Wobble & Growl Bass

Basic LFO filter wobble — dotted and triplet divisions

The original wobble: a saw bass with a tempo-synced LFO on a resonant low-pass. The whole art is which division you choose and when you change it.

1. Analog: Osc 1 Saw, Osc 2 Saw at Octave −1, no detune. Filter 1 LP 24 fully open (we wobble with the external Auto Filter so the LFO shape and sync are easier to automate). Mono voice mode.
2. Auto Filter (12.2): LP, 24 dB, Frequency 2 kHz, Resonance 35%. LFO: sync on, Rate 1/8, Amount 60%, shape sine. Phase 0 so every bar starts at the same point.
3. Automate the LFO Rate in the clip envelope: bar 1–2 at 1/8, bar 3 at 1/8 dotted, bar 4 at 1/8 triplet then 1/16 on the last beat. Rate automation IS the bassline writing.
4. Saturator after the filter, Analog Clip 8 dB, so the resonance peak gets crunchy as it sweeps.
5. Sub: a separate track or rack chain with a sine (see three-way layering).

Automation vs LFO: drawing the wobble by hand

An LFO is predictable; a drawn automation curve lets every wobble be a different length and shape. Most brostep basslines are drawn, not synced.

1. Turn the Auto Filter LFO amount to 0. In the clip's envelope view, select Auto Filter > Frequency.
2. Draw with the pencil tool at 1/16 grid: short ramps for fast wobbles, long curves for slow sweeps. Hold Alt while dragging a segment to bend it into a curve — curved automation gives the "rubber" feel a linear LFO cannot.
3. Use the Live 12 clip envelope "Insert Shape" (right-click) with a triangle or saw shape over a selected region, then adjust its period per region — a middle path between LFO and freehand.
4. Copy the clip, change only the envelope, and you have a bass variation library.

Operator FM growl — ratios and feedback

Frequency modulation with non-integer ratios produces inharmonic sidebands that sound like a snarl. Modulating the FM depth with an LFO or envelope moves the sidebands — that is the growl.

1. Operator, algorithm: the fully serial one (D→C→B→A, leftmost). Osc A: Saw, Coarse 1, level 0 dB (the carrier; a saw carrier is already rich).
2. Osc B: Sine, Coarse 2, level −14 dB. This adds harmonic sidebands — "brightness".
3. Osc C: Sine, Coarse 3, Fine +50 (ratio ≈ 3.5), level −20 dB. The non-integer ratio produces inharmonic sidebands — the "snarl". Try Coarse 7 Fine 0 for a harder metallic tone.
4. Osc D: Sine, Coarse 1, level −30 dB with its envelope looping (Loop mode: Trigger, fast) — a pseudo-feedback that adds noise-like grit.
5. Operator LFO: sync 1/8, shape sine, destination: Osc B level and Osc C level (via the LFO's per-oscillator enable). Amount around 40%. The FM depth now wobbles, which shifts the sideband energy rhythmically.
6. Operator filter: LP 12 at 3 kHz with LFO also enabled on the filter, so cutoff and FM depth move together.
7. Pitch envelope on Osc C only (enable it in the pitch env per-osc): +12 semitones, decay 60 ms — a "yoi" attack on every note.

Wavetable Formant / vocal tables with LFO to position

Wavetable's Formant category holds tables that are already vowel-shaped. Scanning the position with an LFO turns the vowel; a filter on top shapes the "mouth".

1. Osc 1: Formant category, pick a table with clear vowel transitions (scroll with the position knob and listen for "ah–oh–ee"). Osc 2: Basics saw, Octave −1, 40% level for body. Sub oscillator on, level −10 dB.
2. Mod matrix: LFO 1 → Osc 1 Position, 60%. LFO 1 synced, 1/8, shape triangle. Add Env 2 → Position at −30% with a fast decay so each note starts on a different vowel than it sustains.
3. Filter 1: MS2 circuit, Band-pass, 1 kHz, resonance 30%. LFO 1 → Filter 1 Freq 20%. Routing: Serial. Filter 2: OSR LP 5 kHz, Drive 15%.
4. Unison: Classic, 2 voices, small amount, for a little width above the sub — Bass Mono handles the rest.
5. Roar Serial: Soft Sine 6 dB into Digital Clip 3 dB. The clip stage adds the "crunch" that makes a Wavetable growl sound like a brostep bass.

Roar feedback growls

Roar's Feedback routing sends the output back through the saturation stages via a tuned delay. On a bass with the feedback pitched to the key, this creates a self-sustaining resonance that growls on its own when modulated.

1. Roar: routing Feedback. Stage shaper Soft Sine, amount 10 dB. Pre-filter LP 4 kHz.
2. Feedback section: amount 40–50%. Set the feedback pitch (the Feedback section tunes the delay as a note) to the key's root or fifth. Feedback filter: BP around 800 Hz — this is the "mouth".
3. Roar's internal modulation matrix: LFO 1 (sync 1/8) → Feedback pitch, ±5 semitones. Envelope Follower → Feedback amount, negative, so loud attacks tame the feedback and the tail sings. Noise source → Feedback filter freq at a small amount for grain.
4. Keep a Limiter after it. Feedback above 60% can run away; ceiling at −3 dB catches it.
5. Automate Feedback amount per phrase: 30% for verses, 55% for the drop.

Meld Formant and Reed engines

Meld's macro oscillators each have engines built for this: Formant is a vowel source with two morphable formants, Reed is a physically-inspired blown reed whose pressure parameter snarls when pushed.

1. Meld Oscillator A: engine Formant. Its two main macro controls set the formant frequencies; put LFO 1 (synced 1/8, triangle) on the first at 50% and LFO 2 (1/8T) on the second at 30%. The two vowels now move against each other.
2. Oscillator B: engine Reed. Raise the pressure/breath-style control until the tone breaks into overtones; map Envelope 2 to it with a 150 ms decay so each note "spits" then settles.
3. Per-oscillator filters: A → Band-pass 1.2 kHz res 25%; B → LP 3 kHz. Both oscillators to Mono voice mode with Glide 20 ms.
4. Meld's Scrambler (12.2) on Osc B at a low amount for a slightly glitchy, "bitten" edge every few cycles.
5. Saturator Hard Curve after it — Meld's outputs are clean; the growl needs hard edges.

Resampling workflow: bounce → Simpler → re-modulate

Every serious growl goes through at least one resample pass. Once a growl is audio, you can pitch it, reverse it, slice it and modulate it again with a new set of tools, and the layers of movement stack.

1. Play the growl for 4 bars with several different modulation rates. On a new audio track, set the input to Resampling, arm, record. Or in 12.2+, select the clip and Bounce in Place for a pre-fader bounce.
2. Drag the recording into Simpler. Slice mode, Slice By Transient (or by Beat at 1/16 for a wobble). Now every slice is a MIDI note: rearrange the wobble rhythm by playing the slices.
3. After Simpler: Auto Filter BP with LFO at a different division than the original (if the source wobbled at 1/8, use 1/8T now). The two rates interfere — this is "complex" bass.
4. Shifter: Pitch −12 semitones, automate the Dry/Wet in the clip so alternate hits drop an octave. Or use Simpler's own pitch-per-slice.
5. Roar Multiband: leave the low band clean, Fold on the mid band, Soft Sine on the high band.
6. Record the result again. Keep every pass as a clip in Session view with the pass number in the name; you will want to go back.

Full method on the resampling workflow page.

Corpus and Resonators for metallic growls

Physical-model resonators after a growl add a ringing, pipe-like or plate-like body. Corpus tuned to the MIDI note gives a bass that sounds like it is being played through a metal tube.

1. Corpus: resonance type Tube (or Pipe for a more hollow sound). Enable the MIDI frequency option and set its sidechain MIDI input to the bass track so the resonator retunes per note. Transpose 0, Fine 0 for now.
2. Decay 30%, Material 60% (the brightness of the body), Dry/Wet 35%. Higher wet = more "instrument", less "synth".
3. Corpus's internal LFO on the tuning (small amount, 1/8 sync) makes the ring bend with the wobble.
4. Resonators after it, Mode B, 25% wet, resonators I–III tuned to root, +7, +12, decay short. This adds pitched sustain on the harmonics instead of a body.
5. EQ: notch out the strongest ring (usually 2–4 kHz) by a few dB; metallic resonances pile up there fast.

Vocoder talking bass: noise carrier, bass modulator (and the reverse)

Two uses. Classic: the bass is the carrier and a vocal (or a vowel sample) is the modulator, so the bass speaks. Inverted: a noise carrier modulated by the bass gives an airy "breath" layer whose rhythm follows the wobble exactly.

1. Routing matters: Live's Vocoder sits on the modulator track (the signal whose spectral envelope is analysed) and picks the carrier from its Carrier menu. For a talking bass the bass is the carrier, so put the Vocoder on a track playing a vowel loop ("ah-oh-ee-oh" spoken in time, or a formant pad) and set Carrier to External, source = the growl bass track (pre-FX or post-FX). Mute the bass track's direct output or keep it in for a blend.
2. Bands 24–40, Range 100 Hz – 8 kHz, Formant shift +2 to +4 semitones to make the vowel brighter than the human voice.
3. Envelope: Attack 5 ms, Release 60 ms for crisp syllables; raise Release to 200 ms for smeared, slurred growls. Unvoiced 10% adds a little noise on consonants.
4. For the inverted "breath" layer: Vocoder on a Noise carrier (Carrier: Noise), modulator = the bass via sidechain. 12 bands, Depth high. Mix at −18 dB as a top layer above 3 kHz.

Auto Shift on bass: pitch snaps and vibrato growls

Auto Shift is a pitch corrector, but on a resampled growl it is a scale-locked pitch modulator: its internal LFO and vibrato bend the bass in time, and its correction quantizes slides into the key.

1. Auto Shift after a resampled growl in Simpler (monophonic source only — it cannot track chords).
2. Scale: set to the song key (it follows Live 12's set scale if enabled). Correction speed fast so slides snap into notes.
3. LFO: square shape, synced 1/8, depth ±7 semitones, so the bass alternates root/fifth every eighth — a wobble in pitch rather than filter. Try ±12 for octave hops, or a saw shape for rising "yoi" stabs.
4. Vibrato: 0 for riddim; for a tearout wail, 6 Hz vibrato at ±1 semitone with a 200 ms onset delay.
5. Sidechain input on Auto Shift: route the kick so the pitch dips on each hit (set the sidechain to modulate pitch downwards) — a pitch-based "sidechain".

Spectral Resonator "frozen vowel" growl

Spectral Resonator resonates the incoming spectrum at a pitch set by MIDI. On a growl it adds a sustained harmonic halo that follows the bassline; with Freeze automated it holds a single vowel still while the source keeps wobbling under it.

1. Spectral Resonator in MIDI mode, sidechain MIDI from the bass track, so the resonance pitch follows the notes. Polyphony 1 for a tight monophonic bass.
2. Decay 300–500 ms (long enough to bridge the LFO troughs), Harmonics 60% (how many partials resonate), Stretch 0 for harmonic, small positive values for a detuned/metallic edge.
3. Dry/Wet 40%. The dry growl stays rhythmic, the resonated part sings over it.
4. Automate Freeze on for the last beat of every 4 bars: the resonated spectrum holds while the dry signal stops — a built-in "bass tail" riser.
5. Unison/Shift controls inside Spectral Resonator give the halo a chorused width; the HP at 120 Hz afterward keeps that width out of the low end.

"Yoi" tearout stabs and riddim minimal wobbles

Two opposite aesthetics with the same tools. The "yoi" is a fast upward pitch and formant sweep on note attack — a bark. Riddim is the opposite: one short, dull wobble repeated with machine precision and almost no variation.

1. Yoi: Operator with a saw carrier and one sine modulator at ratio 2. Pitch envelope: +19 semitones (an octave and a fifth) with Decay 80 ms, applied to the modulator only (per-oscillator pitch env enable). Filter envelope: Decay 120 ms, amount 60%, so cutoff and pitch fall together — the "yoi" syllable.
2. Roar Serial: Hard 8 dB then Diode 6 dB. The hard clip emphasises the attack transient.
3. EQ Eight bell +4 dB at 900 Hz, Q 1.5: the "nasal" presence tearout bass is known for.
4. Riddim minimal: Wavetable Basics saw, Auto Filter LP 12 with LFO square-ish (morph 70% toward square), 1/8T, amount 50%, resonance 20%. No pitch env, no reverb, no width. Notes on every 1/8, same pitch for 8 bars, one octave drop on bar 8. Compression (Glue 4:1, fast) to make it stiff.
5. The riddim trick is subtraction: EQ Eight LP at 4 kHz on the mid layer. Riddim bass is dark.

Chorus-Ensemble width and multiband processing

Width above 200 Hz makes the growl sound huge on headphones; multiband dynamics keep the mid layer loud and consistent while the sub stays untouched.

1. Multiband Dynamics: crossover at 120 Hz and 3 kHz. Low band: no processing (or solo it to confirm nothing is there if the sub is on another track). Mid band: downward compression above −12 dB at 3:1 plus upward compression below −30 dB — the growl stays dense through the LFO troughs. High band: 2:1 gentle.
2. Chorus-Ensemble in Ensemble mode: 3 voices, Rate 0.5 Hz, Amount 30%, Dry/Wet 35%. The "Classic" mode is narrower; "Vibrato" mode with a small amount gives a pitch shimmer without width.
3. EQ Eight in M/S mode: Side channel high-passed at 200 Hz, so the Chorus width exists only above the body.
4. Utility Width 120% max. Check mono.

Layering three ways: sub, mid, top

The production-standard split. Every 140 BPM bass is three chains driven by the same MIDI; only the mid is modulated, only the top is wide, only the sub is loud below 100 Hz.

1. Sub chain: Operator sine, Amp A 3 ms, R 60 ms, Mono voice mode, glide matched to mid. EQ LP 110 Hz. Nothing else. −3 dB.
2. Mid chain: the growl recipe of your choice, HP 100 Hz, all modulation here. 0 dB.
3. Top chain: a copy of the mid growl or a noise/Wavetable square layer, HP 2.5 kHz, Chorus-Ensemble and Width 150%, Saturator Digital Clip for fizz. −10 dB. This is what makes the growl audible on a phone speaker.
4. Rack macros: Mid LFO rate, Mid filter res, Top level, Dirt. Map to Push for live play.
5. Glue on the rack output ties the layers; the sidechain compressor ducks the sub and mid for the kick (6 dB, Release 80 ms at 140).