RT60 Complete Reference — History, Physics, Measurement, Standards, and Optimal Values

126 years after Wallace Clement Sabine published the first quantitative relationship between room volume, surface absorption, and reverberation time, RT60 remains the single most important parameter in architectural acoustics. It is measured in every acoustic survey, predicted in every design model, regulated by every building code, and misunderstood by a remarkable proportion of the professionals who specify it. In 2024, a study by Acoustic Bulletin found that 42% of architects surveyed could not correctly define RT60, despite it appearing in every acoustic specification they signed.

This reference covers the complete story of RT60: from Sabine's original experiments in the Sanders Theatre at Harvard in 1895, through the physics of sound decay in enclosed spaces, the three principal prediction equations, the measurement methodology defined by ISO 3382-2:2008, the practical distinction between T20, T30, and EDT, the optimal values for every building type in common use, and a catalogue of famous rooms with their published acoustic measurements.

Part 1: History — Sabine and the Birth of Architectural Acoustics

Wallace Clement Sabine was a 27-year-old physics instructor at Harvard when, in 1895, the university president asked him to investigate why the newly completed Fogg Art Museum lecture hall was acoustically unusable. Lectures were unintelligible. Speech echoed for several seconds after the speaker stopped. The room, designed to impress visually, was an acoustic catastrophe.

Sabine had no precedents to draw on. There was no science of room acoustics. No measurement instruments. No theory connecting room geometry to acoustic behaviour. He began with a stopwatch, an organ pipe, and his ears. Over the next five years, working after midnight when the campus was quiet, he systematically varied the absorption in rooms by adding and removing seat cushions from the Sanders Theatre. He discovered that the time for sound to become inaudible was directly proportional to the room volume and inversely proportional to the total surface absorption.

In 1900, he published the equation that bears his name:

RT60 = 0.161 × V / A

where V is the room volume in cubic metres, and A is the total absorption in metric sabins (m²), defined as the sum of each surface area multiplied by its absorption coefficient. The constant 0.161 is derived from the speed of sound (343 m/s at 20°C) and the natural logarithm base, specifically 24 × ln(10) / c = 0.1611 s/m.

Sabine went on to design the acoustics of Boston Symphony Hall (1900), which remains one of the finest concert halls in the world with a measured RT60 of 1.9 seconds at mid-frequencies when fully occupied. He is universally regarded as the founder of architectural acoustics.

Part 2: The Physics of Sound Decay

Exponential Decay

When a steady-state sound source in an enclosed room is abruptly turned off, the sound pressure level does not drop instantaneously. Sound energy continues to reflect between the room's surfaces, losing a fraction of its energy at each reflection. The result is an exponential decay of sound energy over time.

The sound pressure level at time t after the source stops is:

L(t) = L₀ − (60/RT60) × t

where L₀ is the initial steady-state SPL. This linear decrease in dB corresponds to an exponential decrease in sound pressure (and a faster exponential decrease in sound energy). The 60 dB decay convention was chosen because it corresponds to the range from a moderately loud sound (~90 dBA) to the threshold of ambient background noise (~30 dBA) in a typical room.

The Diffuse Field Assumption

Both the Sabine and Eyring equations assume a diffuse sound field — a condition where sound energy density is uniform throughout the room and sound arrives at any point equally from all directions. This assumption is valid when:

1. The room is sufficiently large relative to the wavelength of interest (room dimension > 3× wavelength at the lowest frequency)
2. The room geometry is irregular enough to prevent systematic focusing or modal resonance
3. Absorption is distributed reasonably uniformly across surfaces

Frequency Dependence

RT60 is frequency-dependent. Most materials absorb differently at different frequencies, and air absorption becomes significant above 2000 Hz. ISO 3382-2:2008 §3 requires RT60 to be measured and reported at six octave bands: 125, 250, 500, 1000, 2000, and 4000 Hz. Some standards (BB93, ANSI S12.60) use a mid-frequency average (500, 1000, 2000 Hz) for compliance checking, but this conceals the frequency-dependent behaviour that often determines whether a room sounds good or merely passes a number.

A well-designed concert hall has longer RT60 at low frequencies (bass warmth) than at high frequencies (clarity). The ratio T(125 Hz) / T(500 Hz) of 1.1 to 1.3 is considered ideal for orchestral music (per Beranek, 2004). Conversely, a classroom should have uniform RT60 across frequencies — excessive bass reverberation degrades speech intelligibility even when the mid-frequency average meets the target.

Part 3: The Three Prediction Equations

The Sabine Equation (1900)

RT60 = 0.161 × V / A

where A = Σ(Sᵢ × αᵢ) + 4mV

• Sᵢ = area of surface i (m²)
• αᵢ = absorption coefficient of surface i (dimensionless, 0–1)
• m = air absorption coefficient (Np/m), significant above 2000 Hz
• V = room volume (m³)

Limitation: Sabine overestimates RT60 when ᾱ is high. In the extreme case where every surface is perfectly absorptive (ᾱ = 1), the Sabine equation predicts a finite RT60 = 0.161V/S. This is physically wrong — a room with 100% absorption would have no reflections and no reverberation.

The Eyring Equation (1930)

Carl F. Eyring proposed a correction that accounts for the mean free path between reflections:

RT60 = 0.161 × V / (−S × ln(1 − ᾱ) + 4mV)

where ᾱ = A/S is the average absorption coefficient and ln is the natural logarithm.

Validity: More accurate than Sabine when ᾱ > 0.30 — common in rooms with acoustic ceilings, recording studios, and heavily treated spaces. When ᾱ approaches 1.0, Eyring correctly predicts RT60 → 0.

Worked Example — Comparing Sabine and Eyring:

Consider a 10 m × 8 m × 3 m conference room (V = 240 m³, S = 268 m²) with a Class A acoustic ceiling (NRC 0.90), carpeted floor, and painted plasterboard walls.

Average absorption coefficient: ᾱ = 89.4 / 268 = 0.334

Sabine: RT60 = 0.161 × 240 / 89.4 = 0.432 seconds

Eyring: RT60 = 0.161 × 240 / (−268 × ln(1 − 0.334)) = 0.161 × 240 / (−268 × (−0.406)) = 38.64 / 108.9 = 0.355 seconds

The Sabine equation overestimates by 22%. In a compliance context, this difference matters: a room targeting 0.40 seconds would pass under Sabine but fail under Eyring if the actual construction delivers the Eyring-predicted value.

The Fitzroy Equation (1959)

Daniel Fitzroy proposed a formula for rooms with non-uniform absorption distribution — typically rooms where one surface (the ceiling) is highly absorptive while others (walls, floor) are reflective:

RT60 = 0.161 × V / S² × (Sx²/(-Sx × ln(1 − ᾱx)) + Sy²/(-Sy × ln(1 − ᾱy)) + Sz²/(-Sz × ln(1 − ᾱz)))

where the subscripts x, y, z refer to the three pairs of parallel surfaces (floor/ceiling, end walls, side walls), and S is total surface area.

When to use Fitzroy: When absorption is concentrated on one surface pair. The classic case is a room with an acoustic ceiling (ᾱ = 0.90) and hard walls and floor (ᾱ = 0.05). Sabine and Eyring average the absorption and underestimate RT60 in this configuration because the non-absorptive surfaces sustain reflections along the horizontal plane. Fitzroy captures this directional imbalance.

Practical recommendation: Use Sabine for initial estimates (ᾱ < 0.20). Use Eyring when ᾱ > 0.20, particularly for treated rooms. Consider Fitzroy when absorption is highly non-uniform (one surface pair has ᾱ > 0.60 while another has ᾱ < 0.15). For critical projects, use all three and take the Eyring result as the design basis.

Part 4: T20, T30, and EDT — The Three Measures of Reverberation

ISO 3382-2:2008 §3 defines three reverberation time parameters, each derived from a different segment of the sound decay curve.

T30 (Reverberation Time from -5 dB to -35 dB)

The standard measure of reverberation time. The decay curve is fitted with a linear regression from -5 dB to -35 dB below the initial peak level, and the slope is extrapolated to -60 dB. The -5 dB offset excludes the direct sound and earliest reflections; the -35 dB lower limit ensures the measurement stays above the background noise floor.

T30 requires a dynamic range of at least 45 dB (35 dB evaluation range + 10 dB safety margin per ISO 3382-2 §5). This is achievable with a balloon pop or starter pistol in most rooms.

T20 (Reverberation Time from -5 dB to -25 dB)

A shortened version of T30 that uses only the first 20 dB of usable decay. T20 requires less dynamic range (35 dB minimum) and is used when background noise levels are high or the source level is limited (e.g., mobile phone measurements). T20 values are typically within 5% of T30 in well-diffused rooms.

EDT (Early Decay Time)

EDT uses the first 10 dB of decay (0 to -10 dB) and extrapolates to -60 dB. EDT is the parameter most strongly correlated with perceived reverberance — it captures the initial decay that the human auditory system uses to judge how "live" or "dead" a room sounds.

In a perfectly diffuse room, EDT = T30. In real rooms, EDT is often shorter than T30 because early reflections from nearby surfaces create a faster initial decay rate. A room with EDT/T30 < 1.0 sounds drier and more intimate than its T30 would suggest — a desirable quality in concert halls where "warmth without blur" is the goal.

Part 5: Measurement Methods per ISO 3382-2

Method 1: Interrupted Noise

The classical method. A loudspeaker (omnidirectional or dodecahedral) is driven with broadband noise (pink noise is preferred). The noise is sustained until steady-state conditions are established (at least 3 × expected RT60), then abruptly switched off. The resulting decay curve is captured by a measurement microphone and analysed in octave or third-octave bands. The process is repeated at multiple source-receiver positions and the results averaged.

Advantages: Robust, well-understood, works in high background noise because the source level can be raised. Disadvantages: Requires substantial equipment (amplifier, loudspeaker, measurement system), multiple measurements to average out statistical fluctuation.

Method 2: Integrated Impulse Response (Schroeder Method)

Based on Schroeder's 1965 insight that the backward-integrated squared impulse response produces the same ensemble-averaged decay curve as the interrupted noise method. The room impulse response (RIR) is obtained using either:

• An impulsive source (balloon pop, starter pistol, clapper board)
• A swept sine signal (exponential sine sweep, ESS) deconvolved to extract the impulse response

E(t) = ∫[t to ∞] h²(τ) dτ

where h(t) is the impulse response. The resulting curve E(t) is a monotonically decreasing function whose slope gives the reverberation time.

Advantages: A single measurement yields the complete decay curve. Swept sine provides excellent signal-to-noise ratio (80+ dB). Modern software (DIRAC, REW, EASERA, AcousPlan) automates the entire process. Disadvantages: Impulsive sources are not repeatable; swept sine requires equipment.

Measurement Positions per ISO 3382-2 §5

Source positions should be typical for the room use (lectern in a classroom, conductor position in a concert hall). Receiver positions should cover the audience area uniformly, with minimum 2 m separation between positions and minimum 1 m from any surface. Receiver positions within the critical distance (where direct and reverberant energy are equal) should be avoided, as they yield artificially short EDT values.

Part 6: Optimal RT60 Values by Room Type

The following table represents the consolidated best-practice targets from international standards and leading acoustic design references (Beranek 2004, Long 2014, Barron 2009).

Part 7: Famous Rooms and Their Measured RT60

Note: The "best" concert halls (Musikverein, Boston, Concertgebouw) share common characteristics: rectangular shoe-box geometry, narrow width (< 25 m), occupied RT60 of 1.8–2.0 seconds, high ceiling (17–19 m), and diffuse wall surfaces. Beranek's ranking of the world's concert halls (2004) found that RT60 alone does not predict subjective quality — EDT, lateral fraction (LF), and early decay time uniformity are equally important.

Part 8: Common RT60 Errors and How to Avoid Them

Error 1: Using Sabine when Eyring is required. If your room has an acoustic ceiling with NRC > 0.85, the average absorption coefficient likely exceeds 0.30 and the Sabine equation will overestimate RT60 by 15–25%. Use the Eyring equation for design verification.

Error 2: Ignoring air absorption above 2000 Hz. In large rooms (V > 5,000 m³), air absorption at 4000 Hz can add 0.5–1.0 seconds to the predicted RT60 difference between high-frequency and mid-frequency bands. The air absorption term 4mV in both Sabine and Eyring must be included for the 2000 Hz and 4000 Hz octave bands. The coefficient m depends on temperature and relative humidity per ISO 9613-1.

Error 3: Measuring in the wrong condition. RT60 varies significantly between unoccupied and occupied conditions. Audience absorption at mid-frequencies is approximately 0.85–0.95 m² Sabine per person. A 500-seat concert hall gains approximately 425–475 m² Sabine when full, reducing RT60 by 0.3–0.6 seconds. Standards specify whether targets are occupied or unoccupied — check the standard before comparing measurements.

Error 4: Averaging across frequencies when the standard requires octave-band compliance. DIN 18041 requires RT60 verification at each octave band independently. A room can pass the mid-frequency average target while failing at 125 Hz if low-frequency absorption is insufficient. Always check whether the applicable standard accepts a mid-frequency average or requires octave-band compliance.

Error 5: Neglecting furniture and occupants in predictions. An unoccupied conference room with 20 upholstered chairs, a large table, and curtains has significantly more absorption than the bare room. Ignoring this absorption leads to over-treatment. Standard furniture absorption values (per ISO 354 object absorption tests) should be included in predictions: upholstered chair ≈ 0.25–0.45 m² Sabine per unit, table ≈ 0.10–0.20 m² Sabine per m² surface.

Related Reading:

• You're Calculating RT60 Wrong — Sabine vs Eyring — when the Sabine equation fails and what to use instead
• The Reverberation Time Formula: A Complete Derivation — mathematical derivation from first principles
• The 125 Hz Problem Nobody Treats — why low-frequency reverberation persists when the mid-frequency target is met