RT60 Measurement: T20 vs T30 vs EDT and When Each Matters

The Energy Decay Curve

All reverberation time measurements start with the Energy Decay Curve (EDC), computed by Schroeder backward integration of the squared impulse response:

EDC(t) = 10 log₁₀(∫_t^∞ h²(τ) dτ / ∫₀^∞ h²(τ) dτ)

This is equivalent to averaging an infinite number of individual decay measurements (Schroeder, 1965). The EDC starts at 0 dB and decays monotonically to -∞ dB (in theory). In practice, noise limits the useful range.

The SonaVyx RT60 tool displays the EDC per octave band with the evaluation ranges for T20, T30, and EDT clearly marked.

T20: The Practical Choice

T20 evaluates the decay rate between -5 dB and -25 dB on the EDC. The -5 dB starting point avoids the direct sound and early reflections. The -25 dB end point keeps the evaluation well above typical noise floors.

INR requirement: 35 dB minimum (impulse response noise ratio — the ratio between the peak IR level and the noise floor). With a log sweep excitation in SonaVyx, achieving 35 dB INR is straightforward in most rooms.

When to use T20:

• Noisy environments where achieving 45 dB INR is difficult
• Quick surveys and initial assessments
• ISO 3382-2 (ordinary rooms) where T20 is acceptable
• Building code compliance checks where the code specifies T20

The reverberation time is extrapolated from the 20 dB evaluation range to a 60 dB decay: T20 = 3 × (time for -5 to -25 dB decay).

T30: The Precision Standard

T30 evaluates from -5 to -35 dB — a 30 dB evaluation range extrapolated to 60 dB: T30 = 2 × (time for -5 to -35 dB decay).

INR requirement: 45 dB minimum. This requires a strong excitation signal (long sweep, high output) and low background noise. In occupied spaces or venues near roads, achieving 45 dB INR may require measurements during quiet periods.

When to use T30:

• ISO 3382-1 performance space measurements (formal requirement)
• Commissioning reports for concert halls, theatres, and performance venues
• When T20 and T30 disagree by more than 10% (indicating non-exponential decay)

EDT: The Perceptual Metric

EDT (Early Decay Time) evaluates from 0 to -10 dB on the EDC. It captures the initial decay behavior — the part of the reverberation that human hearing is most sensitive to.

Why EDT matters: In a room with strong early reflections followed by rapid late decay, T30 might be 1.5 s while EDT is 0.9 s. The room sounds like 0.9 s to the listener. Conversely, a room with absorptive early conditions (wide open space, no nearby surfaces) but a long late tail might have EDT > T30.

EDT is more variable across measurement positions than T30 because early reflections depend heavily on the specific source-receiver geometry. ISO 3382-1 recommends reporting both T30 and EDT for a complete characterization.

When T20, T30, and EDT Disagree

In a room with a single exponential decay (ideal diffuse field), all three metrics give the same value. When they diverge, it reveals important acoustic characteristics:

T30 > T20: Coupled Space

The decay has a "knee" — fast initial decay followed by slower late decay. This is typical of coupled spaces: a main hall connected to a reverberant foyer, a church nave connected to a transept, or a room with a large alcove. The T20 captures the fast initial decay; T30 captures the slower tail from the coupled space.

EDT < T30: Strong Early Reflections

A room with many nearby reflective surfaces (low ceiling, side walls within 3 m) has high early reflection density. The initial decay is fast (high EDT) but the late field decays at the room's natural rate (T30). This is generally positive for clarity — STI and C50 will be good.

EDT > T30: Weak Early Field

Rare but occurs in very large spaces with distant boundaries. The direct sound and early reflections are weak relative to the late reverberant field. The room sounds more reverberant than T30 suggests. Treatment focus should be on early reflection management rather than overall absorption.

Octave Band Analysis

Report all three metrics per octave band (125, 250, 500, 1000, 2000, 4000 Hz). The single-number RT60 is typically the average of 500 and 1000 Hz. However, octave-band data reveals frequency-dependent behavior:

• Low frequencies (125-250 Hz): Often 50-100% longer RT60 than mid frequencies due to less absorption. This is normal and desirable in concert halls (warmth), but problematic in speech venues (muddiness).
• Mid frequencies (500-1000 Hz): The reference bands for single-number RT60. Most room type targets are specified at these frequencies.
• High frequencies (2000-4000 Hz): Air absorption becomes significant in large rooms (above 2000 m³). RT60 at 4 kHz can be 0.3-0.5 s shorter than at 500 Hz due to air absorption alone.

The treatment calculator uses per-band RT60 to recommend absorption materials that target specific frequency ranges.

Linear Regression Quality

The r² (coefficient of determination) of the linear regression fit to the EDC within the evaluation range indicates measurement quality. ISO 3382-1 does not specify a minimum r² but values above 0.995 indicate clean, reliable data. Values below 0.99 suggest:

• Insufficient INR (noise contaminating the evaluation range)
• Non-exponential decay (coupled spaces)
• Time-variant conditions during measurement

SonaVyx displays r² per octave band. Bands with r² < 0.99 are flagged with a warning indicator.

Practical Recommendation

For most users: report T20 for quick assessments, T30 for formal ISO 3382 reports, and always include EDT when characterizing listener experience. The room analysis workflow captures all three automatically and highlights any discrepancies that indicate non-standard acoustic behavior.