/ˈjuː.nɪ.sən/
Unison is a synthesizer feature that duplicates a single note across multiple oscillator voices, each slightly detuned and panned, creating a thick, wide, chorusing texture from a single played note.
Every producer has heard that sound — the wall of synth that fills a room before the drop even hits. That's not magic. That's unison, and once you understand its mechanics, you'll never hear a supersaw the same way again.
Unison is a synthesis mode in which a synthesizer generates two or more oscillator voices simultaneously for every note played, with each voice slightly offset in pitch, phase, and stereo position. Rather than hearing a single, stable oscillator, the listener perceives a cluster of near-identical tones whose subtle pitch disagreements create acoustic beating — the slow, rhythmic amplitude fluctuation that the human ear interprets as movement, depth, and organic life. At its tamest, unison imparts a gentle shimmer comparable to a well-maintained ensemble instrument; at its most extreme, it produces the dense, stadium-filling wall of sound associated with trance leads, dubstep basses, and cinematic pads.
The term borrows from music theory, where unison describes two instruments playing the same pitch — but in synthesis, the critical distinction is that voices are intentionally not in perfect unison. They are detuned by a small interval, typically measured in cents (hundredths of a semitone), so that interference patterns emerge between their waveforms. A pair of sawtooth waves tuned zero cents apart produce a single, clean tone. Shift one voice by 6–12 cents and the beating frequency becomes audible, usually in the 0.5–6 Hz range that corresponds to natural vibrato and ensemble shimmer. Push detuning to 25–50 cents and the voices begin to sound like two distinct players slightly out of tune — useful for certain styles, destructive for others.
Most modern software synthesizers implement unison as a dedicated mode accessible from the oscillator section or a global voice panel. Classic parameters include Voice Count (how many copies of each oscillator run simultaneously), Detune (the pitch spread across all voices), Stereo Spread or Pan (how voices are distributed across the stereo field), and, in more sophisticated implementations, Phase randomization, Blend (the dry/wet balance between unison and the base oscillator), and per-voice pitch curves that shape whether detuning is linear, stacked symmetrically, or offset in a chord-like pattern. Understanding how these interact — and how they translate differently across monophonic leads, polyphonic pads, and bass patches — is the central practical skill this entry addresses.
Unison is one of the most powerful tools in a synthesist's palette, but it carries significant costs. Every additional voice multiplies CPU load; a synthesizer running 8-voice unison on a 6-note chord is simultaneously computing 48 oscillators. More critically for mix engineers, unison generates dense mid-frequency energy that competes with every other element in a track. A poorly managed unison synth can mask kick transients, blur vocal intelligibility, and create a muddy low-mid buildup that resists equalization because the smearing is caused by phase interactions rather than simple level. Knowing when to use unison, at what voice count, and with what complementary processing is what separates producers who reach for it reflexively from those who deploy it with precision.
At the signal level, unison works by instantiating multiple copies of the same oscillator algorithm within a single voice allocation. When a MIDI note triggers, the synthesizer's voice manager spawns N oscillators — where N equals the configured voice count — and routes them through the same filter, amplifier, and effects chain before summing them to the audio output. Each voice receives a pitch offset calculated from the master Detune parameter. With two voices, one is typically pitched up by half the detune amount and the other down by half, centering the perceived pitch on the played note. With eight voices, the offsets are distributed across the full detune range, often spaced evenly (e.g., ±0, ±Δ, ±2Δ, ±3Δ) or mapped to a curve that places more voices near the center for a denser fundamental with spread overtones.
The acoustic phenomenon driving the unison effect is beating. When two sine waves of slightly different frequencies are summed, the result is a single tone whose amplitude rises and falls at a rate equal to the frequency difference between the two partials. For example, two oscillators tuned to 440 Hz and 443 Hz produce a 3 Hz beat — a slow tremolo that the ear attributes to movement and richness rather than a detuned instrument. With complex waveforms like sawtooth or square waves, beating occurs not just at the fundamental but at every harmonic pair, creating a rich tapestry of modulations that sounds fundamentally different from any static waveform or modulation-based chorus effect. This is why a hardware supersaw — Roland's term for a seven-voice unison sawtooth — cannot be convincingly replicated by applying a standard chorus plugin to a single oscillator: the harmonic interaction between real detuned oscillators is qualitatively different from the time-delayed copies produced by a chorus algorithm.
Stereo spread distributes voices across the panoramic field. In a simple two-voice implementation, one voice is hard-panned left and the other right. With eight voices, the synthesizer typically assigns voices to pan positions spaced evenly across the stereo field — or uses a dedicated spread algorithm that places voices symmetrically around center so that the summed mono output remains phase-coherent and retains fundamental weight. This is a critical engineering consideration: a wide unison that sounds enormous in stereo can collapse to a thin, phasey mess when summed to mono. Producers targeting broadcast, streaming normalization, or club PA systems (which are often mono in the sub frequencies) must check unison patches in mono and may need to reduce spread, increase voice count to improve mono cancellation averaging, or high-pass the unison layer and blend a separate mono fundamental beneath it.
Phase randomization is a parameter present in synthesizers like Serum, Vital, and Massive X that determines the starting phase position of each voice's oscillator cycle at note onset. With phase set to zero, all voices begin at the same point in their waveform cycle, producing a consistent transient click on every note trigger. With phase fully randomized, each voice starts at an independent random position, smoothing the attack and generating a different-sounding beating pattern on every retrigger — the hallmark of the organic, slightly unpredictable quality in analog hardware unison. Most producers use partial or full phase randomization for pad and lead patches, and locked phase for percussive bass patches where a punchy, consistent transient attack is required.
The CPU cost of unison scales linearly with voice count and is multiplied by the synthesizer's polyphony setting. An 8-voice unison patch on an 8-note polyphonic instrument runs 64 simultaneous oscillators per instance, plus the filter, modulation, and effects overhead for each. On modern hardware this is manageable, but stacking multiple unison instances on a single project — common in dense EDM and film score production — can introduce latency and require freezing or bouncing. Efficient workflow means understanding that 4-voice unison at moderate detune often delivers 80% of the perceived thickness of 8-voice at half the CPU cost, and that careful EQ and stereo processing can recover the remaining perception of width without adding voices.
Diagram — Unison: Unison voice stacking diagram showing 4 detuned oscillator voices distributed across the stereo field and their summed output waveform with beating envelope.
Every unison — hardware or plugin — operates on the same core parameters. Know these and you can work with any implementation.
Sets how many independent oscillator instances are spawned for each MIDI note. Common options are 2, 4, 6, or 8 voices; some synthesizers allow up to 16. More voices produce denser texture and wider stereo spread but multiply CPU load and can muddy the low-mid frequencies — 4 to 6 voices is the sweet spot for most commercial leads and pads.
Controls the total pitch deviation distributed across all voices, typically from 0 to 100 cents. Values of 4–12 cents yield natural ensemble shimmer; 15–30 cents creates an aggressively chorused character; above 40 cents the voices begin to sound intentionally out of tune. The beating frequency — and therefore the speed of the resulting tremolo — increases as detune rises.
Determines how far apart voices are panned, from all voices centered (0%) to voices distributed across the full stereo image (100%). High spread values dramatically widen a patch but can cause low-frequency phase cancellation in mono. For bass-heavy patches, keep spread below 40% or apply a mono low-end crossover, blending the unison only in the mid and high frequencies.
Available in synthesizers like Serum and Vital, Blend sets the level of the detuned voices relative to the center (dry) voice. At 0% only the original oscillator sounds; at 100% all voices are equal. Setting Blend to 70–80% retains a strong, centered fundamental while adding unison shimmer — a technique that preserves low-end punch on bass patches while still achieving width.
Controls whether oscillator voices begin at the same phase position (0) or at random positions each time a note is played. Zero phase produces consistent, punchy transients with a slight click — preferred for bass stabs and plucks. Full randomization gives each retrigger a unique character, emulating the organic variability of analog oscillators drifting in real time.
In advanced synthesizers such as Massive X and Pigments, the distribution of detune offsets can be shaped by a curve — linear spacing distributes voices evenly across the detune range, while exponential or stacked curves cluster more voices near the center pitch. Center-weighted curves preserve fundamental clarity while still producing outer-voice shimmer, a popular choice for supersaw-style leads.
Session-ready starting points. These are starting-point values; always verify mono compatibility before committing a unison patch to a final mix, especially on bass and kick-adjacent frequencies.
| Parameter | General | Drums | Vocals | Bass / Keys | Bus / Master |
|---|---|---|---|---|---|
| Voice Count | 4–6 | 2–3 | 2–4 | 2–4 | 2 |
| Detune (cents) | 8–18 | 4–8 | 5–12 | 3–10 | 3–6 |
| Stereo Spread | 60–80% | 20–40% | 40–70% | 30–50% | 20–35% |
| Blend | 80–100% | 60–80% | 70–90% | 65–80% | 50–70% |
| Phase Randomization | Full | Zero | Full | Partial–Full | Full |
| Mono compatibility | Check | Essential | Check | Essential | Essential |
These are starting-point values; always verify mono compatibility before committing a unison patch to a final mix, especially on bass and kick-adjacent frequencies.
The conceptual ancestor of synthesizer unison is the chorus effect achieved by massed string sections, pipe organs, and keyboard instruments such as the harpsichord, whose double-course stringing placed two slightly out-of-tune strings beneath each key. Instrument builders exploited the resulting beating to create perceived fullness and warmth long before the term was applied to electronics. The pipe organ's Celeste rank — a second set of pipes tuned sharp by a few cents — directly prefigures synthesizer unison detune and was in widespread use by the late 17th century.
In the electronic era, the first practical implementation of voice stacking appeared in the Yamaha GX-1 (1973), an organ-based synthesizer designed for live performance that could layer multiple oscillator stacks. The Roland Juno-106 (1984) popularized a simpler version through its Chorus circuit — technically a bucket-brigade delay device rather than true oscillator stacking — but the sonic result was similar enough that many producers conflate the two. True synthesizer unison in the modern sense arrived with the Roland Jupiter-8 (1981), which included a dedicated Unison mode that collapsed its eight voices onto a single key, detuning each voice independently. Engineers and keyboardists including Jan Hammer, Keith Emerson, and early Depeche Mode collaborator Martin Gore used Jupiter-8 unison leads extensively on recordings throughout the early 1980s.
The defining moment in unison history was Roland's introduction of the JP-8000 in 1996 and its SuperSAW oscillator — a single oscillator algorithm that internally computed seven detuned sawtooth waves simultaneously. The SuperSAW became the defining timbre of trance music, deployed by producers including Ferry Corsten, Paul van Dyk, and Armin van Buuren throughout the late 1990s and 2000s. Its influence was so dominant that virtually every subsequent software synthesizer included a dedicated supersaw or multi-saw mode, and Roland's JP-8000 patch architecture remains the template for modern unison implementations. Steve Duda and Lennar Digital's Sylenth1 (2007) brought high-quality eight-voice unison to a generation of laptop producers, and Xfer Records' Serum (2014) further refined the concept with independent blend, phase, and detune curve controls that gave producers surgical precision over the voice-stacking architecture.
The role of unison in popular music expanded dramatically in the 2010s as EDM production moved into mainstream pop. Producers such as Skrillex, Deadmau5, and Martin Garrix built signature sounds almost entirely around high-voice-count unison leads. Simultaneously, film composers including Hans Zimmer's Remote Control Productions team adopted massive unison string and synth layers — often stacking live ensemble recordings with synthesized unison patches — to create the dense, emotionally overwhelming textures heard in scores from 2010 onward. Today, unison is considered a fundamental synthesis parameter alongside filter cutoff and envelope attack, present in every major software synthesizer and many hardware instruments, from the sequential Prophet-6's unison mode to the Arturia MatrixBrute's oscillator detuning matrix.
Leads and supersaws: The quintessential application of unison is the trance or EDM lead — a sawtooth oscillator with 4–8 voices, 10–20 cents of detune, and 70–100% stereo spread. At these settings, the patch occupies the full stereo field and generates the beating pattern that gives trance its characteristic forward motion. To keep the mix from collapsing, experienced producers high-pass the unison lead at 150–200 Hz and blend in a separate mono oscillator or a sub-octave layer to anchor the low end. Sidechain compression triggered by the kick compresses the unison lead rhythmically, turning the wide, static texture into a breathing, pumping element that locks to the groove.
Pads and atmospheres: For cinematic pads and ambient textures, unison is set to lower detune values (4–10 cents) with partial or full phase randomization, producing slow beating that evolves over several seconds. Adding a long reverb after the unison stage further diffuses the voice interactions into a seamless wash. The detune speed — the rate of amplitude fluctuation — can be tuned to the tempo of the track by calculating the beat frequency: for a 3 Hz beat at A4 (440 Hz), set detuning such that adjacent voices are 3 Hz apart, which at 440 Hz corresponds to approximately 11.8 cents. This synchronizes the pad's natural movement to approximately 180 bpm — useful for high-tempo tracks where the pad needs to feel rhythmically engaged rather than floating.
Bass patches: Unison on bass requires the most careful handling. The standard approach is to use 2–4 voices with minimal detune (3–8 cents), reduce stereo spread to 30% or less, and use a mono low-end crossover to ensure that all energy below 150–200 Hz is summed to center before it hits the mix bus. Some producers use unison exclusively in the upper harmonics of a bass patch — achieved by high-passing the unison layer and layering it with a separate mono fundamental — producing a bass that sounds wide and textured in the mids while remaining club-system-safe in the lows. Serum's Blend parameter is particularly useful here, allowing the detuned voices to be mixed back in subtly while the center voice anchors the fundamental.
Vocal doubling and instrument emulation: Unison at very low voice counts (2) and extremely subtle detune (1–4 cents) can simulate the natural imprecision of doubled vocal performances or ensemble instruments. Applied to a synthesized flute, violin, or vocal patch, this technique adds the slight pitch instability that makes synthetic textures feel human. Unlike chorus plugins, which introduce comb filtering artifacts from time-delayed copies, oscillator-level unison modulates pitch directly, producing a cleaner, more organic doubling effect. This application is common in lo-fi and indie pop production where the goal is warmth rather than width.
One email a week. The techniques behind the terms — curated by working producers, not algorithms.
Abstract knowledge becomes practical when you can hear it in music you know. These tracks demonstrate unison used intentionally, at specific moments, for specific purposes.
One of the earliest chart hits built around a supersaw unison lead, rendered on a Roland JP-8000 with its native SuperSAW oscillator. At the 1:28 drop, the seven-voice detuned sawtooth occupies the entire upper midrange, demonstrating how high stereo spread and moderate detune (~15 cents) allows a single synth line to fill a mix without harmonic clutter. Notice how the bass is kept fully mono beneath the wide lead — a production practice that became standard in trance. The track reached number one in Germany and introduced the supersaw to mainstream pop production.
The main dubstep bass employs extreme unison detuning — likely 6–8 voices with 20–35 cents of detune — to create the aggressive, growling wobble that defined brostep. The critical production choice here is that the unison voices are processed through heavy saturation and a moving filter, so the beating interacts with the harmonic distortion to create sidebands that are not present in clean unison patches. Listen for the way the sound narrows and widens perceptibly as the filter sweeps, demonstrating the relationship between filter resonance and unison beating frequency. The track won a Grammy for Best Dance Recording in 2012.
The Rhodes-style synthesizer chord layer uses subtle 2-voice unison at very low detune (estimated 3–5 cents) to simulate the natural imperfection of a double-tracked keyboard performance. This is textbook application of unison for organic warmth rather than width — at no point does the patch feel electronically chorused or artificially widened. The result sits comfortably behind Nile Rodgers' guitar without competing. Compare this to the more aggressive unison in the synthesizer pads at 2:48, which use wider spread to fill space in the arrangement break.
The iconic portamento lead is a textbook 6-voice supersaw with detune around 15–20 cents, high-passed at approximately 180 Hz and reinforced by a separate mono pluck sample on the attack transient — a common production trick to give heavily detuned unison patches a focused front edge. The stereo spread at nearly 100% creates the enormous width that makes the drop feel physically large on a club system. The detune is audible as a fast shimmer (approximately 5–8 Hz beating rate), which locks perceptually to the 128 bpm tempo of the track.
The bass patch uses 2-voice unison with minimal detune and almost zero stereo spread — an example of unison used for subtle harmonic thickening rather than width. The mono-compatible approach is critical here since the track is heavily consumed on smartphone speakers and earbuds in mono. FINNEAS has discussed using Xfer Serum for much of the bass work; the Blend parameter at approximately 75% keeps the center voice dominant while the detuned voice adds slight movement in the upper harmonics. The result is a bass that sounds fuller than a single oscillator without introducing the phase problems that would compromise playback on small speakers.
The defining unison type in EDM production, consisting of multiple sawtooth waves (typically 7) detuned symmetrically around a center pitch. The Roland JP-8000's SuperSAW algorithm is the canonical reference; its specific voice spacing and detune curve are widely imitated in software. Best used for trance leads, dance pop chords, and any context requiring maximum width and density.
Produced by hardware synthesizers whose oscillators exhibit natural, temperature- and component-dependent pitch drift. Rather than precise cent-based offsets, drift unison creates slowly wandering detuning that varies in rate and depth over time. The result is warmer, less predictable, and less perfectly symmetric than software unison — valued for its organic character in vintage and neo-soul production.
A mode in which all voices begin at the same phase position on each note trigger, producing a consistent, punchy attack transient rather than a randomized onset. Phase-locked unison is preferred for bass patches, stabs, and percussive sounds where attack consistency matters. The sonic character is brighter and more defined at the transient than fully randomized phase unison.
Some synthesizers allow unison voices to be pitched not just in cent offsets but at musical intervals — thirds, fifths, octaves — so that a single key plays a pre-configured chord using the unison voice stack. This is distinct from a chord mode or arpeggiator; it uses the same oscillator architecture as standard unison but maps voice detuning to harmonic ratios. Produces instant harmonically rich textures without requiring MIDI chord input.
Unison applied to wavetable oscillators that cycle through different waveform shapes over time. Each voice not only detunes independently but also occupies a slightly different position in the wavetable, so the beating interacts with the timbre evolution of the wavetable scan. The result is a uniquely animated texture — constantly shifting in both pitch and harmonic content — widely used in cinematic pads and progressive electronic music.
These MPW articles put unison into practice — specific techniques, real tools, and applied workflows.