only the mid band (1–5 kHz) set to compress by 3–5 dB with 10 ms attack, 150 ms release. The lead stays full in the lows and highs and gets out of the way only where the vocal lives. A full-band Compressor sidechain is cruder but works; see sidechain ducking.
5. Arrangement: the lead should not play when the vocal sings the hook; call and response. The best EQ is a rest.
6. Reference on a phone: the lead must still be audible with no bass and no sub; if it is not, the 2–4 kHz region needs more, or the layers need the air double from the rack recipe.

See EQ strategies, transient design.

Live 12 tuning systems and Scale awareness for leads

Live 12 can load a tuning system for the whole set (.ascl/.scl files or the included ones) so every native instrument plays in it; clip Scale mode highlights or folds notes to a scale; the Scale MIDI effect forces incoming notes to it; and MIDI generators (Seed, Shape, Stacks) respect the scale. For leads this means two things: microtonal leads with native synths, and quick generation of melodies that cannot be wrong.

1. Set the Scale and root in the Control Bar; enable the Scale awareness for the set. New clips take the scale; the piano roll shows in-scale notes and Push folds the pad layout to it.
2. Tuning system: in the Tuning section of the set (opened from the Control Bar tuning chooser), load a system (e.g. a 19-EDO, a Pythagorean or a just-intonation file). Operator, Wavetable, Analog, Drift, Meld, Collision, Electric and Tension follow it; Simpler/Sampler follow where tuning is per-note; external plug-ins do not. A lead in 19-EDO on a Wavetable supersaw is a genuinely new timbre because the unison beating interacts with the different intervals.
3. Scale MIDI effect on the lead track with the same scale: anything you play on an unfolded controller lands in key. Set it to "fold" behaviour for the nearest in-scale note so slides from Push stay in tune.
4. Generate: Seed with a scale-aware setting and a limited range (an octave and a fifth) for a lead sketch; Shape to draw a melodic contour; then Ornament (12.1) to add grace notes and Strum for arpeggiated chord-leads. Humanize with Velocity Shaper afterwards.
5. Audio leads (samples, resampled lines): Auto Shift with its scale set to the set’s scale corrects them to the same tuning, including non-12-EDO tunings if the device is set to the set’s tuning.
6. Hardware: send the lead MIDI to the TD-3 or a eurorack oscillator via CV Tools; those will not follow a tuning system, so keep microtonal lines on native instruments and give the hardware the 12-EDO bass.

See tuning, scales and MPE, MIDI tools, generative sounds.