Voice & play

The voicing section decides how notes become sounding voices: up to 16-voice polyphony, five play modes (Poly / Duo / Mono / Legato / Latch) with note priority and glide, unison stacking (count / detune / spread), and the analog-character controls (Voice Detune, Drift, Variance, and Transient) that loosen voices away from perfect digital precision for an “alive” sound.

Polyphony

The Voices control sets the polyphony cap from 1 to 16. When every allotted voice is busy, a new note steals the oldest. The effective cap follows the play mode and unison count; mono modes and heavy unison stacks fold down to fewer sounding notes automatically.

Play modes

Five modes decide how held keys become sounding voices:

Mode	What it does
Poly	Full polyphony: every key sounds its own voice
Duo	True two-note duophonic: the two priority notes of what you hold sound. A third note pushes one out; releasing one of the pair while a third is held brings that third note back (a note reservoir). Note Priority (Last/Highest/Lowest) picks the two winners; glide and unison apply per note
Mono	A single sounding voice; each new note retriggers the envelopes
Legato	A single voice that glides to each new note without retriggering, so envelopes keep running
Latch	A single held note that new notes replace, and that keeps sounding after you lift the key

With the arpeggiator on, Latch becomes arp hold: the chord keeps cycling after you release the keys, and a fresh press with no keys down replaces the latched chord.

Note priority

In mono modes, the sounding note is chosen by priority:

• Last: the most recently pressed key.
• Highest: the highest held key.
• Lowest: the lowest held key.

Glide

Glide slides the pitch from the previous note to the new one over a settable time. Two modes:

• Always: glide happens on every note, regardless of whether keys overlap.
• Fingered: glide only happens between overlapping notes. A fresh note after all keys are released snaps cleanly to pitch, so you can shape legato passages without glide following you everywhere.

A whole chord glides from the previous note cleanly, without each voice smearing its own start.

For slide steps in the step sequencer, Glide has to be non-zero for the slide to actually bend pitch. See the step sequencer for details.

Unison

Unison stacks up to eight voices per note. Two controls shape the stack:

• Detune: spreads the stacked voices in pitch, on top of any base detune. Because each stacked voice carries its own stable offset, the stack fattens into a wide, slowly beating cloud rather than a chorus that pulses in lockstep.
• Spread: pans the stacked voices across the stereo field.

At Unison 1 there is no stacking; Detune and Spread have no effect. Turn Unison up and Detune up together for the classic “megasaw” texture; add Spread for width.

Analog character: the aliveness layer

Four controls loosen each voice away from perfectly matched digital precision, modeling the way separate analog voice cards each have their own personality. At zero every voice collapses to identical. Higher settings model older, looser hardware.

• Voice Detune: a fixed per-voice pitch spread across the stack. Even a small amount makes a chord shimmer and feel live rather than static.
• Drift: slow, quasi-random per-voice pitch wander that keeps the tone moving over time.
• Variance: one knob that adds per-voice slop to cutoff, envelope times, pulse width, and glide time simultaneously. Push it up and the synth starts to feel like individual analog voice cards that each have their own character.
• Transient: two settings. Analog leaves oscillator phase and filter state running when a voice is stolen and reused, for a punchy, varying attack. Reset zeroes them at each note-on for a consistent, clicky attack. Transient governs the DSP state reset only; it is separate from Env Restart, which controls the envelope’s starting level when a new note arrives.

The analog-character controls are what the philosophy chapter describes as “aliveness.” See The Aconite philosophy for the broader framing of why these exist.

Velocity curve

The velocity curve is a drawable map that reshapes how hard you play into the velocity the synth actually uses. It sits in the performance area alongside the pitch and mod wheels.

The horizontal axis is the velocity you play; the vertical axis is the velocity Aconite receives. The default is a straight diagonal, a 1:1 pass-through. Bending the line away from the diagonal changes the response:

• Soften: flatten the upper end so hard playing lands gentler, for a more forgiving touch.
• Harden: steepen it so subtle differences in strike force become dramatic dynamic swings.
• Compress: pull the whole range into a narrow band, so loud and soft notes sit close together.
• Expand: push the extremes apart, widening the gap between your softest and hardest playing.
• Fix: a flat horizontal line for one constant velocity regardless of how hard you play.
• Reverse: slope it downward so soft playing is loud and hard playing is soft.
• Freehand: draw any contour; the curve follows exactly what you draw.

Because the curve reshapes velocity before anything reads it, it moves everything velocity drives at once: note loudness and every velocity-based route in the modulation matrix respond to the same drawn shape. It also feeds the arpeggiator and step sequencer, so their accents ride the reshaped velocity too.

A live indicator dot shows where your last played note landed on the curve, so you can see your touch mapping as you play. A reset returns the curve to the straight linear default at any time.

Preset modes

A small box in the lower-right corner of the curve plot shows a mini thumbnail of the current shape. Click it to cycle through a set of common preset modes, loading each one straight into the curve:

• Linear — a straight 1:1 response, the same as you played.
• Soft — a soft touch still plays loud, for a gentler, more forgiving feel.
• Hard — it takes a firm hit to get loud, so you have to dig in for the top end.
• S-Curve — gentle at both ends and steep through the middle, easing your softest and loudest notes.
• Fixed — a constant output regardless of how hard you play.

Each click advances Linear → Soft → Hard → S-Curve → Fixed and then wraps back to the start. Draw your own contour by hand at any time and the box switches to Custom. Whichever mode you land on saves with the patch, just like a hand-drawn curve.

Pitch-bend range

The pitch-bend range controls how many semitones the pitch wheel reaches at full throw. The standard is two semitones; wider ranges suit expressive lead playing or trombone-style glides.

When MPE is active, each voice tracks per-note bend from its own MIDI channel, and the per-note pitch-bend range is adjustable from 1 to 96 semitones, 48 semitones by default, to suit expressive controllers. Per-note pressure and vertical slide are exposed to the modulation matrix as the sources MPE Press and MPE Slide, so you can route them to any destination (filter cutoff, timbre, volume, or anything else in the matrix) for fully independent expression on every finger.

Panning

Two pan controls are available:

• Voice Pan: places the overall voice output in the stereo field. Because this follows the modulation matrix Pan destination, an LFO or other source can auto-pan on top of the static setting.
• Scene Pan: places a whole scene in the field when both scenes are live under Layer or Split routing.

Sequencer: Shared / Per-Scene

The SEQ: Shared / Per-Scene control chooses whether the two scenes share one sequencer, arp, chord lane, and probability filter, or whether Scene B runs its own. Shared (the default) drives both scenes from one pattern; Per-Scene gives Scene B an independent sequence you edit via the Edit: Scene B box. See Scenes for how to build two fully independent scenes.

Panic

A PANIC button is always visible in the header. One click instantly silences everything: it stops all notes, kills any stuck or hung voices across both scenes, clears held and latch state, and flushes the effect bus so ringing reverb and delay tails go quiet. Use it any time a note will not release or a patch runs away on you.

Quality and CPU

The oversampling Quality setting and the Auto HQ on bounce option live in the Master band. For a full explanation of Quality levels and CPU tradeoffs, see Performance & CPU.

Tip

For thick pads: set Poly to 6–8 voices, nudge Voice Detune to about 10–15%, add a little Drift, and set Variance to taste. The result sits between “digital precise” and “vintage unstable”: the sweet spot that makes a chord feel alive without going out of tune.