/rɪˈliːs/
Release is the time a compressor or dynamic processor takes to stop attenuating the signal after it falls below the threshold. It shapes sustain, groove, and the natural decay of instruments in a mix.
Get the attack wrong and your transients suffer. Get the release wrong and your entire mix stops breathing — and most producers never hear it until someone else points it out.
Release is the fourth stage of the classic ADSR envelope model as applied to dynamic processors — compressors, limiters, expanders, and noise gates. Specifically, it defines the amount of time, measured in milliseconds, that a processor takes to cease gain reduction once the input signal has fallen below the threshold level. Where attack governs the initial clamp on a signal's peak, release governs the recovery: the moment the compressor "lets go" and allows the signal to return to its unprocessed, or unity-gain, state. A fast release of 10–50 ms restores full level quickly; a slow release of 300 ms to several seconds eases the signal back gradually.
The perceptual consequence of release time is profound and often underestimated. Too short a release on a mix bus can cause the compressor to pump rhythmically — the classic "breathing" artifact where the background noise and room ambience swell audibly between transient hits. Too long a release on a snare drum can cause sustained gain reduction that robs the snare of its crack and makes the kit feel sluggish and glued to the floor. The right release setting finds the boundary where dynamics are controlled but the music still feels alive and intentional.
Release interacts directly with attack time, threshold, ratio, and the musical tempo of the material. On rhythmic content — drums, bass, rhythm guitar — producers often synchronize release time to the tempo of the track, ensuring the compressor recovers in time for the next transient hit. On sustained material — pads, strings, vocals — a longer, more musical release smooths level variations without imposing audible gain-change artifacts. Understanding this relationship is what separates producers who use compression instinctively from those who use it correctly.
In the context of noise gates and expanders, release operates inversely: it controls how long the gate stays open or how slowly the expander allows a signal to return to full attenuation after the signal drops below the threshold. A slow gate release allows a drum's natural reverb tail to fade naturally rather than being cut off abruptly. A fast gate release is used on dialog, drum samples, or gated synth effects where a hard, defined chop is intentional — think the gated reverb snare that defined 1980s pop production.
Modern dynamics processors offer several variations on release behavior: fixed-time release, program-dependent release (where the release time adapts automatically to the incoming signal's dynamics), dual-stage release (a fast initial release followed by a slower secondary release), and look-ahead release for limiters. Each approach solves a different sonic problem, and knowing which mode to reach for is as important as knowing the millisecond value to dial in.
At the circuit level, a compressor's release behavior is implemented through an envelope follower — an RMS or peak detector that tracks the signal's level over time. When the signal drops below the threshold, the envelope follower begins discharging, and the gain reduction applied to the VCA (voltage-controlled amplifier), optical element, or digital gain stage follows this discharge curve. The release time constant defines how quickly that discharge happens. In analog designs, this is often a capacitor-resistor (RC) circuit: a larger capacitor or resistor increases the time constant, yielding a slower release. In digital compressors, the equivalent is a coefficient applied to a first-order IIR (infinite impulse response) filter that smooths the gain reduction signal.
The shape of the release curve matters as much as its duration. A linear release rises at a constant rate from maximum gain reduction back to zero dB of gain change — predictable but sometimes mechanical-sounding. An exponential release, which is what most analog RC circuits produce naturally, releases quickly at first and then tapers off, which tends to sound more natural and musical because it mirrors the way acoustic instrument envelopes behave. Program-dependent release algorithms, pioneered in units like the dbx 160 and later refined in Neve's 33609 and API's 2500, analyze the incoming audio and shorten or lengthen the release on the fly to minimize pumping on complex, dynamic material. This is why program-dependent compressors often sound "transparent" even at aggressive settings.
In the digital domain, look-ahead buffering allows a limiter's release to begin before the signal technically drops below the threshold, by pre-reading an incoming signal buffer of 1–20 ms. This is why brickwall limiters on mastering chains can apply extreme gain reduction with minimal audible artifacts — the release is already unwinding before the gain reduction peak is fully applied. It does, however, introduce a latency equal to the look-ahead window, which must be compensated for in the DAW's delay compensation system.
A commonly misunderstood interaction is the relationship between release time and the compressor's ratio and threshold. A high ratio (say, 8:1) combined with an aggressive threshold creates deep gain reduction events. If the release is short, the gain reduction snaps back fast — creating a pronounced punch and transient shape. If the release is long, the gain reduction lingers, causing level buildup and potential inter-transient distortion if the compressor doesn't fully recover before the next peak arrives. This "over-compression" artifact — where the gain reduction from one transient is still active when the next arrives — is sometimes called "riding the gain" or compression buildup, and it is one of the most common causes of a mix feeling dense, flat, and fatiguing at louder playback levels.
The practical upshot: use your compressor's gain reduction meter as a visual guide for release behavior. Watch it drop on a transient and observe how quickly it returns to zero. If it does not fully recover between beats at your track's tempo, either shorten the release or raise the threshold to reduce the depth of gain reduction events. Either approach restores the compressor's ability to respond cleanly to every new transient rather than accumulating static, always-on attenuation.
Diagram — Release: Waveform diagram showing compressor gain reduction over time, with Attack, Hold, and Release phases labeled, demonstrating fast vs slow release recovery curves.
Every release — hardware or plugin — operates on the same core parameters. Know these and you can work with any implementation.
The primary parameter: the time from when signal drops below threshold to when gain reduction returns to 0 dB. Typical ranges are 10 ms (very fast, punchy, pump-prone) to 2,000+ ms (very slow, transparent on sustained material). On kick drums, 40–100 ms is a common starting point; on bus compression, 150–400 ms is typical.
When engaged, the compressor's internal envelope follower dynamically adjusts release time based on the density and duration of gain reduction events. On dense, mixed material this prevents pumping by slowing the release when many transients arrive in quick succession. Found on the dbx 160, SSL G-Bus compressor, and most modern mastering-grade plug-ins such as FabFilter Pro-C 2.
Found on gates, expanders, and some compressors, hold keeps the gain reduction (or gate closure) in place for a fixed duration before the release phase begins. Essential on gates for drums: a hold of 50–200 ms prevents the gate from chattering on noisy snare hits with multiple transient peaks. On compressors, hold is rare but exists on vintage limiters like the Fairchild 670.
Linear release applies a constant rate of gain-reduction recovery — useful for predictable, transparent ducking effects and parallel compression. Exponential (logarithmic) release, the default in most analog-modeled processors, releases quickly at first and slows toward the end, mimicking acoustic instrument decay curves. Exponential release tends to sound more natural on melodic content; linear release is more controllable for rhythmic sidechaining.
Used in high-end dynamics processors such as the Neve 33609 and the SSL 4000 channel strip compressor, dual-stage release employs an initial fast release followed by a slower secondary release. This lets the compressor recover its gross gain reduction quickly (avoiding slug-like punch) while easing the final few dB back smoothly (avoiding the click or "thud" of abrupt recovery). Reproduced in plug-in form in the UAD SSL 4000 E channel strip and Waves SSL G.
Exclusively available in digital compressors and limiters, look-ahead reads 1–20 ms into the future via a buffered delay on the signal path. This allows the release ramp to begin before gain reduction is at its maximum, producing a smoother transition. Used in brickwall mastering limiters (Fabfilter Pro-L 2, iZotope Ozone Maximizer) to maintain perceptual loudness while minimizing distortion at true-peak ceiling settings of −1 dBTP.
Session-ready starting points. These are session-usable starting points — always confirm by watching the gain reduction meter recover fully between transients at your track's tempo.
| Parameter | General | Drums | Vocals | Bass / Keys | Bus / Master |
|---|---|---|---|---|---|
| Typical Range | 10 ms – 2 s | 30 – 150 ms | 100 – 400 ms | 80 – 300 ms | 150 – 500 ms |
| Starting Point | 120 ms | 60 ms | 200 ms | 150 ms | 250 ms |
| Pumping Risk | Medium | High (fast tempos) | Low | Medium | High (dense mixes) |
| Auto Mode Advised? | Sometimes | Rarely | Often | Sometimes | Yes |
| BPM-Linked Tip | 1 bar = 60000/BPM ms | Set to 1/8 note value | Not usually tempo-linked | ½ bar or 1 bar | ½–1 bar at tempo |
| Interaction w/ Ratio | High ratio → shorten release | 4:1+ → 40–80 ms | 2:1–4:1 → 150–300 ms | 3:1–6:1 → 80–200 ms | 2:1–4:1 → auto preferred |
| Gate Release Range | 10 ms – 1 s | 50 – 300 ms | 100 – 500 ms | 80 – 400 ms | N/A |
These are session-usable starting points — always confirm by watching the gain reduction meter recover fully between transients at your track's tempo.
The concept of automatic gain control (AGC) — the technological ancestor of the modern compressor — dates to Harold Stephen Black's work at Bell Laboratories in the 1920s, where it was used to regulate telephone transmission levels over varying line conditions. The first broadcast-oriented limiter, the Western Electric 110-A, appeared in 1932 and used tube-based amplifier gain control to prevent transmitter overloading. These early units had release times governed entirely by the RC time constants of their circuit components, with no user-adjustable parameter — the release was simply a property of the hardware. Engineers chose hardware based on the release behavior they wanted, not a knob they could turn.
User-adjustable release controls entered professional audio in the 1950s and 1960s as variable-mu (vari-mu) tube compressors became studio staples. The Fairchild 670, introduced in 1959 and designed by Rein Narma for Fairchild Recording Equipment Corporation, offered six fixed time-constant positions selected by a switch — each position setting both attack and release simultaneously via different combinations of resistors and capacitors in its sidechain. Engineers at Capitol Studios, Abbey Road, and Columbia Records learned to use position 3 or 4 on drum buses for its characteristic slow-attack, medium-release profile that added weight without killing transients. The Fairchild's release behavior, with its program-dependent secondary characteristic on the vari-mu element, is still considered one of the most musical ever produced.
The 1970s saw the democratization of adjustable release through solid-state VCA compressors. The dbx 160, released in 1971 and designed by David Blackmer, introduced RMS detection with program-dependent release that automatically adapted to incoming audio — a genuine engineering innovation that made the 160 usable on a wider range of material than any prior unit. The UREI 1176 (1967, designed by Bill Putnam Sr.) used a fixed-ratio FET topology with fast attack and faster release options than tube units, enabling the aggressive parallel compression techniques later codified by engineers like Chris Lord-Alge and Andrew Scheps. On the SSL 4000 console, introduced in 1976, the bus compressor's "auto" release became one of the defining sounds of 1980s and 1990s pop and rock — audible on records from Michael Jackson's Thriller (1982, engineered by Bruce Swedien) to Oasis's (What's the Story) Morning Glory? (1995, produced by Owen Morris).
In the digital era, plug-in developers gained the ability to implement release curves of arbitrary complexity — look-ahead, multi-stage, frequency-selective, and fully program-dependent algorithms all became practical. Waves's introduction of the Renaissance Compressor in 1999 popularized "Electro" and "Opto" release modes in the plug-in domain. FabFilter's Pro-C 2 (2012) introduced a continuously variable release-shape parameter, allowing producers to morph between linear and exponential recovery curves in real time. iZotope's Neutron series (2016–present) brought machine-learning-assisted release time suggestion, analyzing incoming audio and recommending settings based on transient density — a direct evolution from Blackmer's program-dependent concept of five decades earlier. Today, release remains one of the most actively researched parameters in dynamics processing, with ongoing work in adaptive, multi-band, and perceptually weighted release algorithms appearing in academic publications from the AES.
Drums and percussion: Release time on drum compression is arguably the most mix-critical application of the parameter. On a kick drum, a release of 40–80 ms allows the compressor to recover before the next kick hit at 120 BPM (each beat = 500 ms, so even a 150 ms release is safe at that tempo). Shorter releases of 20–40 ms increase the sense of punch and transient snap; longer releases of 100–200 ms add weight and sustain to the body of the kick. On a full drum bus, producers like Greg Wells and Ryan Hewitt use the SSL-style bus compressor with release times of 100–300 ms in auto mode, deliberately allowing slight, musical pumping that creates a sense of the kit "breathing" with the groove. For snare parallel compression, a fast release of 30–60 ms brings up the room tail while preserving the crack of the initial hit — a technique popularized in the New York parallel compression workflow.
Vocals: Vocals require release times long enough to avoid audible gain-change artifacts between words but short enough to track the dynamic range of a performance. A useful starting framework: set release so the gain reduction meter returns to zero between phrases, not between syllables. On lead vocals, 150–300 ms is a common working range. For heavily processed lead vocal tracks — hip-hop or hyperpop, for example — faster releases of 60–100 ms create an aggressive, squashed quality that glues the vocal into the beat. Engineers tracking this approach include Alex Tumay (Young Thug's mixer) and Jahlani Bakari. For backing vocals, slower releases of 300–500 ms smooth out level differences between singers without audible compression movement.
Bass guitar and synth bass: Bass compression release interacts directly with the low-frequency content of the instrument. Because low frequencies take longer to complete a waveform cycle (a 60 Hz fundamental = 16.7 ms per cycle), release times shorter than 30–40 ms can cause intra-cycle distortion, where the gain reduction changes within a single low-frequency waveform cycle, introducing harmonic distortion. Keeping release at 60 ms or above on bass sources avoids this. For synth bass with tight, percussive attacks — trap 808s, for instance — release times of 80–200 ms allow the sustain to bloom naturally after the initial transient is caught by a fast attack setting.
Mix bus and mastering: On a mix bus, release time governs the macro-dynamic feel of an entire record. A release of 200–400 ms on a bus compressor applying 2–4 dB of gain reduction creates the "glue" effect: individual elements lock together rhythmically as the bus compressor's release modulates their collective level in sync with the track's tempo. The key is ensuring that release is short enough that the compressor fully recovers between the loudest transient events (typically kick + snare on the downbeat), but not so short that it pumps audibly. In mastering, release times on brickwall limiters are typically set short (10–50 ms) with look-ahead enabled, while the preceding compression stage uses a slower program-dependent release to smooth density before the limiter stage.
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 release used intentionally, at specific moments, for specific purposes.
The mix bus compression on this track is a textbook example of correctly set release time on a funk-influenced arrangement. The drum kit, bass, and rhythm guitar breathe together with a unified dynamic envelope — the release on the bus compressor (reportedly an SSL-style unit) is long enough to create cohesion but short enough to let Nile Rodgers's guitar attack punch through on every sixteenth-note chord. Listen on headphones for the subtle swell of room ambience between guitar strokes: that is the compressor releasing. The track sits at approximately 1–2 dB of bus compression gain reduction throughout, which is ideal for learning how release interacts with groove at moderate compression depths.
The kick and snare on this track demonstrate deliberate short-release compression on drums. The kick hits with a sharp, almost distorted transient that speaks through the mix with unusual clarity — a product of fast attack catching the body while a fast release (estimated 30–50 ms) allows the compressor to recover and restore full level within the kick's sustain phase, adding perceived punch. The snare clap has a similar quality: a dry, present snap followed by immediate recovery. This is contrast-based compression: short release on drums, long release on the 808, creating a mix where elements occupy separate dynamic spaces.
The vocal compression on this track is a masterclass in slow release matching a low-energy, intimate vocal performance. Eilish's voice exhibits minimal dynamic variation between syllables — a function of both her controlled delivery and the compressor's release being long enough (estimated 200–300 ms) to smooth level changes without creating audible gain pumping. The effect is a vocal that sits unnervingly close in the mix without ever jumping forward aggressively. The bass, conversely, uses a faster release to allow its transient attack to poke through the compression, creating a contrast that defines the track's quiet-but-present low end.
The drum bus compression on this track — tracked at Sound City Studios on a Harrison console with outboard compression — uses a slower release that creates a pronounced pumping effect on the chorus transitions. As the full band enters after the quiet verse, the bus compressor is suddenly overwhelmed by the guitar wall and allows the room ambience to swell behind the kit. Rather than correcting this, Butch Vig embraced it as part of the track's explosive dynamic contrast. The release pumping is most audible in the room mics on the snare, where the background noise visibly rises between snare hits after the chorus drops in. This is a historically important example of intentional release-induced pumping as an artistic decision.
The release time constant is set by the producer and does not vary with the program material. This makes the behavior highly predictable and repeatable, ideal for rhythmic material where precise, tempo-synchronized recovery is desired. Fixed-time release requires more active monitoring of the gain reduction meter to avoid buildup on dense passages.
The compressor internally modulates its release time constant based on the incoming signal's dynamics — shortening release on brief, isolated transients and lengthening it on sustained, loud signals. This approach, pioneered by David Blackmer at dbx, minimizes pumping on complex material and is widely considered the most transparent behavior for bus and mastering applications. Most modern plug-ins implement this as a toggleable "Auto" mode.
Optical compressors use a light-dependent resistor (LDR) and electroluminescent panel to control gain reduction. The LDR's inherent non-linear response to light produces a release that begins quickly after the signal drops but then slows dramatically as gain reduction approaches zero — a dual-speed characteristic that sounds exceptionally natural on vocals, bass, and acoustic instruments. The LA-2A's release is effectively program-dependent by nature, varying from approximately 40 ms to 3 seconds depending on the depth and duration of gain reduction applied.
Variable-mu tube compressors use the gain characteristics of a vacuum tube biased into different operating regions by the sidechain signal. Release in these units is inherently program-dependent — the tube's recovery from deep compression is slower than recovery from shallow compression, giving vari-mu units a smooth, gradual release that is widely used in mastering for adding warmth and cohesion to a mix bus. The Fairchild 670's six time-constant positions each set a distinct release character, from 0.3 s (position 1) to over 3 s (position 6).
In noise gates and downward expanders, release controls how quickly attenuation is restored after the signal drops below the threshold. A fast gate release (10–50 ms) produces a sharp, chopped-off sound used deliberately on gated snare effects or synth chops. A slow gate release (100 ms – 1 s) allows reverb and room tails to decay naturally, avoiding the audible guillotine cut that can make gated drums sound artificial. The Drawmer DS201 became standard on 1980s and 1990s recordings specifically because its variable release allowed natural-sounding gate behavior on live drum room mics.
These MPW articles put release into practice — specific techniques, real tools, and applied workflows.