Unison stacks detuned copies of an oscillator to create a thick, wide, chorus-like sound

Unison plays N simultaneous copies (voices) of the same oscillator, each slightly detuned from centre pitch. The pitch offsets cause continuous amplitude/frequency beating that the ear perceives as richness, warmth, width, or — at large detuning — roughness. This emulates true analog oscillators, which drift a few cents and beat naturally; digital oscillators are perfectly stable, so the effect must be created deliberately. Two controls trade off independently: voice count (more voices = denser beating) and detune spread (in cents, or absolute Hz). Small detune gives subtle warmth; large detune gives the chorused ‘supersaw’ cloud used for leads and pads. A blend control can balance centre versus outer voices. Boundary conditions: detune trades thickness against pitch clarity, so it is undesirable where precise pitch matters; absolute-Hz detuning behaves differently at low versus high pitches than cent-based detuning; and unison CPU cost scales with voice count and oscillator complexity. When summing many voices, divide amplitude by voice count to avoid clipping.

Examples

SuperCollider (5 detuned saws):

~numVoices = 5; ~distance = 0.3;
Mix.fill(~numVoices, {|i|
  var d = (i - floor(~numVoices/2)) * ~distance;
  LFSaw.ar(220 + d)
}) / ~numVoices !2

Serum: 4–7 voices, sweep detune from ~5 cents (subtle warmth) to ~40 (aggressive supersaw); lower blend to push outer voices back. Surge Classic osc: 16 voices at modest CPU; switch to absolute-Hz detune for pitch-independent low-bass detuning. Massive: 3 voices give a balance of ‘solidity and filth.‘

Assessment

Sweep unison detune from 0 to maximum at a fixed voice count and describe the transition from warm to extreme. Compare 1, 2, and 3 voices and relate voice count and spread to beating density. Explain how absolute-Hz detune differs from cent-based at low vs high pitch, and name a scenario where unison detune is undesirable.