/priː dɪˈleɪ/
Pre-Delay is the time gap—measured in milliseconds—between a dry signal and the first reflection of a reverb tail. It preserves the attack and intelligibility of a sound while still placing it inside a convincing acoustic space.
Most producers reach for less reverb when a mix sounds washed out. The real fix is almost always more pre-delay — it's the gap that lets your sounds breathe inside the space instead of drowning in it.
Pre-delay is the controlled silence inserted between the arrival of a dry, direct signal and the moment the reverb's first reflection begins to sound. Measured in milliseconds (ms), it is one parameter on virtually every hardware reverb unit, software reverb plugin, and DAW stock effect manufactured in the last four decades. Although the number itself can be small — anywhere from 0 ms to roughly 250 ms in most implementations — its perceptual impact on a mix is enormous. Pre-delay is the single most direct tool a producer has for governing where a sound appears to sit spatially: right at the front of the room, or sunk deep inside it.
In natural acoustics, when a sound source fires in a real room, the direct sound reaches your ears first. Reflections off walls, floors, and ceilings arrive fractions of a second later — this gap is determined by the physical distance those surfaces are from the source. The brain uses this delay to localize a sound: a long gap signals a large, distant space; a short gap signals proximity. By artificially controlling that gap in a mix, a producer can decouple a sound's perceived position in the stereo field from its apparent room size. You can have a voice sitting close and intimate while still wrapping it in the bloom of a cathedral reverb — an acoustic impossibility in the real world, but a standard tool of recorded music.
Pre-delay is distinct from the decay time (RT60), diffusion, and early-reflection level of a reverb, though all four interact. While decay time determines how long the reverb tail rings and diffusion controls how quickly reflections merge into a dense wash, pre-delay determines when any of that reverb begins at all. It is the threshold of space. Increasing it pushes the reverb further behind the dry signal, making the attack of a transient — a snare hit, a consonant, a plucked string — land with full clarity before any spatial smear arrives. Decreasing it to zero causes the reverb to fuse with the attack, creating an intimate, enveloping quality that can sound lush on sustained pads or oppressively dense on rhythmic material.
The parameter is sometimes labeled differently across platforms — Logic Pro's Space Designer calls it simply "Pre-Delay," Valhalla's reverbs use the same label, Lexicon hardware historically used "Pre-Delay" as well, while some budget plugins shorten it to "Predelay" or fold it under a "Room" or "Space" tab. Regardless of labeling, the function is identical: delay the onset of the wet signal relative to the dry. Understanding pre-delay is foundational not just for mixing, but for sound design — in synthesizer and sampler reverb algorithms, pre-delay is equally available and equally powerful.
Mechanically, pre-delay is implemented as a digital delay line inserted at the input stage of the reverb algorithm, before any early reflection, diffusion, or late-reverb processing begins. When an audio sample enters the reverb engine, a buffer stores it for the specified number of milliseconds. During that buffer time, the dry signal passes through unaffected. Only after the buffer has elapsed does the delayed copy of the signal feed into the reverb's reflection network. The dry signal and the reverb output are then summed at the plugin's wet/dry mix stage. The result is that the listener hears the direct, unprocessed sound followed, after the pre-delay gap, by the spatial bloom of the reverb.
The perceptual mechanism underlying pre-delay exploits the Haas Effect, also known as the precedence effect. Discovered by physicist Helmut Haas in 1951, the effect describes how the human auditory system fuses sounds arriving within approximately 30–40 ms of each other into a single perceived event, attributing its location to the first-arriving sound. When the reverb's reflections arrive within that fusion window (0–30 ms), the brain integrates them with the attack, and the source sounds embedded in the space — present but not always well-defined. When reflections are delayed beyond the fusion threshold (above ~30–40 ms), the brain begins to perceive the reverb as a separate, trailing event, and the direct signal snaps into sharp focus at the front of the image. This is why a 25 ms pre-delay on a lead vocal can make the difference between a voice that sits in the mix and one that sits on top of it.
Pre-delay interacts directly with the tempo of a track. Because reflections that land on or near rhythmic subdivisions reinforce the groove rather than blurring it, many engineers sync pre-delay to the session BPM. A 16th note at 120 BPM equals exactly 125 ms; an 8th note equals 250 ms; a dotted 8th note equals 187.5 ms. The classic formula is: pre-delay (ms) = 60,000 ÷ BPM × note value, where note value is 0.25 for a 16th note, 0.5 for an 8th, 0.75 for a dotted 8th, and 1.0 for a quarter note. Many modern reverb plugins (Valhalla Room, FabFilter Pro-R, Exponential Audio R4) include a tempo-sync toggle that automates this calculation. Synced pre-delay is especially powerful on percussion busses and lead vocals in uptempo music, because it prevents reverb smear from blurring the rhythmic grid.
The interaction between pre-delay and reverb decay time also shapes the apparent size of a space. A short pre-delay (5–15 ms) combined with a long decay (2–4 s) produces an unrealistic but frequently useful sound — the source sounds close and clear, but the room sounds enormous. A long pre-delay (60–100 ms) with a short decay (0.4–0.8 s) creates a different unrealism: the source sounds like it fired in a small, tight space, but the reflection arrives conspicuously late, producing a slap-back-plus-bloom quality prized in certain vintage recordings. Understanding these combinations allows a producer to sculpt spatial character independently of acoustic plausibility, which is precisely what most modern mix aesthetics demand.
It is worth noting that while pre-delay is typically a single static value, some advanced reverb units and plugins offer modulated or tempo-synced pre-delay that can drift subtly over time, add gentle pitch variation, or even automate changes across a song section. The underlying signal-path logic remains the same — a delay line before the reverb input — but these extensions allow the spatial character of a sound to evolve dynamically within an arrangement.
Diagram — Pre-Delay: Signal flow diagram showing dry signal, pre-delay buffer gap, early reflections, and reverb tail with timeline and millisecond labels.
Every pre-delay — hardware or plugin — operates on the same core parameters. Know these and you can work with any implementation.
Measured in milliseconds, typically 0–250 ms. Values below 20 ms create an intimate, enveloping quality where dry signal and reverb fuse perceptually. Values of 20–60 ms separate the attack from the reverb cleanly without an audible gap. Values above 60 ms introduce a perceptible space and can be synced to BPM subdivisions — at 120 BPM, a 16th note pre-delay equals exactly 125 ms.
Pre-delay and decay time work as a pair: pre-delay governs when the reverb starts, decay governs how long it lasts. A long pre-delay with a short decay (e.g., 80 ms pre-delay, 0.6 s decay) creates a distinct slap-bloom; a short pre-delay with a long decay (e.g., 10 ms, 3 s) creates a pillowy, immersive wash. Matching decay time to song tempo — so the tail decays near the next beat — keeps dense arrangements from accumulating mud.
On a send/return routing, this is typically set to 100% wet. On an insert, 15–30% wet is common for subtle room treatments. Pre-delay's effectiveness depends on this balance: if the wet signal is too loud relative to the dry, even a generous pre-delay may not save intelligibility. A useful rule of thumb is to set wet mix conservatively, then extend pre-delay until the transient clicks forward, rather than burying the dry signal in reverb and trying to recover clarity with pre-delay alone.
Early reflections are the first set of discrete echoes that arrive after the pre-delay gap. Their level relative to the tail governs how 'real' the room sounds — high early reflection levels make spaces sound concrete and specific (a studio, a hall), while low levels blur into an abstract wash. Pre-delay effectively slides the entire early-reflection-plus-tail structure backward in time; raising early reflection level then restores the perception of room proximity without moving the reverb's onset forward.
High diffusion causes reflections to smear quickly into a smooth tail — combining this with a moderate pre-delay (20–40 ms) produces the classic lush-but-clear vocal reverb heard on countless pop records. Low diffusion keeps early reflections discrete and room-specific; in this setting, a shorter pre-delay (5–15 ms) is often preferred so the discrete reflections don't sound like audible delays trailing behind the source.
Real rooms absorb high frequencies more readily than low frequencies, so HF damping simulates natural material absorption. When pre-delay is long (60+ ms), a well-damped reverb tail feels cohesive because the ear has registered the attack before it is colored by warm, diffuse energy. Without damping, a long pre-delay can expose the algorithmic or convolution reverb's texture, making the transition from silence to reverb sound artificial.
Session-ready starting points. Values are starting points at 120 BPM; scale BPM-synced times proportionally for faster or slower tempos.
| Parameter | General | Drums | Vocals | Bass / Keys | Bus / Master |
|---|---|---|---|---|---|
| Pre-Delay Range | 10–40 ms | 5–25 ms | 20–60 ms | 0–15 ms | 5–20 ms |
| BPM Sync? | Optional | Recommended | Recommended | Optional | Rarely |
| Typical Note Value | 16th note | 16th note | Dotted 8th | 8th note | 16th note |
| Decay Time Pairing | 0.8–2.5 s | 0.3–0.8 s | 1.5–3.5 s | 0.3–1.0 s | 0.5–1.5 s |
| Wet Mix (insert) | 20–35% | 10–25% | 20–40% | 8–20% | 15–30% |
| Diffusion | Medium–High | Low–Medium | High | Low–Medium | Medium |
| HF Damping | Moderate | Light–None | Moderate–Heavy | Heavy | Moderate |
Values are starting points at 120 BPM; scale BPM-synced times proportionally for faster or slower tempos.
The concept of pre-delay is inseparable from the history of artificial reverberation. Early hardware reverb solutions — the EMT 140 plate reverb, introduced in 1957 by the German company EMT (Elektromesstechnik), and the spring reverb systems built into guitar amplifiers of the same era — offered no meaningful pre-delay control. Their electromechanical designs coupled the input signal almost instantaneously to the reverberant medium. Engineers working at studios like Abbey Road and Capitol in the late 1950s and 1960s compensated by physically distancing performers from microphones or routing signals to purpose-built echo chambers, creating natural pre-delay via the speed of sound across physical distance. The distances involved were small — 10 feet yields roughly 9 ms — but the practice established the perceptual principle that would later be codified in hardware.
The digital reverb era brought explicit pre-delay control to engineers' fingertips. The Lexicon 224, released in 1978 and designed by David Griesinger, was among the first widely adopted digital reverb units to offer a dedicated pre-delay parameter with programmable values. Griesinger, who held a PhD in experimental physics and would go on to shape the science of room acoustics, understood from psychoacoustic research that early reflections arriving after a short gap were critical to the perception of source localization. The 224 became a studio standard throughout the 1980s, used by engineers including Hugh Padgham, who applied it extensively on Phil Collins's records — most famously the gated reverb snare on Peter Gabriel's "Intruder" (1980) and Collins's own "In the Air Tonight" (1981), where a short, carefully tuned pre-delay helped the cavernous reverb co-exist with an exposed, punchy snare attack.
The Lexicon 480L, released in 1986, and the Sony DRE-2000, introduced in 1983, both expanded pre-delay range and resolution, allowing engineers to dial values in single-millisecond increments up to 500 ms. Bruce Swedien, mixing engineer for Michael Jackson, used the 480L extensively on Thriller (1982) and Bad (1987), calibrating pre-delay on Jackson's vocals against the session's tempo to keep the voice articulate inside dense, layered reverb treatments. Concurrent with hardware development, academic research by Manfred Schroeder at Bell Labs — building on Schroeder's landmark 1961 and 1962 papers on natural-sounding artificial reverb — gave engineers a theoretical framework: the Haas fusion zone of approximately 30–40 ms was the critical boundary separating perceptual fusion from perceptual separation, directly informing optimal pre-delay settings for different instruments.
By the 1990s and into the 2000s, software reverb plugins inherited all of these parameters. Waves TrueVerb (1997) and the Altiverb convolution reverb by AudioEase (2002) brought pre-delay into Pro Tools sessions, while freeware and commercial plugins democratized the parameter entirely. The Valhalla DSP lineup, beginning with ValhallRoom in 2010, developed by Sean Costello, brought high-quality algorithmic reverb with precisely calibrated pre-delay into the sub-$50 price range and became ubiquitous in electronic and hip-hop production throughout the 2010s. Today, pre-delay appears in every DAW's stock reverb, every serious third-party reverb plugin, and in modular synthesis environments where it is often patched as a literal delay module feeding into a reverb module — returning the concept to its physical roots with full voltage-controlled flexibility.
Vocals: The most common and consequential use of pre-delay is on lead vocals. A value of 20–60 ms, ideally synced to a musical subdivision, allows a singer's consonants and vowel attacks to land cleanly before the reverb blooms. Engineers working in pop and R&B frequently use a dotted 8th note pre-delay (187.5 ms at 120 BPM) paired with a hall or chamber reverb at 1.5–2.5 s decay — the combination creates the impression of a large space without the vocal sounding distant or washed. For rap and hip-hop vocals, shorter pre-delays (10–25 ms) with shorter decays (0.4–0.8 s) are common, preserving the rhythmic articulation of fast syllables while adding just enough spatial dimension to lift the voice off the dry, flat quality of a close-mic recording.
Drums and Percussion: On snare drums, pre-delay is used to preserve the crack of the attack before room reflections arrive. Values of 5–20 ms are typical; beyond 25 ms, the reverb can begin to sound detached from the hit rather than part of it. The legendary gated reverb sound of the 1980s, achieved with the Lexicon 224 or 480L gated to close before the tail fully decays, often used a moderate pre-delay (10–18 ms) so the initial transient retained its punch before the reverb gate slammed shut. For room reverb on a full drum kit bus, 5–15 ms of pre-delay can make the difference between a kit that sounds wet and recorded versus a kit that sounds like it lives in a convincing, believable room.
Guitars, Synths, and Pads: Sustained instruments benefit from shorter pre-delays than transient-heavy ones, because their attacks are inherently softer and the risk of reverb smear obscuring an attack is lower. On guitar chords and synth pads, 5–20 ms of pre-delay adds space without conspicuous delay, blending the reverb into the sustain phase. For synth leads or guitar solos where note separation matters, push pre-delay to 20–40 ms to keep melodic lines intelligible. On ambient and drone textures, pre-delay can be pushed to 100+ ms for deliberately artificial, layered spatial effects where the 'wrong' pre-delay becomes a compositional texture in its own right.
Bus and Mix Bus: Applying reverb with a short pre-delay (5–20 ms) to the full mix bus or a stem bus is a technique borrowed from mastering engineers and film mixers who want to 'glue' elements into a single acoustic environment. The pre-delay is kept short enough that the reverb onset does not register as a separate event, but long enough that the reverb does not smear transient information on the bus. Most mix engineers working in rock, country, and live-recording genres use this subtly — 10–15 ms pre-delay, 0.4–0.6 s decay, 8–12% wet — to imply the acoustic unity of a live performance without making a studio record sound obviously processed.
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 pre-delay used intentionally, at specific moments, for specific purposes.
The cavernous gated snare reverb that defines this track uses a carefully set pre-delay — estimated at approximately 10–14 ms based on the SSL 4000E console's room microphone setup at Townhouse Studio — that allows the snare's transient crack to register fully before the enormous Lexicon 224 reverb tail explodes outward. Listen on headphones: the snare attack is bone-dry and precise; the reverb arrives as a distinct, cathedral-scale bloom a perceptible moment later. The pre-delay is the architectural reason this reverb sits behind the drum rather than swallowing it.
Greg Kurstin applied a generous pre-delay — widely cited in interviews as synced to the track's 79 BPM, placing it near a 16th-note value of approximately 190 ms — to Adele's lead vocal reverb, creating the distinctive sense that her voice exists in a cavernous hall while still sounding immaculately present. The pre-delay is long enough to hear as a brief ghost of silence before the reverb bloom, especially audible on the sustained vowels at the ends of phrases. This setting is a masterclass in using pre-delay to make a big, theatrical reverb work in a ballad without obscuring diction.
The reverb on Kendrick's vocals in the opening bars is short and tightly controlled — a room or small hall reverb with a pre-delay estimated at 15–20 ms, keeping the delivery sharp and confrontational while adding just enough spatial body to separate the vocal from the dry, minimalist beat. The short pre-delay means the reverb fuses nearly with the attack, creating an aggressive, close-mic'd presence rather than a recessed, dreamy distance. Compare this with the longer pre-delays used on the harmonized sections later in the track, where the vocal is intentionally pushed back in the image.
The acoustic guitar and vocal treatment on this track use long pre-delays — in the range of 50–80 ms — combined with spacious hall reverbs, creating a sense of vast, outdoor distance consistent with the song's lyrical imagery. The pre-delay on the vocal is long enough to register as a perceptible gap before the reverb arrives, giving Justin Vernon's falsetto an ethereal quality where the voice seems to exist outside normal room physics. This is a textbook example of pre-delay used for aesthetic, non-realistic spatial design rather than simply for clarity.
Bruce Swedien's drum sound on Thriller is widely studied. The snare on 'Billie Jean' uses a pre-delay on its room reverb that allows the tight, close-mic'd snare crack to fully declare itself before any spatial information arrives — Swedien has described using precise pre-delay calibration on the Lexicon 224 to ensure the attack remained punchy on any playback system, from the radio to the dance floor. The result is a snare that sounds simultaneously like it was recorded in a world-class room and placed precisely in a dry, groove-locked arrangement.
Pre-delay locked to a musical subdivision (16th note, 8th note, dotted 8th note) of the session tempo. The primary advantage is that reverb reflections align with the rhythmic grid, preventing temporal smear that blurs the groove. This is the most common approach in modern pop, hip-hop, and electronic production, and most current reverb plugins implement it with a single sync toggle.
Pre-delay set to an arbitrary millisecond value without reference to tempo. This approach is preferred in film scoring, ambient music, and orchestral mixing where tempo relationships may not apply, or where the specific psychoacoustic distance implication of a particular ms value (e.g., 32 ms implying a room approximately 11 meters deep) is the design goal. Free-time pre-delay is also used when slightly 'wrong' rhythmic relationships create a desirable looseness or dream-like quality.
In convolution reverb, the IR (impulse response) file may already contain a period of silence before the first recorded reflection — this is the room's natural pre-delay captured at the time of measurement. Plugins like Altiverb and Space Designer allow an additional pre-delay to be dialed on top of this inherent IR pre-delay, or they offer tools to trim the IR's head to remove it. Understanding the distinction between IR-inherent and plugin-added pre-delay is important for accurate convolution reverb calibration.
A pre-delay value that drifts slowly over time according to an LFO or random modulation source. The subtle pitch variation this introduces to the reverb tail — because a changing delay time causes a small Doppler-like pitch shift — adds organic movement to otherwise static algorithmic spaces. This technique is common in lush synth pad treatments and in cinematic sound design where a reverb must evolve over a long cue.
Setting pre-delay to 0 ms or as close as the hardware allows, causing the reverb to begin simultaneously with the direct signal. This fuses the two, creating maximum envelopment at the cost of transient clarity. Zero pre-delay is appropriate for sustained instruments (strings, organ, pads) where the attack is soft enough to survive the merging, and for deliberate vintage or lo-fi aesthetics where the separation of dry and wet signals would sound anachronistically clean.
These MPW articles put pre-delay into practice — specific techniques, real tools, and applied workflows.