Question 1

How is an impulse response used to calculate RT60?

Accepted Answer

RT60 is derived from the impulse response using Schroeder backward integration. The squared impulse response h²(t) represents the instantaneous energy at each time point. Integrating from the end backwards gives the energy remaining at each point — this is the Energy Decay Curve (EDC). The EDC is equivalent to the ensemble average of infinitely many interrupted noise decays. A straight line fitted to the EDC on a dB scale gives the decay rate, from which RT60 is extrapolated. T20 fits the line from -5 to -25 dB and multiplies the slope by 3, T30 fits from -5 to -35 dB and multiplies by 2. This method requires a single sweep measurement and is far more reliable than directly observing noise decay. SonaVyx performs this analysis automatically and displays the EDC with the fitted regression lines.

Question 2

Can I use an impulse response to predict speech intelligibility (STI)?

Accepted Answer

Yes. STI can be calculated directly from the impulse response without needing a separate STIPA test signal. The method (described in IEC 60268-16 Annex A) computes the Modulation Transfer Function from the IR: MTF(fm) = |integral of h²(t) × exp(-j2π×fm×t) dt| / integral of h²(t) dt, where fm is each modulation frequency. This gives the MTF for each of the 7 octave bands at each of the 14 modulation frequencies. Background noise correction can be applied if the noise spectrum is known. The advantage is that a single IR measurement yields both RT60 and STI. The limitation is that the IR-derived STI assumes time-invariant conditions and does not capture the effect of fluctuating noise. SonaVyx computes STI from impulse responses on the IR analysis page.

Question 3

What is auralisation and how does it use impulse responses?

Accepted Answer

Auralisation is the process of creating an audible simulation of a room by convolving a dry audio signal with a room impulse response. The result sounds as if the original audio were played in that room. This has powerful applications: previewing how a venue will sound before construction or treatment, comparing the acoustic effect of different treatment options, demonstrating to clients what their room sounds like versus a target, and archiving the acoustic signature of historic spaces. SonaVyx supports A/B listening via the Web Audio API ConvolverNode, which performs real-time convolution. You can load a captured IR and listen to music or speech processed through it. The before/after comparison mode lets you switch between the original room IR and a post-treatment IR to hear the improvement.

Question 4

How do I use impulse responses for system time alignment?

Accepted Answer

The impulse response reveals the exact propagation delay from source to microphone as the first significant peak. For multi-speaker alignment: measure the IR of each speaker individually. Note the time of the first peak for each. The difference in arrival times is the delay that must be compensated. Set delays in the DSP so that all speakers arrive simultaneously (or with a controlled offset for Haas effect). The IR also reveals if there are significant early reflections that could cause problems. A strong reflection within 1-5 ms of the direct sound creates comb filtering and may indicate a physical obstruction or nearby surface that should be treated. SonaVyx delay finder automatically detects the direct sound arrival time and calculates the required delay offset.

Question 5

Can impulse responses be used to identify room acoustic problems?

Accepted Answer

Yes, the impulse response is a diagnostic goldmine. Problems visible in the IR include: discrete echoes appearing as late peaks more than 50 ms after the direct sound (these cause perceptible echoes and reduce intelligibility). Flutter echo appearing as a regular series of decaying pulses. Comb filtering from a strong early reflection (visible as a peak within 1-10 ms of direct sound). Coupled space decay (double-slope decay visible in the EDC). Excessive late energy (long reverberant tail). Weak direct sound relative to reflections (low D50). Missing early reflections (can indicate dead spots in the room). SonaVyx problem detector analyzes the IR for these patterns and provides diagnostic cards with severity ratings and recommended remedies.

Question 6

How are impulse responses used in loudspeaker design?

Accepted Answer

Loudspeaker designers measure the impulse response of a driver in a controlled environment (anechoic chamber or outdoors) to characterize its behavior. The IR reveals: the frequency response via FFT, the phase response and group delay through crossover regions, the time alignment between drivers in a multi-way system, the decay behavior (ringing, resonances), and the step response showing how the speaker reacts to a transient. By windowing the IR to exclude room reflections, ground-plane measurements can substitute for anechoic chambers. The minimum phase component of the IR helps design crossover filters that align the drivers in time. SonaVyx speaker measurement suite uses IR capture with windowing to characterize speakers even in non-ideal measurement environments.

Question 7

What is IR windowing and why would I use it?

Accepted Answer

IR windowing applies a time-domain window to the impulse response to select a specific time segment before taking the FFT. A short window (5-10 ms around the direct sound) excludes all room reflections, revealing the quasi-anechoic speaker response. A longer window (20-50 ms) includes early reflections, showing how the speaker interacts with nearby surfaces. The full-length IR includes the complete room response. Windowing is essential for: speaker measurements in non-anechoic environments, separating speaker behavior from room behavior for targeted EQ, analyzing specific reflection groups, and creating minimum-phase approximations. The trade-off is frequency resolution: a 5 ms window gives only 200 Hz resolution, so low-frequency analysis requires longer windows. SonaVyx IR mode supports configurable window length with real-time FFT update.

Question 8

How do I compare impulse responses from different rooms?

Accepted Answer

To meaningfully compare IRs from different rooms, normalize them to the same reference. Common approaches: normalize to the direct sound peak (sets the direct sound to 0 dB, showing relative reflection and reverberation levels). Normalize to Sound Strength G (referenced to free-field at 10 m, per ISO 3382). Compare derived parameters (RT60, C50, C80, D50, Ts, STI) rather than the raw waveforms. When overlaying IR waveforms, align them at the direct sound arrival so the time axes match. SonaVyx trace store lets you save IRs from different measurements and overlay them on the same plot. The ISO 3382 parameter table provides a standardized set of comparable numbers. For auralisation comparison, convolve the same dry recording with different room IRs and listen to them sequentially.

Impulse Response Applications: RT60, STI, Alignment, and Room Diagnostics

Frequently Asked Questions