reverse reverb: how it works and how to make it
reverse reverb is the swell that grows into a sound instead of trailing behind it. how the trick works, why it sounds so unnatural, and three ways to make it, from the manual render to a built-in mode.
the sound that arrives before the sound
you know the sound even if you have never named it. a vocal is about to come in, and just before the word lands there is a rising wash, a ghostly intake of breath that swells up out of nowhere and then snaps into the dry voice. or a snare is about to hit, and something breathes in underneath it first, pulling your ear toward the beat before the beat is there.
that is reverse reverb, and it does something no real space can do. a normal reverb tail is an aftermath: the sound happens, then the room answers, then the answer fades. reverse reverb flips the aftermath to the front. the tail crescendos into the sound instead of trailing off behind it. the room answers before the question.
here is the thing: this is not a subtle spatial nicety. it is a small act of impossible physics, and your ear catches it instantly.
key takeaway
reverse reverb is a reverb tail played backwards, so it grows louder as it approaches the sound and then cuts to the dry hit. every version of the effect, from a hand-rendered bounce to a built-in mode, is a way of making the tail run in the wrong direction.
how the trick actually works
the mechanism is almost stupidly simple once you see it, and it is the same whether you do it by hand or a plugin does it for you: you reverse, you reverberate, you reverse again.
start with a dry clip. flip it so it plays backwards. now print a reverb onto that backwards clip, exactly as you always would, so the tail trails off after the (reversed) sound. then take that whole rendered result, reverb and all, and flip it back to forward.
the double flip is the whole magic. flipping twice returns the source to playing forward, so the word or the snare sounds normal again. but the reverb tail was added in between the two flips, pointed the normal way, so the second flip leaves it running backwards. a tail that was decaying after the sound is now swelling into it.
notice what the swell is made of: it is the reverb of the sound, so it is spectrally the sound’s own reflection, only rising and time-smeared. that is why a reverse swell on a bright vocal shimmers and a reverse swell on a dark pad glowers. you are not adding a generic riser. you are hearing the sound’s own tail, walking toward you backwards.
why it sounds unnatural, in the good way
there is a real perceptual reason reverse reverb feels eerie, and it is worth understanding because it tells you where the effect earns its keep.
almost every sound in the physical world shares one shape: a fast attack and a slower decay. a string is plucked, a stick strikes a drum, a glass is tapped, and in each case energy appears suddenly and then bleeds away as the object gives up its motion. your hearing has spent your whole life inside that regularity, and it has drawn a conclusion from it: the sharp part is the event, the fading part is the room. researchers have shown that this asymmetry is not just a description, it is something the ear actively uses. a tone that fades out (“damped”) and the exact same tone reversed so it fades in (“ramped”) are perceptibly different sounds even though they carry identical energy.[^1]
the tail your ear throws away
this goes deeper than “they sound different.” in a set of loudness experiments, sounds that swelled in (slow attack, sharp cutoff) were judged louder than their time-reversed twins of equal energy, and the authors proposed a striking explanation: the ear parses incoming sound into a direct part and a reverberant part, and it discounts a slowly decaying tail because it treats that tail as room reflections rather than as the source itself.[^2] a normal decay gets quietly written off as “that is just the space.” reverse reverb refuses that trick. its tail arrives on the rising edge, where the ear expects the source, so the brain cannot file it as room, cannot discount it, and shoves it into the foreground. the effect you hear as anticipation is your auditory system failing to explain away a tail it usually ignores.
and there is a blunter violation underneath all of it. in reverse reverb, the loudest part of the reflected energy arrives before the sound that caused it. cause after effect. nothing you have ever heard in a real room does this, because sound cannot precede its own source, and your brain treats that impossibility the way it treats a face in a photo negative: recognisable, wrong, and impossible to un-see. that is the whole emotional payload of the effect. use it where you want unease, suction, or the feeling of being pulled toward a moment. do not reach for it when you just want a sense of space, because it will never sit quietly in the background. it is an event, not an ambience.
making it yourself, three ways
the manual render (works in any DAW, sounds the best, takes the longest). this is the technique that has been on records since tape, and it still gives the most controlled result because you are matching the swell to one specific printed phrase.
note
the four-minute version, in any DAW: (1) isolate the clip you want the swell to lead into, a single vocal phrase or a snare, on its own. (2) reverse it (most DAWs have a reverse audio function, or an offline “reverse” process). (3) put a reverb on that track, wet-heavy, decay long enough to make the swell you want, and bounce or freeze the reversed clip with the reverb printed in. (4) reverse that bounced file back to forward. (5) slide it so the swell ends right as the original dry sound begins, then blend the dry sound back in on top. the swell tail should crash into the dry hit and stop. nudge the timing by a few milliseconds until the landing feels like a landing.
the tedium is real, which is why the effect used to feel like a special occasion. every phrase you want it on is its own reverse-render-reverse chore, and if you change the vocal take you do the whole dance again. (i have absolutely bounced the wrong region at 1am and wondered why the swell was leading into silence.)
the reverse-gate shortcut (fast, retro, narrower). some reverbs offer a “reverse” or “gate” program that fakes a rising envelope in real time without the render. it is the classic 80s trick, instant and automatable, but it tends toward a specific hard, synthetic sound and can stumble on dense polyphonic material where it cannot tell where one event ends and the next begins. good for a stylised build, less good for a delicate vocal swell.
a real-time reverse mode (automatable, matched to the mix). the newer approach plays the wet tail backwards continuously, inside the reverb, so there is nothing to bounce. you get the swell as a mode you can switch on, automate, and A/B, and you can move the swell length live. the honest trade is that a live effect has to swell across a fixed window rather than being hand-matched to one printed phrase, so for a single hero moment the manual render can still edge it. for everything else, the ability to just turn it on and ride it wins.
this last approach is the one i built into OPEN, the algorithmic reverb for synths and drum machines that is in beta now. its reverse mode swaps the wet tail for a backwards, equal-power version of itself, and because OPEN’s mix-aware carving keeps shaping the tail underneath, the reverse swell is still reacting to how busy the track is. if you want first access, tick “i want to beta test future plugins” on the signup at /#signup.
frequently asked questions
frequently asked questions
what is reverse reverb?
reverse reverb is a reverb tail played backwards so it swells up into the sound instead of decaying away after it. you get a rising wash that crescendos and then cuts to the dry hit, most often heard as a ghostly intake of breath before a vocal line or a build into a snare. it is sometimes called pre-verb because the reverb arrives before the source instead of after it.
how do you make a reverse reverb?
the classic method has three steps: reverse the audio clip so it plays backwards, print a reverb onto the reversed clip so the tail trails off the (now backwards) sound, then reverse that whole rendered result back to forward. because you flipped it twice, the source plays forward again but its reverb tail is now running backwards, swelling into the note. many reverbs also now include a reverse mode that does this in one control without the render dance.
why does reverse reverb sound so haunting?
because it breaks a rule your hearing treats as physics. in the real world energy always arrives with the source and decays after it, so your brain learns to expect a sharp attack followed by a fading tail and to file that tail as the room. reverse reverb hands you the opposite: a tail that grows louder as it approaches the sound, energy that precedes its own cause. nothing in nature does that, so your ear cannot file the swell as a room, and it reads as anticipation, suction, or dread.
what is reverse reverb good for?
transitions and entrances. a swell into a vocal phrase, a build into a downbeat or a snare, a riser under a scene change, a lead-in to a drop. it works anywhere you want the ear pulled toward a moment before that moment lands. it is less useful as a general sense of space, because the effect is an event, not an ambience.
can you do reverse reverb without rendering the audio?
yes, if the reverb has a built-in reverse mode. a reverse mode plays the wet tail backwards internally, in real time, so you can automate it, A/B it, and change the swell length without bouncing anything. the trade is that a live reverse tail has to guess where the sound is going, so it swells across a fixed window rather than being perfectly matched to one printed phrase the way a manual render is.
references
a note from the developer
the thing that surprised me building a reverse mode was how unforgiving the swell is about smoothness. a normal reverb tail can wobble a little and you will never notice, because your ear is already discounting it as room. a rising swell is the opposite: it is on the foreground edge where the ear is paying full attention, so every seam shows. my first version had two playback heads sweeping the tail backwards, and after a length change they drifted a hair out of alignment and the swell dipped about five decibels in the middle, a little hole right where the ear was leaning in hardest. no measurement flagged it as broken. it just felt like the swell stumbled.
the fix was to make the two heads sum to constant power no matter how they drift, so the rise is always monotonic. that is the lesson reverse reverb taught me twice over: the effect lives entirely on the attack, and the attack is where you cannot hide anything.
if you make something eerie with it, or you have a swell that will not sit right, send it over. jonas@kernaudio.io. i read every email.