/pætʃ/
Patch is a complete, saved configuration of a synthesizer's sound-shaping parameters — oscillators, filters, envelopes, LFOs, and modulation routings — that together define a single playable sound or timbre.
Every iconic synth sound you've ever heard — the Oberheim brass stabs on 'Jump,' the Minimoog bass crawling under 'Superstition,' the Juno pad dissolving through 'Blue Monday' — began as someone deciding exactly where to set a knob, and then another, and then another, until a patch existed that had never existed before.
A patch is the complete, holistic configuration of a synthesizer at a given moment: every oscillator waveform and tuning, every filter cutoff and resonance setting, every envelope stage, every LFO rate and depth, and every modulation routing, treated as a single unified entity and saved under a name for instant recall. The word functions simultaneously as a noun (the patch itself, the saved preset) and as a verb (to patch — meaning to connect signal paths, particularly in modular and semi-modular contexts). In modern DAW-based production, patch and preset are used interchangeably, though patch carries stronger connotations of hands-on sound design while preset implies factory-delivered material.
The term derives directly from analog modular synthesizer hardware, where sound designers used short lengths of wire called patch cables to physically connect the output of one module to the input of another — routing an oscillator's audio output into a filter's input, or a low-frequency oscillator's control voltage output into a filter's cutoff frequency control input. The specific physical arrangement of those cables on the synthesizer's front panel was called the patch. Before digital memory existed, a patch was so thoroughly identified with its physical cable configuration that synthesizer manufacturers sold patch sheets — printed grids of the front panel that engineers could sketch onto, recording jack positions so a valuable patch could be manually recreated after the cables were removed.
In modern software synthesis, the physical cables are gone, but the conceptual architecture remains entirely intact. A patch in Serum, Massive, or Vital is still a description of signal flow — audio generated by oscillators, shaped by filters and amplifiers, modulated by envelopes and LFOs, with the routing between these blocks defining the patch's character as much as the individual parameter values do. The difference is that all of this is now stored in a few kilobytes of data, recallable in milliseconds, and exchangeable between producers worldwide through preset packs and sound libraries. What once required a careful sketch on graph paper now survives as a .fxp file.
The creative importance of understanding patches rather than merely loading them cannot be overstated. A producer who can read a patch — who sees a filter envelope with a fast attack and high initial depth and immediately hears the vowel-like sweep it will create, or who recognizes a slowly cycling LFO on oscillator pitch and knows it will produce subtle vibrato — has access to the entire history of electronic music as a toolkit rather than a mystery. Patch literacy is to synthesis what chord theory is to harmony: it doesn't constrain creativity, it makes creativity deliberate.
Every patch, regardless of synthesizer architecture, describes a signal flow through four functional stages. The first stage is generation: one or more oscillators produce raw audio waveforms (sine, sawtooth, square, triangle, or noise) at a pitch determined by the note played. In some architectures this stage also includes sample playback, wavetable scanning, or FM operator stacks, but the fundamental role is identical — creating the raw material that subsequent stages will shape. In a typical two-oscillator subtractive patch, one oscillator might run a sawtooth wave at concert pitch while the second runs a slightly detuned square wave, creating the beating, chorused quality characteristic of classic analog leads.
The second stage is filtering, where the harmonic content of the generated signal is sculpted. Most classic patches route the mixed oscillator signal through a low-pass filter whose cutoff frequency and resonance define the timbral brightness at any given moment. The cutoff determines which frequencies pass through (everything below the cutoff) and which are attenuated (everything above). Resonance boosts frequencies immediately around the cutoff point, creating the characteristic self-oscillating peak that defines sounds from the Moog ladder filter's saturated squelch to the Roland TB-303's liquid acid blip. A patch without interesting filter behavior is often a patch without personality — the filter is where synthesis stops sounding like synthesis and starts sounding like music.
The third stage is amplitude shaping via a Voltage Controlled Amplifier (VCA) governed by an amplitude envelope — classically described by the ADSR model: Attack (time to reach peak level after a note is triggered), Decay (time to fall from peak to Sustain level), Sustain (held level while the key is depressed), and Release (fade time after key release). A fast attack and no sustain creates a plucked transient; a slow attack with full sustain creates a swelling pad. The fourth stage, modulation, is what separates a static tone from a living sound: LFOs, additional envelopes, and modulation matrices apply time-varying changes to parameters across all prior stages — vibrato on oscillator pitch, filter movement over time, tremolo on amplitude. A patch's modulation routing is often more important than its base parameter values.
In modular synthesis the same four stages apply, but the physical routing between modules using 3.5mm or 6.35mm patch cables makes every routing decision explicit and visible. A patch on a Eurorack system might route an oscillator's audio output into a wavefolder before the filter, then send a second, inverted envelope to the filter cutoff while the primary envelope controls VCA level — a configuration impossible on many fixed-architecture synthesizers but immediately buildable from discrete modules. This explicit signal-flow visibility is why experienced synthesists often recommend building patches on modular systems even if you primarily work in software: it makes the causal relationships between parameters physically undeniable.
When a patch is saved, the synthesizer or DAW writes the complete parameter state — every knob value, every routing assignment, every modulation depth — to a data structure that varies by format (.fxp for VST, .aupreset for Audio Units, proprietary binary formats for hardware). On hardware synthesizers from the mid-1970s onward, patch memory was implemented in battery-backed SRAM or, later, EEPROM, allowing patches to persist without power. The Roland Juno-106 stored 128 patches in RAM; the Yamaha DX7 shipped with 32 factory ROM patches and 32 user RAM locations, a limitation that partly explains why certain DX7 patches — Electric Piano 1, Brass, Bass 1 — became inescapable in 1980s pop production: most engineers simply left the factory patches loaded.
Diagram — Patch: Signal flow diagram of a subtractive synthesis patch: Oscillators feed into a Filter, then VCA, with Envelope and LFO providing modulation.
Every patch — hardware or plugin — operates on the same core parameters. Know these and you can work with any implementation.
Sawtooth waves contain all harmonics (odd and even) and are the brightest, most harmonically rich choice — ideal for leads, brass, and strings. Square waves contain only odd harmonics and sound hollow and woody, like a clarinet. Sine waves contain no harmonics and sit almost invisibly in a mix as sub bass or FM carrier. The waveform choice is the single most foundational decision in a patch because every subsequent stage is shaping what the oscillator provides.
Expressed in Hz, cutoff sets the frequency above which the filter begins attenuating signal in a low-pass configuration. At 200 Hz a patch sounds dark, muffled, almost underwater; at 8 kHz it sounds wide open and bright. The critical insight is that cutoff is rarely set statically in a musical patch — envelope modulation of cutoff by even ±20% creates the vowel-like movement that makes synthesizers feel alive. On Moog-architecture synths, cutoff responds to keyboard tracking so higher notes open the filter proportionally.
Resonance (also labeled Q or Emphasis) boosts the frequencies immediately surrounding the cutoff point. At moderate settings (20–60%) it adds presence and sibilance to pads and leads. Above about 80% on most analog-modeled filters, the filter begins to self-oscillate, generating its own pure tone at the cutoff frequency independent of any oscillator input — a technique used to create laser sfx, acid squelches, and the trademark TB-303 blip. High resonance combined with a fast filter envelope produces the 'wah' effect central to classic funk synthesis.
Attack time fundamentally determines whether a sound is percussive or swelling. Under 5ms produces a click-fronted transient useful for bass plucks and lead stabs; 5–50ms gives a natural struck quality; 200ms+ creates a slow bloom characteristic of orchestral string pads and cinematic atmospheres. Both amplitude and filter envelopes have independent attack parameters, allowing a patch to have an instant-on loudness profile while the filter blooms open slowly, creating the 'reverse' timbral movement of many Vangelis-style pads.
LFO rate is typically expressed in Hz or as a tempo-synced note value (1/4, 1/8, 1/16). Below 1 Hz produces the gentle vibrato and slow filter sweeps of expressive leads and moving pads. 1–10 Hz covers musical vibrato and tremolo ranges. Above ~20 Hz the LFO crosses into audio-rate territory, producing sidebands and FM-like timbral effects rather than perceptible movement. Tempo-syncing the LFO rate to the project BPM creates rhythmic timbral motion — a 1/16 LFO on filter cutoff produces the stuttering, rhythmic filter chops central to house and techno production.
Modulation depth (also called amount, intensity, or modulation index depending on architecture) scales how dramatically a modulation source moves its target parameter. In bipolar configurations, a depth of ±0 produces no movement; ±100% sends the parameter through its full range each cycle. In FM synthesis, modulation index has a direct mathematical relationship to the number and amplitude of sidebands generated — an index of 0 produces a pure sine carrier, while an index above 2 creates complex, spectrally rich tones. Setting depth too high creates obvious, cartoonish modulation; subtle depths (5–15% of the destination's range) often produce more musical results.
Portamento sets the time in milliseconds or seconds that the synthesizer takes to slide pitch from one note to the next rather than jumping immediately. At 0ms, pitch changes are instantaneous; at 200–500ms, a characteristic slide connects notes that gives bass lines and leads an expressive, vocoder-adjacent quality. The TB-303's portamento — applied only when notes overlap in legato playing — is responsible for the characteristic slide of acid house, where non-overlapping notes jump cleanly but connected notes glide. On polyphonic patches, portamento is generally avoided as it creates incoherent pitch slides between simultaneously-sounding voices.
Session-ready starting points. These ranges apply to subtractive synthesis patches; FM and wavetable architectures will have equivalent parameters with different scales.
| Parameter | General | Drums | Vocals | Bass / Keys | Bus / Master |
|---|---|---|---|---|---|
| Oscillator Waveform | Sawtooth | Square / Noise | Sine / Triangle | Sawtooth / Pulse | N/A (source stage) |
| Filter Cutoff | 2–8 kHz | 500 Hz–3 kHz | 3–6 kHz | 200–900 Hz | N/A |
| Filter Resonance | 15–35% | 0–20% | 0–15% | 20–55% | N/A |
| Amp Envelope Attack | 2–30 ms | 0–4 ms | 10–80 ms | 5–25 ms | N/A |
| Amp Envelope Release | 80–400 ms | 40–200 ms | 200–800 ms | 100–500 ms | N/A |
| LFO Rate | 0.3–6 Hz | Sync 1/16–1/8 | 4–7 Hz (vibrato) | Sync 1/4–1/16 | N/A |
| Portamento | 0–50 ms | 0 ms | 0–30 ms | 30–200 ms | N/A |
These ranges apply to subtractive synthesis patches; FM and wavetable architectures will have equivalent parameters with different scales.
The word patch entered synthesizer vocabulary through the telephone and radio industries, where a patch bay — or jackfield — allowed engineers to route audio and control signals between equipment by inserting short cables into a panel of sockets. Robert Moog's first commercial modular synthesizer systems, introduced at the 1964 Audio Engineering Society convention and developed commercially through 1965–67, borrowed this architecture directly: each module (oscillator, filter, envelope generator, VCA) had input and output jacks, and a patch cable physically routed signal from one module to another. The resulting configuration — the patch — was the sound. Don Buchla's independently developed West Coast modular systems, emerging from the San Francisco Tape Music Center around 1963–64, used the same cabling concept but with a radically different signal-flow philosophy emphasizing complex waveshaping and low-frequency oscillators over filter-based timbre sculpting.
Through the late 1960s and into the 1970s, the ephemeral nature of patches was one of electronic music's defining challenges. The Columbia-Princeton Electronic Music Center's RCA Mark II Sound Synthesizer, used extensively by Milton Babbitt, Vladimir Ussachevsky, and Charles Wuorinen beginning in 1957, stored its routings on punched paper rolls — an early, mechanical form of patch memory. Engineers at studios including Cologne's WDR and London's BBC Radiophonic Workshop developed meticulous personal notation systems to document Moog and EMS patches, sketching cable positions and writing parameter values in the margins of score paper. Delia Derbyshire at the BBC and Wendy Carlos (recording Switched-On Bach in 1968) were famous for their obsessive patch documentation practices. Carlos reportedly took hours rebuilding patches for the Bach album because even a half-millimeter deviation in slider position changed the sound.
The transition from analog patch cables to digital patch memory occurred with the Prophet-5, designed by Dave Smith at Sequential Circuits and released in January 1978. The Prophet-5 was the first commercially available polyphonic synthesizer with fully programmable patch storage: 40 patches in RAM and 40 in ROM, recallable instantly via numbered buttons. This was technically enabled by a microprocessor reading the current state of every front-panel potentiometer and writing those values to battery-backed CMOS RAM. The impact on music production was transformative — live keyboardists could now change sounds mid-performance reliably, and studio engineers could save the exact patch used on a session and return to it months later. The Prophet-5's patch architecture became the template for every hardware synthesizer that followed. Roland's Juno-106 (1984) offered 128 patches; Yamaha's DX7 (1983) carried 32 user and 32 ROM patches, with external cartridges expanding capacity.
The DX7 era (1983–1990) created a new sub-discipline of patch programming as FM synthesis required a fundamentally different mental model than subtractive patching. Programming a DX7 patch meant assigning operator roles (carrier vs. modulator), setting modulation indices, and designing algorithms — possible combinations of the six operators — rather than simply setting filter cutoffs. Manufacturers like Yamaha, Korg, and Roland hired dedicated sound designers to create factory patch banks, and an entire industry of third-party patch libraries emerged. Keyboard magazine published patch parameter lists monthly. By the early 1990s, companies like Patchman Music were selling thousands of DX7 patch cartridges by mail order, establishing the commercial preset marketplace that now encompasses millions of DAW-format files distributed via Splice, Loopmasters, and individual artist stores. The word patch had completed its journey from a physical cable arrangement to a software data object, but the underlying concept — a complete, nameable, shareable sound configuration — remained unchanged across sixty years of technological evolution.
In electronic dance music production, patches function as both the melodic and textural skeleton of a track. A deep house producer building a chord progression in Ableton Live might load a vintage Juno-style pad patch as a starting point, then systematically edit: opening the filter cutoff from 40% to 65% to let more high-frequency content through the mix, slowing the LFO rate to sync at 1/4 note against the 124 BPM project tempo, extending the release from 300ms to 800ms so chords breathe between hits, and finally adding a subtle velocity-to-filter-cutoff modulation so played notes feel dynamic. What began as a factory preset is now an original patch tailored to the track. This process — called patch editing or sound design — is the core practical skill of synthesizer-based production.
In hip-hop and trap, patches most commonly appear in two roles: tuned 808 bass patches and atmospheric lead patches layered under vocal hooks. The classic trap 808 patch is deceptively simple — a sine oscillator with a very fast amplitude attack (under 2ms for the initial transient click), a slow exponential decay into a sustained sine tone, portamento set to 80–150ms for characteristic pitch slides between notes, and minimal filtering. The 'knock' of the 808 comes from a very brief additional noise burst at the attack stage, added either in the patch itself or via a parallel layered sample. Producers like Southside and Metro Boomin have developed signature 808 patches with specific key tracking behaviors and pitch envelope amounts that differentiate their bass from generic library presets.
For synthesized lead sounds in pop and film scoring, patch design often centers on creating movement through careful modulation routing. A patch that sounds compelling throughout a 16-bar section typically has at least three simultaneous modulation sources: an amplitude envelope giving each note its shape, a filter envelope adding timbral movement over the first 200ms of each note, and an LFO adding gentle vibrato or filter breathing. Composers like Hans Zimmer, who uses Xfer Serum, Omnisphere, and custom Eurorack rigs extensively, often begin film patches by designing the modulation architecture before setting base parameter values — deciding how the sound changes over time before deciding what it fundamentally sounds like.
Modular synthesizer patches used in studio production (as opposed to live performance) are often treated as extended instruments rather than fixed presets. A producer building a Eurorack patch for a rhythmic texture might route a clock signal through a clock divider to trigger both an envelope generator (controlling a filter VCA) and a sample-and-hold circuit (generating stepped random voltages that modulate pitch), creating an evolving, semi-random melodic texture that was designed rather than programmed in the conventional sense. This approach — common among producers like Jon Hopkins, Actress, and Floating Points — produces patches with a generative quality that purely software-based patches rarely match, because the physical patch cable routing creates constraints that force unexpected musical results.
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 patch used intentionally, at specific moments, for specific purposes.
The opening clavinet sound is a heavily processed patch built around Wonder's Hohner Clavinet D6 run through a Maestro Funk Machine envelope follower and a wah pedal — an early example of real-time patch modulation on a keyboard instrument. Listen for how the filter responds to picking dynamics rather than keyboard velocity, creating an organic, touch-sensitive quality that purely static patches cannot replicate. The bass line uses a separate Minimoog patch with a low cutoff (approximately 300 Hz), moderate resonance, and a very fast filter envelope giving each bass note a subtle front-end click. The two patches occupy completely non-overlapping frequency spaces, demonstrating practical patch complementarity in arrangement.
The main melodic hook uses a vocoder patch — not a simple vocoder preset but a carefully constructed carrier signal (a sawtooth-wave synthesizer patch tuned to the formant frequencies of the French phrase being encoded) combined with the modulator (vocal recording). The bass that enters at 0:16 is a classic subtractive patch: sawtooth oscillator through a low-pass filter set around 600 Hz with moderate resonance, amplitude envelope with 5ms attack and 200ms release. At 0:28 the filter cutoff sweeps open to approximately 2 kHz over four bars — a manually automated filter on a static patch creating the track's most recognizable textural moment. This demonstrates that patch parameter automation is as compositionally important as the patch design itself.
The opening pad is widely attributed to an Oberheim Matrix-12 with a patch featuring detuned oscillators (approximately ±8 cents between the four voices), a low-pass filter at about 60% cutoff with slow LFO modulation (approximately 0.15 Hz) creating the gentle filter breathing, and an amplitude envelope with a 400ms attack and 1.2-second release. The slow LFO rate means each filter cycle takes about 6.5 seconds — longer than a full bar at the track's tempo — so the modulation feels like movement in the sound rather than an audible rhythm. This is the distinguishing characteristic of textural pad patches: modulation rates that operate below or far above musical rhythm to avoid drawing attention to the mechanism.
The arpeggiated lead synth from the chorus demonstrates a classic pulse-width modulation patch: a single square wave oscillator whose pulse width is modulated slowly by an LFO (approximately 0.4 Hz), creating a continuously shifting timbral quality that makes the single oscillator sound like two detuned voices. The filter cutoff is relatively high (5–7 kHz) with near-zero resonance, putting all the character in the PWM modulation. At 1:05, a second patch — a shorter, harder square bass — enters below. The two patches share the same basic oscillator type but are distinguished entirely by their filter settings, demonstrating how dramatically filter cutoff alone can differentiate patches with identical source material.
The most common patch type: oscillators generate harmonically rich waveforms that are then filtered (subtracted) down to the desired timbre. Subtractive patches are defined primarily by their filter character — Moog ladder filters have a warm, saturating low-end; Roland's IR-3 chip in the Juno series has a cleaner, slightly glassy quality. The majority of vintage analog sounds in contemporary production — bass lines, leads, pads, brasses — are subtractive patches.
FM (Frequency Modulation) patches use audio-rate oscillators (operators) to modulate each other's frequency rather than filtering a source signal. The modulation index and operator ratio are the primary timbral parameters; there is no filter in the classical sense. FM patches are capable of bell-like tones, metallic textures, electric piano simulations, and percussive transients that subtractive architectures cannot easily replicate. The DX7's Electric Piano 1 patch is the single most-heard synthesizer sound in 1980s pop music.
Wavetable patches use oscillators that scan through a table of single-cycle waveforms rather than playing a fixed waveform, creating evolving timbral movement even without LFO or envelope modulation. The scan position — where in the wavetable the oscillator currently reads — can be modulated by envelopes, LFOs, or MIDI controllers for dramatic spectral morphing. Serum's wavetable engine, capable of importing custom wavetables from audio samples, has made this architecture the dominant format in contemporary electronic music sound design.
Physical cable-routed patches where the routing itself is a fundamental design parameter. Semi-modular systems like the ARP 2600 and Make Noise 0-Coast have normalled (default) connections that work without cables, making them accessible for producers who want to explore signal routing without committing to a full modular case. Eurorack modular patches can involve dozens of modules and can produce self-generative, evolving textures that are effectively unique at every playback — suitable for ambient, experimental, and film score applications.
ROMpler patches combine short looped samples (attack transients, sustain loops) with subtractive or simple synthesis processing. The Roland D-50's famous patches — such as the widely imitated 'Digital Native Dance' — layered sampled attack transients with purely synthesized sustain tones, a technique called Linear Arithmetic synthesis. The Korg M1's piano and organ patches defined the sound of late-1980s and early-1990s pop production. Contemporary sample-based patches in Kontakt and similar platforms extend this architecture to extremely large (multi-GB) sample sets with sophisticated round-robin and velocity layering.
These MPW articles put patch into practice — specific techniques, real tools, and applied workflows.