What Is Reverb? The Controls, What They Do, and Why Decay Is Not Echo
Data engineer who loves building high-performance data and web-related tools. Creator of SlowedReverbMaker.net, implementing browser-side digital signal processing (DSP) to democratize audio editing.
Reverb is the single most misunderstood effect in music production, and the confusion usually starts with an analogy that sounds helpful and is not: that reverb is a kind of echo.
It is not. An echo is a distinct repetition — you hear a sound, then you hear it again, separated clearly enough that your brain counts two events. Reverb is what happens when the repetitions arrive so densely and so close together that you stop being able to count them at all, and they merge into a continuous wash. The difference is not one of degree but of perception: at some point your auditory system gives up on separating the reflections and starts hearing a space instead.
That distinction matters, because it explains what all the controls are for. Every reverb parameter is a knob on the illusion of a room, and once you know which aspect of a room each one governs, the guesswork disappears.
Try this while you read: Slowed Reverb Generator
Slow songs, add reverb, preview, and export MP3 or WAV. Free, in your browser, no signup.
1. What Actually Happens in a Room
Clap your hands in a hall and three distinct things reach your ears in sequence. Every reverb effect ever built is an attempt to manufacture that sequence, so it is worth having the vocabulary straight.
- Direct sound
- The part that travels straight from the source to your ears without touching anything. It arrives first because it takes the shortest path.
- Early reflections
- The first bounces off the nearest large surfaces — floor, ceiling, closest walls. Still sparse enough to be individually distinct, and their timing tells your brain the size and shape of the space.
- Late reverberation (the tail)
- After enough bounces, reflections arrive thousands of times per second from every direction and no pattern remains. What is left is a decaying cloud of energy, fading as each bounce loses a little to absorption.
- Reverb vs echo
- An echo is a distinct repetition you can count as a separate event. Reverb is what happens when repetitions arrive so densely that your ear stops separating them and hears a space instead. The difference is perceptual, not one of degree.
Why the Early Reflections Matter Most for Realism
The tail is what people notice, but the early reflections carry the information. A reflection arriving very soon after the direct sound means a surface is close; one arriving much later means it is far away.
Your brain reads that pattern and infers the room before you have consciously noticed anything at all. It is why a bathroom and a stairwell sound obviously different even though both are just 'reverberant'.
2. Decay and RT60
Decay is the control people reach for first, and it is usually described as 'how long the reverb lasts'. That is close enough to be useful and slightly wrong in a way worth correcting.
- RT60
- The time it takes for sound to fall by 60 decibels from where it started — a factor of one thousand in amplitude. It is the standard measure of a space's reverberation.
- Decay (the control)
- On most tools, the length of the underlying impulse response. It sets how absorbent the imagined space is, which is what determines how large it feels.
RT60 Is Not the Moment of Silence
Silence never technically arrives. The decay is exponential and approaches zero without reaching it. Sixty decibels is a well-chosen threshold because it is roughly where a decaying sound disappears beneath the background noise of an ordinary room — so RT60 is the moment the tail stops mattering, not the moment it stops existing.
Most of the Action Happens Early
This has a consequence people find counter-intuitive. Measuring the impulse response used by this site's slowed reverb generator, whose envelope falls as a power curve, the level has already dropped 20 dB by about 60% of the way through the tail, and 40 dB by 84% of the way.
The last 40% of the decay covers the final 40 dB — audible if the mix is quiet, and completely buried the moment anything else is playing. This is why long decay settings do far less than expected on dense material and far more on sparse material. The tail did not change; what changed is whether anything was covering it.
3. Pre-Delay: The Control That Rescues Vocals
Pre-delay is the most underused control on any reverb, and the one that most reliably fixes a vocal you can no longer understand.
- Pre-delay
- The gap between the direct sound and the start of the reverb. Physically it corresponds to how far away the nearest reflecting surface is.
Why It Preserves Intelligibility
During the pre-delay gap, the dry sound is heard alone and unobstructed. A vocal with a healthy pre-delay gets its consonants out cleanly before the wash arrives behind it. The same vocal with no pre-delay has reverb starting on top of the syllable that produced it, and clarity suffers immediately.
Stand in a large hall and the first reflection has a long way to travel, so there is a noticeable gap. Stand in a small tiled bathroom and the walls are close enough that reflections arrive almost instantly — which is exactly why bathrooms are unflattering to speech.
4. Wet, Dry, and What They Really Control
The wet/dry mix sets how much processed signal you hear against the unprocessed original. Wet is the reverb; dry is the source.
Wet Is Not a Distance Control
It partly behaves like one — a more reverberant signal reads as further away, because the ratio of direct to reflected energy is one of the documented cues your ear uses to judge distance.
But past a certain point, raising wet stops pushing the sound further away and starts simply making it indistinct. You lose the source without gaining the space.
A more reliable framing: the wet control decides how much of what you hear is information about the room rather than about the performance. At 30% you have a performance in a place. At 70% you have a place with a performance somewhere in it. Both are legitimate — they are just different pieces of music.
If You Want a Bigger Space, Use Decay
This is the single most common mix-up. Wanting something more atmospheric, people push wet to 60% and get a vocal that has disappeared without the room feeling any larger.
A long decay at a modest wet mix gives you the cathedral. A high wet mix at a short decay just gives you less vocal.
Reverb Adds Level — and That Can Distort
Reverb is additional signal summed on top of the dry, so raising the wet mix raises the overall level. Where one note's tail overlaps the next note's attack, those overlaps add and produce peaks higher than either sound alone.
This is a common and unrecognised cause of distortion in heavily reverbed edits. If your track starts crackling after you raise the reverb, the reverb did not distort — the sum ran out of headroom. There is a full explanation of that failure mode if you have hit it.
5. The Two Ways Reverb Gets Built
Every reverb you encounter is made one of two ways, and knowing which you are using explains most of its behaviour.
- Convolution reverb
- Starts from an impulse response — a recording of how a real space answers a sharp burst of sound. Convolving audio with it makes the audio sound as though it were played there. Not a model of a hall; a measurement of one.
- Algorithmic reverb
- Builds the effect from networks of delays and filters feeding back into each other, tuned until the output resembles a decaying room. Nothing is measured — it is a synthetic space.
The Trade-Off
Convolution is more realistic and less flexible: change the decay and you are stretching or truncating a recording, which can sound unnatural. Algorithmic is less literal and endlessly adjustable — you can ask for a six-second decay in a small room, which is not a thing that exists.
This is why most of the famous studio reverbs are algorithmic. They were prized for sounding good, not for sounding like anywhere in particular.
What This Site Uses, and Why
The reverb in these tools is convolution-based but uses a synthesised impulse response rather than a recorded one: a burst of decaying noise shaped by an exponential envelope.
That is a deliberate middle path. It gives the dense, natural tail that convolution produces, while leaving decay as a free parameter rather than a property of some specific room someone once recorded — which matters when the whole point of the tool is a decay slider.
6. Common Mistakes
Most reverb problems are one of a small number of things.
- Using decay to add drama on a fast track. If the tempo is quick, long tails overlap and the result is mud rather than space. Shorten the decay and raise the wet mix instead.
- No pre-delay on vocals. The single most common cause of a reverbed vocal that has lost its words.
- Judging reverb in solo. Reverb that sounds perfect on an isolated vocal is almost always far too much in the full mix, because nothing else was competing for the space.
- Ignoring what the source already has. A recording made in a live room arrives with its own reverberation. Adding more puts a room inside a room, and it rarely sounds like either.
- Forgetting reverb adds level. Long tails and high wet mixes consume headroom, and the overlaps are where clipping starts.
The Short Version
Reverb is not echo, decay is not the moment of silence, and the wet control is not a distance knob. Each parameter governs one property of an imagined space: decay sets how absorbent it is, pre-delay sets how far away the walls are, and wet sets how much of the room you are hearing relative to the performance.
The fastest way to internalise any of this is to move one control at a time on audio you know well. Try it on your own track — set the speed first, then decay, then wet, and listen to what each one changes on its own.
Method
The decay-curve figures are derived from this site's own `createImpulseResponse`, whose envelope falls as (1−t) raised to the power 2.5 across the impulse length. Solving for the −20, −40, and −60 dB points gives 60.2%, 84.2%, and 93.7% of the total length respectively, which is where the RT60-is-94%-of-the-setting figure comes from. RT60 as a −60 dB measure is the standard acoustic definition, not a site-specific one.