← LearnSound Design
Sound DesignSerum 2Synthesis
Sound Design
with Serum 2
Wavetable synthesis is why so much electronic music sounds the way it does right now. Serum 2 is the go-to tool, not because it's the most powerful synth ever built, but because it's fast to patch, the modulation is drag-and-drop, and the presets are actually usable. This covers the whole thing: signal flow, each parameter, and step-by-step recipes for real genre sounds.
■ orange — note■ green — tip■ yellow — heads up
Why wavetable, and why it dominates
A wavetable is a stack of single-cycle waveforms (typically 256 to 2048 frames, each a snapshot of a different timbre). WT POS scrubs through every frame, morphing the timbre as you go. Modulate it with an LFO and the sound evolves constantly without touching the filter, which is most of what makes wavetable synthesis worth using.
Compared to subtractive, you're not locked to saw/square/sine: you can use any timbre you can capture in a single cycle. Serum also has FM built in via OSC B, so you can layer both approaches.
What Serum 2 adds over Serum 1
| Area | What Changed in Serum 2 |
|---|---|
| Oscillator Warp | New warp modes including spectral warping, additional window sync types, and improved FM/AM fidelity at high modulation depths. |
| Filter Bank | Expanded filter types including new multi-mode morphing filter, updated Dirty models with per-pole saturation, and a dedicated Filter FX slot in the effects rack. |
| FX Rack | Expanded from 8 to 10 FX slots. New effects types added, improved Reverb with plate/room/hall distinction, better Delay tape emulation. |
| Modulation | Matrix view improved with per-mod curve shapes. LFO has additional playback modes including triggered step-sequencer and envelope-follower mode. |
| Wavetable Import | Drag-and-drop any audio file to create a wavetable. Improved spectral editing inside the WT editor for sculpting frames manually. |
| Unison Engine | Up to 16 voices per oscillator (up from 8 in Serum 1). Improved stereo spread algorithm reduces low-frequency phase cancellation. |
| Voice Architecture | Added per-voice random seed for more organic unison spread. MPE (MIDI Polyphonic Expression) support for per-note pitch bend and pressure. |
The signal flow
Oscillators → filter → amp → FX. Know the path and you always know which stage to fix. Click a stage below for parameters and tips.
Signal Flow — click a stage to explore
→
→
→
→
→
Oscillators
OSC A + B
The raw sound source. Each oscillator reads through a wavetable — a looped single-cycle waveform table — and outputs audio at the pitch set by your MIDI note. OSC A is the primary voice; OSC B is identical in architecture and can be pitched, detuned, or phase-shifted relative to A. Both support up to 16-voice unison, which layers slightly detuned copies of the oscillator across the stereo field. Wavetable position (WT POS) sweeps between frames in the table, morphing the timbre without envelopes or filters.
Key Parameters
WT POSSweeps through wavetable frames. Automating or modulating this is the core of moving, evolving timbres.
UNISON1–16 voices. Each voice is a separate oscillator detuned slightly from centre. Higher counts = wider, thicker sound.
DETUNEControls spread between unison voices in semitones (0–2 range). 0.10–0.20 is the supersaw zone; above 0.30 creates noticeable beating artefacts.
BLENDMix between the centre voice and the outer unison voices. Low blend = tighter, more mono-compatible. High blend = wider but thinner centre.
WARPWaveshaping applied inside the oscillator before it exits. Modes include Sync, Bend+, FM (from B), Window Sync, Asym, Remap, and more.
PHASE / RANDStarting phase of the oscillator on each note-on. RAND adds randomness to reduce phase-correlation between notes.
SEMI / FINECoarse and fine pitch of each oscillator. Offset OSC B by +12 for an octave layer, or +7 for a fifth.
Sound Design Principle
For a supersaw lead, use 7 unison voices on OSC A, Detune at 0.18, Blend at 0.60–0.70. Stereo Spread at 0.35–0.50 keeps it wide without causing mono phase cancellation below 200 Hz. Keep OSC B off — two supersaws fight each other more than they thicken.
Tip
Troubleshoot top-down. Too bright? Start with the filter before touching the oscillator. Too thin? Check oscillator unison before reaching for FX. Too wet? Dial back reverb or delay mix before adjusting oscillator level. Each stage has a clear job, so working in the right place saves a lot of time.
Heads up
Stacking FX slots without checking what each one does is the fastest way to make a patch sound cloudy and CPU-heavy. Bypass each slot individually. If you can't hear a difference, delete it. Less is cleaner.
Oscillators
OSC A is your main voice. OSC B layers on top, detuned, shifted in timbre, or at a different octave. Switch OSC A's warp to FM or AM "from B" and OSC B becomes a modulator instead of a mixed voice.
WT POS is your timbre knob. A filter cutoff sweep only removes harmonics, but WT POS can add them: routing an LFO into WT POS at 0.2–0.5 Hz gives pads a kind of movement a filter sweep can't quite replicate.
Warp modes
| Warp Mode | What it does |
|---|---|
| Sync | Hard-syncs a hidden oscillator to OSC A's reset. Creates the classic sync sweep sound — harsh, bright, with a sweepable formant peak. The WARP knob controls the sync ratio. |
| Bend+ | Warps the wavetable playback time, creating asymmetric waveforms with enhanced even harmonics. Subtle WARP values add weight; extreme values create hard-edge distortion. |
| Bend− | Same as Bend+ but bends in the other direction. Emphasises odd harmonics. Alternating polarity between OSC A and B adds complexity. |
| Window | Applies a window function to each cycle, reducing the edges. Lower-aliasing alternative to hard sync. Useful for smooth, breath-y leads. |
| Asym | Asymmetric waveshaping — applies a different gain curve to positive and negative cycles. Adds second-harmonic (even) content, simulating tube saturation. |
| FM (from B) | OSC B's output frequency-modulates OSC A. The WARP knob controls FM depth. Classic FM bell, metal, and DX7-style timbres — but with full wavetable source flexibility. |
| AM (from B) | OSC B amplitude-modulates OSC A. Creates sum and difference sidebands (ring modulation family). Metallic, bell-like when tuned harmonically; dissonant when not. |
| Remap | Applies a user-drawn curve to the waveform. Essentially arbitrary waveshaping. Good for adding harmonics in a musically controllable way. |
| Flip | Flips the top or bottom half of the waveform. Creates hard transitions and additional harmonic content at extreme settings. |
Tip
Hard sync lead: Warp mode = Sync, WARP knob at 30–50%, then modulate that WARP knob with ENV 2 sweeping high to low. The formant dives down across the attack of every note. That's the sound of Robert Miles' "Children" and basically every 2000s trance lead.
Unison
Stacks detuned copies of the oscillator. Up to 16 voices per oscillator.
| Parameter | Effect |
|---|---|
| UNISON (count) | 1 voice = monophonic oscillator. 3–5 voices = subtle thickening. 7–9 voices = classic supersaw. 12–16 voices = extreme width and wash, expensive on CPU. |
| DETUNE | Pitch spread across all voices in semitones. 0.05–0.10 = tight chorus. 0.15–0.25 = supersaw zone. 0.30+ = audible beating and detuned character. Above 0.50 = intentionally out-of-tune. |
| BLEND | Mix between centre voice (0%) and full unison spread (100%). Low blend keeps a focused centre with wide spread. High blend is lush but loses punch. |
| STEREO | Panning spread of the unison voices across the stereo field. 0% = all voices centre. 100% = outer voices at full L/R. Use 30–50% to avoid over-widening that collapses in mono. |
Heads up
7 voices at 0.18 detune is the mono-safe supersaw. Outer voices are ~10 cents off centre, so cancellation below 200 Hz stays minimal. Push to 0.30+ and low-mid energy disappears the moment your mix sums to mono.
Tip
OSC B as FM modulator: set OSC A warp to FM (from B), dial in WARP knob depth, then set OSC B level to 0. It shapes OSC A's frequency invisibly. Classic DX7-style timbres with full wavetable flexibility.
Filter
Most of the movement in house, techno, and DnB bass comes from the filter envelope sweeping the cutoff over time. ENV AMT is usually the most important knob to start with, not the cutoff position itself.
Filter types
| Type | Character and use |
|---|---|
| MG Low 24 | Moog ladder filter emulation, 24 dB/octave low-pass. The warmest, most musical low-pass in the bank. Self-oscillates cleanly at resonance > 0.85. Default choice for bass and warm leads. |
| MG Low 12 | 12 dB/octave Moog ladder. Less steep slope, more midrange content passes through. Useful for bass sounds you want to retain upper harmonics in. |
| Dirty Low 24 | MG Low 24 with added harmonic saturation at the filter stage. Each pole introduces subtle distortion. Sounds 'analog' and gritty without explicit distortion. Excellent for reese bass. |
| SV Low / High / Band | State-variable filter with clean, precise response. 12 dB/octave. More neutral than MG types. Band-pass mode useful for telephone/radio formant effects and mid-focused timbres. |
| Comb+ | Comb filter producing a series of resonant peaks spaced at the comb period. The CUTOFF knob controls comb spacing. RESONANCE controls peak sharpness. Metallic, tubular bell, or plucked string character. |
| Flanking | A through-zero flanger algorithm baked into the filter stage. The cutoff controls the flanger delay time and the resonance controls feedback depth. |
| Formant | Two resonant bandpass peaks tuned to vowel formant positions. CUTOFF sweeps between vowel shapes (A → E → I → O → U). Classic for talking bass effects. |
Filter Envelope Shape — Typical Bass
ENV 2 (Filter Envelope): Attack: 0 ms — opens instantly on note-on Decay: 300 ms — sweeps down from peak cutoff Sustain: 25% — settles at a partially open position Release: 150 ms — closes as note ends Filter: MG Low 24 Cutoff: 35% — base position ENV AMT: +65% — peak cutoff = 35 + 65 = ~100% at attack RES: 0.30 — slight resonance adds presence at peak DRIVE: 20% — harmonic warmth throughout
Tip
Low base cutoff + high ENV AMT = blows open on attack, closes as it decays. Classic electronic bass wah. Flip ENV AMT negative and it starts open and snaps shut: tight pluck character.
Heads up
Resonance above 0.70 at high cutoff can trigger self-oscillation, an audible pitch at the cutoff frequency. Back off resonance if ENV AMT is sweeping high.
Drive
DRIVE saturates before the filter, so distortion harmonics get shaped along with everything else. 10–25% adds warmth. 60%+ is acid techno: crunchy input that the filter then sculpts. Very different from putting distortion after the filter.
Modulation
Drag-to-assign: drag from any source to a destination knob. The mod amount shows as a ring around the target.
Envelopes
ENV 1 is hardwired to amplitude. ENV 2 defaults to filter cutoff but routes anywhere. Need more? Add sources in the mod matrix.
| Stage | What it actually does to your sound |
|---|---|
| Attack (A) | Time from note-on to peak level. 0 ms = instant transient. 500 ms = slow pad swell. Attack on ENV 1 shapes the volume opening; attack on ENV 2 delays how long before the filter opens fully. |
| Decay (D) | Time to fall from peak to the sustain level. On a pluck (sustain = 0), decay is the entire length of the note — the only thing you hear. 200–400 ms = snappy pluck. 600–900 ms = longer pluck. |
| Sustain (S) | The level the envelope holds at while the note is pressed. 0% = percussion/pluck. 50% = partial hold. 100% = full hold (leads, pads, sustained bass notes). |
| Release (R) | Time to fall from sustain level to zero after key-off. Long release = reverberant, ambient trail. Short release = tight, separated notes. Always set release ≥ your reverb pre-delay. |
Tip
Techno/house bass: 50–150 ms release or notes smear at 130 BPM. Ambient pads: 2–4 s. Quick test: play fast 8th notes. Blurring? Shorten release. Chopped? Lengthen it.
LFOs
Four LFOs, all routable. Into WT POS: evolving timbre. Into filter cutoff: auto-wah. Into pan: auto-panner.
| LFO Parameter | What to know |
|---|---|
| RATE | In Hz (free) or note divisions (BPM-synced). 1/4 = one LFO cycle per quarter note. 1/8 = per eighth note. Synced LFOs lock movement to your DAW tempo. |
| SHAPE | The waveform of the LFO. Sine = smooth continuous movement. Triangle = linear ramp. Square = abrupt switching between two states. Random = stepped random (sample & hold). User-drawn shapes available. |
| TRIGGER | How the LFO resets on note-on. 'Free' = continuous, notes don't reset it. 'Trigger' = resets to phase 0 on each note-on. 'Env' = LFO runs once like an envelope (one shot). Envelope mode converts the LFO into a free-form envelope. |
| DELAY / ATTACK | A fade-in time before the LFO reaches full depth. Useful for vibrato — start with no wobble, introduce it gradually after 200–500 ms, mimicking a human player's technique. |
| AMT | Modulation depth — how much the LFO moves the target parameter. Dragging to assign sets the initial amount; you can trim it after assignment. |
Macros
Four knobs, each assignable to any number of parameters with individual amounts and directions. Map to DAW automation or MIDI CCs.
Tip
Think of each macro as a "character" axis: Macro 1 = Brightness (filter + WT POS), Macro 2 = Thickness (unison blend + sub level), Macro 3 = Movement (LFO rate + depth), Macro 4 = Space (reverb wet + delay wet). Four knobs sitting on top of a deep patch is usually all the performance control you actually need.
Other modulation sources
| Source | Musical use |
|---|---|
| Velocity | Harder notes open the filter more (velocity → ENV AMT or filter cutoff). Adds dynamics to otherwise static pads and leads. |
| Note (pitch) | Modulate filter cutoff with note pitch at low depth — keeps the filter proportional to the note frequency, preserving tonal consistency across octaves. |
| Aftertouch / MPE | Channel aftertouch or per-note MPE pressure modulates vibrato, filter, or LFO depth in real time during held notes. Requires a pressure-sensitive controller. |
| Modwheel | Classic performance controller. Assign to vibrato depth (LFO 1 → pitch) or filter opening. The standard go-to for live expressiveness. |
| Random | A new random value per note. Randomize WT POS slightly for organic variance between voices. Randomize filter cutoff slightly for unpredictability. |
FX chain
Ten slots in series, each with wet/dry. Distortion before reverb = the whole tail is saturated. Distortion after = clean reverb, gritty dry signal. General rule: distortion first, modulation effects second, reverb and delay last.
| Effect | What it adds, practical use |
|---|---|
| Hyper/Dimension | Hyper mode adds a second unison layer at the FX level — width without oscillator CPU cost. Dimension emulates the Roland Dimension D, adding a subtle chorus+spread that doesn't add flutter. Both are excellent for pads. Use Dimension at 70–100% wet for transparent width; Hyper at 30–50% for more obvious effect. |
| Distortion | Multiple modes with distinct character. Soft Clip = tube-like warmth at low drive. Hard Clip = digital edge and presence. Warm = low-frequency saturation bias. Shred = extreme harmonic generation. Down Sample = lo-fi aliasing. Use at 5–20% drive for presence; 40%+ for intentional grit. |
| Flanger | Modulated comb filter. Short delay time (0.5–3 ms) with feedback. Creates the classic flanging sweep. Good on pads and arpeggios but heavy on leads — use sparingly. |
| Phaser | All-pass filter chain producing a frequency-dependent phase shift. Subtler than flanging. At low depth and slow rate it adds movement without obviously modulating. At high depth it sweeps noticeably. |
| Chorus | Modulated short delay for width and shimmer. Classic use: light chorus on a mono lead to widen it. Depth 30–50%, Rate 0.3–0.8 Hz, Width 50–70%. |
| EQ | 4-band parametric. Inside Serum, primarily used to high-pass below 60–80 Hz (prevents sub mud on non-bass sounds) and high-shelf cut above 16 kHz (tames harshness on bright supersaws). |
| Compressor | Dynamic range control. Used post-distortion to tame peaks and after Hyper to even out the multi-voice level. Avoid heavy ratios inside the synth — that's your DAW channel strip's job. |
| Reverb | Plate, room, and hall modes. Pre-delay (10–30 ms) separates the dry sound from the reverb onset, preserving punch. Size = room scale. Decay = reverb tail length. Use 15–25% wet inside the synth and add more on the DAW return channel. |
| Delay | Stereo and ping-pong delay with tempo-sync. Feedback controls repeat count. Use 1/8 or 1/4 note sync for rhythmically locked echoes. Keep mix at 10–20% inside the synth for space without washing. |
| Filter FX | An additional filter stage at the end of the chain. Useful for post-reverb high-passing or shaping the overall spectral content of the wet signal. |
Tip
The most versatile FX chain template for electronic music production: slot 1 = Distortion (Soft Clip, drive 8–15%) for harmonic warmth, slot 2 = Hyper/Dimension (Dimension mode, 60% wet) for stereo width, slot 3 = EQ (high-pass at 70 Hz), slot 4 = Reverb (Plate, pre-delay 18 ms, decay 1.2 s, mix 18–22%), slot 5 = Delay (1/8 note, feedback 25%, mix 12%). Adjust per sound: a bass probably only needs slot 1; a pad needs all five.
Heads up
Hyper generates additional voices at the FX level, which gets expensive fast on polyphonic patches. If your CPU is spiking, check Hyper first. Oscillator unison gets you similar width for less cost, so worth trying before defaulting to Hyper.
Recipes
Build these from scratch. Once working, break them: push DETUNE, throw an LFO at WT POS, swap the filter type.
Recipe 1 — Trance Supersaw Lead
No filter processing. All thickness from unison and stereo spread.
OSC A: Wavetable: Analog > Saw Analog (or Basic Shapes > Saw) WT POS: 0% (pure saw — no morphing needed) Unison: 7 voices Detune: 0.18 Blend: 0.65 Stereo: 0.40 Octave: 0 Semitone: 0 Fine: 0 OSC B: OFF Sub: OFF Noise: OFF Filter: OFF (supersaw is designed to run filter-free) ENV 1 (Amp): Attack: 8 ms (just enough to avoid click) Decay: 200 ms (slight peak smoothing) Sustain: 100% Release: 400 ms (medium tail) ENV 2: not used (filter off) LFO 1: Shape: Sine Rate: 0.25 Hz (free, very slow) Target: WT POS (+12%) — barely perceptible drift for life Macros: Macro 1 → Detune (+0.12 range): opens to 0.30 for extra width Macro 2 → Reverb mix (+20%): performance control FX: Slot 1: EQ — HPF at 60 Hz (clean sub), HF shelf -2 dB at 12 kHz Slot 2: Chorus — Depth 0.35, Rate 0.5 Hz, Width 60% Slot 3: Reverb — Plate, Pre-delay 20 ms, Decay 1.5 s, Mix 20% Slot 4: Delay — 1/8 note, Feedback 20%, Mix 12% Master: Poly, Portamento OFF, VOL -3 dB
Note
7 voices: odd count keeps stereo spread symmetric and holds the mono centre. Even counts (6, 8) can introduce subtle mono imbalance. 7 is the sweet spot for supersaw density at 0.18 detune.
Recipe 2 — Drum & Bass Reese Bass
The growl is beating interference between two detuned saws. OSC B at −14 cents is the whole trick.
OSC A: Wavetable: Analog > Saw Analog WT POS: 0% Unison: 1 (mono — no unison on Reese) Detune: 0 Fine: 0 OSC B: Wavetable: Analog > Saw Analog (same table as A) WT POS: 0% Unison: 1 Semitone: 0 Fine: -14 cents (core detuning — the growl source) Level: same as OSC A (~0 dB relative) Sub: Wave: Sine Level: 25% (reinforces fundamental below 80 Hz) Noise: OFF Filter: Dirty Low 24 Cutoff: 45% Resonance: 0.35 Drive: 35% (pre-filter saturation is essential for Reese) ENV AMT: +40% ENV 1 (Amp): Attack: 0 ms Decay: 10 ms (fast click, then full) Sustain: 100% Release: 80 ms (tight — DnB patterns are fast) ENV 2 (Filter): Attack: 0 ms Decay: 400 ms Sustain: 30% Release: 150 ms LFO 1: Shape: Sine Rate: synced 1/1 (one LFO cycle per bar) Target: OSC B Fine (+8 cents range) — slow beating drift Trigger: Free FX: Slot 1: Distortion — Warm mode, Drive 20%, Mix 60% Slot 2: EQ — HPF 40 Hz, cut -3 dB at 600 Hz (clean mids) Master: MONO, Portamento 40 ms, Poly = 1 voice
Tip
Lower notes beat slower, which is why Reese basses feel heavier as you go down the keyboard. Keep it around −14 cents; wider makes it buzz, not growl.
Recipe 3 — Tech House / Minimal Pluck
No sustain: just attack and decay. The filter envelope does all the tonal work in that decay window.
OSC A: Wavetable: Analog > Saw Analog (or PWM wavetable at WT POS ~40%) Unison: 3 voices Detune: 0.08 (tight — barely any spread) Blend: 0.50 Stereo: 0.20 (mono-leaning) OSC B: OFF Sub: OFF Noise: Type: White Pitch: +20 (rolls off low noise, keeps it bright) Level: 15% (adds click/transient body) — route Noise through its own short ENV → Noise Env: Attack 0, Decay 60 ms, Sustain 0 Filter: MG Low 24 Cutoff: 30% Resonance: 0.45 Drive: 15% ENV AMT: +75% ENV 1 (Amp): Attack: 0 ms Decay: 350 ms (the note length) Sustain: 0% (100% pluck — no hold) Release: 180 ms ENV 2 (Filter): Attack: 0 ms Decay: 280 ms (slightly shorter than amp decay) Sustain: 0% Release: 120 ms LFO 1: not modulating (static pluck is intentional) FX: Slot 1: Distortion — Soft Clip, Drive 12%, Mix 55% Slot 2: EQ — HPF 80 Hz, narrow cut -4 dB at 1.2 kHz (boxiness) Slot 3: Reverb — Room, Pre-delay 12 ms, Decay 0.8 s, Mix 15% Master: Poly (up to 4 voices), Portamento OFF, VOL -2 dB
Tip
The noise is what makes it feel physical. White noise at 10–20%, 40–80 ms decay. Mimics the attack texture of a real plucked string.
Recipe 4 — Evolving Pad
All motion from WT POS modulation and a slow filter LFO. Continuous timbral drift with no obvious arc.
OSC A: Wavetable: A spectral or complex wavetable (e.g. "Vocal") WT POS: 50% (middle of the table) Unison: 8 voices Detune: 0.22 Blend: 0.55 Stereo: 0.50 OSC B: Wavetable: Different timbre (e.g. "Strings" or "Pads") WT POS: 20% Unison: 1 Semitone: +12 (octave up — adds shimmer) Level: -6 dB relative to OSC A Sub: Wave: Sine Level: 20% Filter: SV Low 12 (cleaner, less coloured for pads) Cutoff: 75% Res: 0.10 ENV AMT: +15% ENV 1 (Amp): Attack: 900 ms Decay: 500 ms Sustain: 85% Release: 2.5 s ENV 2 (Filter): Attack: 800 ms Decay: 300 ms Sustain: 80% Release: 2 s LFO 1: Shape: Sine Rate: 0.18 Hz (free, very slow) Target: WT POS OSC A (+30% depth) LFO 2: Shape: Sine Rate: 0.07 Hz (even slower) Target: WT POS OSC B (+20% depth) — offset drift LFO 3: Shape: Triangle Rate: 0.30 Hz Target: Filter Cutoff (+8%) — gentle breathing FX: Slot 1: Hyper/Dimension — Dimension mode, Mix 75% Slot 2: Chorus — Depth 0.55, Rate 0.4 Hz, Width 80% Slot 3: Reverb — Hall, Pre-delay 25 ms, Decay 3.0 s, Mix 35% Slot 4: Delay — 1/4 note dotted, Feedback 30%, Mix 18% Slot 5: EQ — HPF 50 Hz Master: Poly (8 voices), Portamento 0, VOL -4 dB
Quick reference cheat sheet
Supersaw detune values
| 0.05–0.08 | Subtle chorus, mono-safe |
| 0.12–0.18 | Classic supersaw, clean mono |
| 0.22–0.28 | Wide, slight mono loss at low end |
| 0.30–0.40 | Extreme, audible beating, thin mono |
Envelope shapes by sound type
| Pluck | A: 0 ms D: 200–500 ms S: 0% R: 150 ms |
| Lead | A: 5 ms D: 100 ms S: 90% R: 200–400 ms |
| Pad | A: 400–1200 ms S: 80% R: 1–3 s |
| Bass | A: 0 ms D: 10 ms S: 100% R: 60–120 ms |
| Stab | A: 0 ms D: 80 ms S: 0% R: 80 ms |
Filter type quick pick
| MG Low 24 | Warm bass, classic lead, pad |
| Dirty Low 24 | Reese bass, gritty analogue |
| SV Band | Formant/vowel, telephone effect |
| Comb+ | Bell, plucked string, metallic |
| Formant | Talking bass, vocal effect |
| SV Low 12 | Clean pad, subtle filtering |
FX chain order rules
| Distortion | Early — shapes the source timbre |
| Hyper/Dimension | After dist — width pre-reverb |
| EQ | After dist — shape before space |
| Chorus/Flanger | Mid-chain — on already-shaped sound |
| Reverb | Late — places sound in space |
| Delay | Last — temporal echoes of the full sound |
Key warp mode use cases
| Sync | Trance sweep, formant lead |
| FM (from B) | Bell, DX7 electric piano |
| AM (from B) | Metallic, ring-mod bass |
| Bend+ | Fat harmonics, warm deform |
| Remap | Custom harmonic sculpting |
| Window | Low-alias sync alternative |
Mono vs Poly — when to use which
Mono: basses, monophonic leads, anything that should only play one note at a time. Add Legato mode for glide between connected notes without re-triggering the envelope.
Poly: pads, chord stabs, plucks played in parallel. Set voice count to match your max chord size (4–8 is typical).
Portamento in Mono mode: 0 ms for tight DnB bass. 30–80 ms for sliding reese. 100–200 ms for expressive lead. Only audible between notes in mono/legato mode.
← All ArticlesParameter values are starting points. Trust your ears over any number.