What is reverb decay time (RT60)?

What does decay mean on a reverb?

When a sound stops in a real room, its reverb carries on. The reflections keep bouncing between the walls, losing some energy at every surface and to the air in between, so the wash fades out over a fraction of a second in a small room and over several seconds in a big stone church. The decay control on a reverb sets how long that fade takes.

Plug-ins label it in different ways. Ableton Live's Reverb calls it Decay Time and Logic's ChromaVerb calls it Decay, while others say Time, Length or Reverb Time; nearly all of them mean the same thing, given in seconds. It's the control that changes a reverb the most, because the length of the tail decides what happens in the gaps between notes, and the gaps are where you hear reverb.

This page is about decay alone. For the other controls and the different kinds of reverb, start with what is reverb?

What is RT60?

RT60 is the standard way to put a number on decay. It's the time, in seconds, a sound takes to fall by 60 dB once the source stops. Set a reverb's decay to 2 seconds and that's its RT60: two seconds after the last note, the tail is 60 dB quieter than it was when the note ended.

Sixty decibels is a big drop. It's a millionth of the starting energy, or a thousandth of the amplitude, which takes a loud sound down to around the edge of hearing in a quiet room. The idea goes back to Wallace Clement Sabine, a Harvard physicist who was asked in the 1890s to fix the lecture hall in the university's Fogg Art Museum, where speech was hard to follow. He timed how long the sound of an organ pipe took to become inaudible and found that the time depended on two things, the volume of the room and how much its surfaces absorbed. Boston's Symphony Hall, which opened in 1900, was the first concert hall designed with his results.

Measured in decibels, a reverb tail falls in a straight line, so the RT60 also tells you the rate. A 2 second decay drops 30 dB every second: it's 15 dB down after half a second and 30 dB down after one. That's why the impulse response screen in the lab draws the tail as a slope, and why turning the decay knob swings the slope flatter or steeper.

Sabine's formula

Sabine's formula is simple enough to work out by hand. RT60 = 0.161 × V ÷ A, where V is the room's volume in cubic metres and A is its total absorption: each surface's area multiplied by how much sound it soaks up, from 0 for a perfect reflector to 1 for an open window.

Say you have a room 5 metres long, 4 wide and 2.5 high. Its volume is 50 m³ and its surfaces add up to 85 m². If they absorb 20% of the sound that hits them on average, A is 17 m², and the RT60 comes out at 0.161 × 50 ÷ 17, or 0.47 seconds. Double every dimension and the volume goes up eight times while the surface area only goes up four, so the decay doubles to about 0.95 seconds. Bigger rooms ring for longer because sound travels further between surfaces, so it hits fewer of them each second and loses energy more slowly.

The formula assumes the sound spreads evenly through the room and the surfaces don't absorb much. In very dead rooms it overestimates, and acousticians switch to Carl Eyring's version from 1930, which handles high absorption better.

How is RT60 measured?

Catching a full 60 dB of decay is harder than it sounds. A real room has background noise from air conditioning, traffic and the building itself, and a test signal often gets only 40 or 50 dB above it. The tail sinks into the noise before it reaches −60 dB.

So acousticians measure part of the fall and extend it. T30 times the drop from 5 dB down to 35 dB down and doubles it, and T20 times −5 to −25 dB and multiplies by three. Both assume the slope stays the same all the way down. They're defined in ISO 3382, the standard for measuring room acoustics, alongside EDT, the early decay time, which times the first 10 dB and multiplies by six. EDT tends to match how reverberant a space feels, because the loud start of the tail is the part you hear most.

The raw decay of a room is ragged, so it's smoothed first. In 1965 Manfred Schroeder showed that you can get a clean decay curve from a single impulse response by adding up its energy backwards from the end, and that backward integration is still the usual method. The lab measures its own tails the same way. Every time you move a control it renders one click through the reverb you're hearing, integrates it backwards, times the fall from −5 to −35 dB and doubles it.

Reading decay in the lab

In the lab at the top of this page, the controls that only shape the start of the reverb (pre-delay, early reflections and size) are locked. Everything left changes how long the tail lasts or how loud it is.

Keep an eye on the impulse response screen. The dashed line marked 60 dB DOWN is where the measured tail has fallen 60 dB, and the Tail figure in the Right now panel gives the same time. On the hall it's usually a little longer than the decay setting, because the lab's hall lets its lows ring on longer than its mids, as real halls do.

On the big screen, the colour that carries on after each hit is the tail. With the 120 BPM loops, the Right now panel also counts how many beats the tail lasts, which is the most useful way to think about decay once it's in a song.

Typical decay times for real spaces

These are rough mid-frequency figures. Reverb presets copy the same spaces, which is why a room preset is usually under a second and a hall preset around two.

In a mix, under a second reads as a room and 1.5 to 3 seconds as a hall or a plate. Past about 4 seconds the reverb stops sounding like a place you could stand in and becomes a texture of its own.

How long should reverb decay be?

It should be long enough to hear in the gaps and short enough to have mostly gone when the next important sound arrives. Tempo decides most of it, because tempo sets how long the gaps are. A beat lasts 60,000 ÷ BPM milliseconds, so at 120 BPM it's 500 ms, a half note is 1 second and a bar of 4/4 is 2 seconds.

A good starting point is to set the decay to a note length: a beat or two for drums, a half note to a bar for a vocal, a bar or more for pads. With the decay set to a half note, the tail is already 30 dB down after one beat, which in a full mix is usually enough to keep it out of the way of the next hit.

How busy the part is matters as much as the tempo. A slow ballad with long held notes has few gaps, so a 3 second hall just sounds big. A drum loop at the same tempo has a gap after every hit, and the same 3 seconds turns it into a wash, as you heard in the lab at 6 seconds.

Setting the decay, step by step

1. Start shortPut the reverb on a send at 100% wet with the decay around 1 second, then bring the send up until you can clearly hear it.
2. Play the busiest sectionDecay problems show up in the fastest, densest part of the song, so judge it there.
3. Lengthen it until the gaps fillRaise the decay until the tail carries through the gaps the way you want. On a vocal, listen to the ends of lines.
4. Back off when the rhythm blursIf the next hit or word lands in a loud tail and the groove gets harder to follow, shorten the decay until it clears.
5. Listen with the mix playingThe other parts hide the end of the tail, so you may want a little more length than you'd choose in solo.

Decay at high and low frequencies

In real rooms, different frequencies die away at different rates. Soft furnishings, people and the air itself soak up treble faster than bass, so the top end of a tail goes first. Concert halls are often designed so the low end decays a little more slowly than the mids, which listeners describe as warmth.

Reverbs copy this with damping, which some plug-ins call high damp, HF decay or treble decay. In the lab, damping shortens the bands above 1 kHz, and the impulse response draws the top end on its own as a pale line. At 85% damping that line falls far more steeply than the tail as a whole, and the reverb sounds darker and shorter while the decay knob stays exactly where it was.

The low end causes the opposite problem. A long decay on bass notes and kick drums piles up into mud, because each low note's tail is still going when the next one starts. You can keep the length and lose the mud by filtering the reverb: a low cut at 200 to 500 Hz on the reverb return takes the bass out of the tail and leaves the dry parts alone. Some reverbs have a bass multiplier, which sets the low-frequency decay as a multiple of the main one; set it below 1x and it does the same job from inside the reverb.

Decay time vs room size

Size and decay are separate controls on most reverbs. Size sets how far apart the walls are, which spreads out the early reflections and slows down how quickly the tail builds to full density. Decay sets how long the tail takes to fade, and in a real room that depends on how absorbent the surfaces are as well as how big the room is.

Big rooms usually ring longer, as Sabine's formula shows, so plug-ins that model real spaces often lengthen the decay when you raise the size. Plenty of real places break the pattern, though. A small tiled bathroom can ring longer than a large living room full of sofas, and a big hall hung with heavy curtains can be surprisingly dead. A large size with a short decay gives you the spread-out start of a big space and a quick finish, which works well on a vocal in a fast song.

Plates and springs have no walls at all, and they still have a decay. The EMT 140 plate set its decay with a porous damping pad that could be moved closer to the steel sheet, so the plate lost energy faster. In the lab, pick the plate card and the decay knob sets its RT60 directly. The plate vs hall vs room vs spring comparison covers how the types differ apart from their length.

Very long and gated decays

Pads, ambient textures and sound design use decays far past 5 seconds on purpose, often with heavy damping so the long tail stays dark. Many reverbs go to 20 seconds or more, and some can hold the tail indefinitely; Ableton Live's Reverb has a Freeze button that does exactly that.

Going the other way, a gate can cut a long tail short. The big drum sound on Phil Collins's “In the Air Tonight” came from room microphones compressed hard and then cut off by a noise gate, so the room's sound stays loud for a fraction of a second and then stops dead, which no natural decay does.

Common decay mistakes

Setting the decay in solo

With the track soloed you can hear the whole tail down to the noise floor, so every decay sounds long. Bring the mix back in and the last 20 or 30 dB of it disappears under the other parts. Set the decay with everything playing.

Lengthening the decay to make the reverb louder

A longer decay keeps the tail audible for longer, but so does a louder send, and the two are easy to mix up. Turn the reverb up 10 dB and it stays above any given level for an extra sixth of its RT60, which is 0.33 seconds on a 2 second tail. If the space sounds the right length and just too quiet, turn up the send and leave the decay where it is.

A long tail full of bass

Before you shorten a muddy reverb, cut the lows from it. Quite often the length was fine and the bass in the tail was the problem.

One decay for everything

Drums usually want a much shorter tail than vocals or pads. Two sends, a short room for the drums and a longer plate or hall for the rest, will cover most songs.

Training your ears to hear decay

The quiz under the lab plays a hidden hall with either 0.6 or 3.5 seconds of decay and asks which one you heard. The controls and screens are covered while it plays, and both versions are matched for loudness. Listen to the gaps after each sound stops. Once that's easy, close your eyes, set the decay knob by ear to a length you'd like, then check the Tail figure to see how close you got.

Questions people ask

Keep going