Reverberation Time (RT60 and EDT)
Reverberation time is the headline measure of how long a room keeps ringing after a sound stops.
When a sound source switches off, the energy already in the room does not vanish — it keeps bouncing between surfaces, losing a little at every reflection, until it fades into the noise floor. How long that takes is what a room sounds like: a treated vocal booth dies away almost instantly, a stone cathedral washes on for many seconds.
RT60: the 60 dB decay
RT60 is the time, in seconds, for the reverberant tail to drop by 60 decibels — a factor of one million in energy — after the source stops. Sixty dB is the historical definition because it spans roughly the range from a loud sound down to inaudibility in a quiet hall.
In practice the tail rarely stays clean across a full 60 dB before noise swamps it, so RT60 is almost always extrapolated. You measure the slope of the decay over a cleaner, smaller range and scale it up to 60 dB:
| Measure | Decay range measured | Scaled to |
|---|---|---|
| T20 | −5 dB to −25 dB | ×3 → 60 dB |
| T30 | −5 dB to −35 dB | ×2 → 60 dB |
| EDT | 0 dB to −10 dB | ×6 → 60 dB |
T20 and T30 are the same RT60 number measured over different spans: T30 uses more of the decay and is the usual reported value, while T20 falls back to a shorter span when noise swamps the tail early. EDT is the odd one out — it measures only the first 10 dB and answers a different question (below).
The first few dB are skipped because the very start of the decay is dominated by the direct sound and the strongest early reflections, which are not yet "reverberant."
EDT: what the ear actually hears
Early decay time (EDT) measures only the first 10 dB of decay and scales that slope up to 60 dB. It weights the beginning of the tail — exactly the part the ear uses to judge how live a space feels.
EDT and RT60 agree in an ideal, perfectly diffuse room. They diverge in real ones:
- EDT much shorter than RT60 — strong early reflections steepen the start of the decay, so the room sounds tighter than its long late tail suggests. Common in rooms with a reflective near field but absorptive far surfaces.
- EDT close to RT60 — a smooth, well-diffused decay; the perceived reverberance matches the measured tail.
Because perception tracks EDT, two rooms with identical RT60 can feel quite different. EDT is the better predictor of "reverberance"; RT60 is the better predictor of how long a sustained note will smear into the next.
Reading the numbers
| RT60 | Feel |
|---|---|
| < 0.3 s | Dead — booths, treated control rooms |
| 0.3–0.6 s | Tight — bedrooms, small studios |
| 0.6–1.2 s | Lively — live rooms, large lounges |
| 1.2–2.5 s | Reverberant — concert halls |
| > 2.5 s | Cavernous — cathedrals, large stone spaces |
RT60 scales with room volume and falls with absorption: a big, hard-surfaced space rings long; a small, soft-furnished one does not. That relationship is what lets libsonare estimate an equivalent room from a decay alone.
How libsonare measures decay time
libsonare derives reverberation time from the energy decay curve (the Schroeder backward integral of the squared impulse response). It fits a line to the decay over the T20/T30 region, rejects segments where the curve is not linear enough, and extrapolates the slope to 60 dB; EDT is fit over the first 10 dB. Per-band reverberation time is computed by running the same fit on octave-band-filtered versions of the response, which is what the per-band decay table shows. When the input is ordinary music rather than a clean impulse, the decay is recovered blindly from the signal's gaps and offsets, and the confidence score reflects how clean that recovered decay was.
Related: Clarity and Definition (C50, C80, D50), Per-Band Decay and Absorption, Inverse Room Estimation, Acoustic Analysis