Question 1

What is the difference between T20, T30, and EDT?

Accepted Answer

T20 and T30 are derived from the energy decay curve (EDC) by measuring the slope of decay over different ranges and extrapolating to 60 dB. T20 uses the decay from -5 dB to -25 dB (20 dB range), while T30 uses -5 dB to -35 dB (30 dB range). EDT (Early Decay Time) measures from 0 dB to -10 dB and is most perceptive to early reflections. T30 is the most reliable for ISO 3382 compliance because it uses a larger dynamic range, but requires a higher signal-to-noise ratio (at least 45 dB INR). T20 is more practical in noisy environments (needs 35 dB INR). EDT correlates best with perceived reverberance during music and speech. SonaVyx calculates all three simultaneously and flags which are reliable based on the measured INR.

Question 2

Should I use a sweep or interrupted noise to measure RT60?

Accepted Answer

Logarithmic sine sweeps (Farina method) are preferred for most applications. Sweeps provide high signal-to-noise ratio, immunity to time-varying noise, and the ability to separate the impulse response from harmonic distortion. Interrupted noise (the traditional method per ISO 3382-2) excites the room with broadband noise that is abruptly cut off, and the decay is measured directly. Interrupted noise requires multiple measurements (at least 3) averaged together because each decay is random. Sweeps produce deterministic impulse responses from a single measurement. SonaVyx uses logarithmic sine sweeps (configurable 2-20 seconds) and deconvolution to extract the room impulse response, then computes the EDC via Schroeder backward integration.

Question 3

What is MLS and when should I use it instead of a sweep?

Accepted Answer

MLS (Maximum Length Sequence) is a pseudo-random binary signal used for impulse response measurement via circular cross-correlation. MLS has a flat power spectrum and distributes energy evenly over the measurement period, providing good SNR in noisy environments. However, MLS is sensitive to time variance (if anything changes during the measurement, the result is corrupted) and cannot separate harmonic distortion from the linear response. Use MLS when: you need fast measurements in known time-invariant conditions, or the system under test has limited dynamic range. Use sweeps when: the environment has time-varying noise, you want to characterize nonlinearities, or you need the highest possible quality impulse response. SonaVyx supports both methods.

Question 4

How many measurement positions do I need for a valid RT60 result?

Accepted Answer

ISO 3382-1 (performance spaces) requires at least 6 source-receiver combinations for engineering-grade results and 12 for precision-grade. ISO 3382-2 (ordinary rooms) requires at least 2 source positions and 3 receiver positions (6 combinations minimum) for engineering grade. The microphone should be at least 1 m from any surface, 2 m from the source, and at different distances from the source. Receiver positions should be separated by at least half a wavelength of the lowest frequency of interest (about 2 m for 125 Hz). In practice, for system tuning purposes, 3-5 positions in the audience area give a good spatial average. SonaVyx lets you take multiple measurements and computes the spatial average with standard deviation.

Question 5

What is INR and how much do I need for a valid RT60 measurement?

Accepted Answer

INR (Impulse-to-Noise Ratio) is the difference between the peak level of the impulse response and the noise floor, measured in dB. It determines the usable dynamic range for decay curve fitting. ISO 3382 requires: at least 35 dB INR for a valid T20 measurement (20 dB evaluation range + 5 dB start offset + 10 dB safety margin) and at least 45 dB INR for T30. If INR is insufficient, the decay curve bends upward at the bottom as it merges with the noise floor, causing underestimated reverberation times. To improve INR: use a longer sweep, increase source level, reduce background noise, or measure during quiet hours. SonaVyx displays INR for each measurement and warns when it is too low for reliable analysis.

Question 6

What is Schroeder backward integration and why is it used?

Accepted Answer

Schroeder backward integration computes the energy decay curve (EDC) by integrating the squared impulse response from the end backward to each time point. The result is a smooth, monotonically decreasing curve that represents the ensemble average of infinitely many interrupted noise decays. This is mathematically superior to directly observing individual noise decays, which are random and require averaging many trials. The key advantage is that a single sweep measurement yields a clean decay curve equivalent to averaging thousands of interrupted noise measurements. SonaVyx applies Schroeder integration after truncating the impulse response at the noise floor crossing point (Lundeby method) to prevent noise floor contamination from biasing the decay curve upward.

Question 7

Can I measure RT60 with a balloon pop or hand clap?

Accepted Answer

Balloon pops and hand claps are impulsive sources that can give a rough estimate of RT60, but they have significant limitations. Their spectrum is uncontrolled and non-repeatable: balloon pops are weak below 250 Hz, hand claps are weak below 500 Hz. Neither provides sufficient energy for reliable low-frequency RT60 measurements. The INR is typically limited to 20-30 dB, which may be insufficient for T30. ISO 3382 allows impulsive sources but recommends that they produce omnidirectional radiation and sufficient energy across all octave bands. For quick site surveys where a rough estimate is acceptable, a balloon pop works. For reliable data, use a dodecahedron speaker with a swept signal. SonaVyx can capture and analyze balloon pop impulse responses but will flag bands with insufficient INR.

Question 8

What sweep duration should I use for RT60 measurement?

Accepted Answer

Longer sweeps provide higher signal-to-noise ratio because the same source energy is spread over more time, and the deconvolution concentrates it into a short impulse. As a rule of thumb, the sweep should be at least twice the expected RT60. For a room with 1.5 s RT60, use at least a 3-second sweep. For reverberant spaces (churches, concert halls with RT60 above 3 s), use 10-20 seconds. Longer sweeps also improve low-frequency resolution because more cycles of low frequencies are captured. The trade-off is measurement time and susceptibility to time-varying conditions. SonaVyx defaults to 5 seconds, which works well for most rooms, and offers 2, 5, 10, 15, and 20 second options. The measurement wizard recommends a duration based on room size.

RT60 Measurement Methods: Sweep, MLS, T20, T30, EDT, and ISO 3382 Compliance

Frequently Asked Questions