Question 1

What is propagation delay in audio?

Accepted Answer

Propagation delay is the time sound takes to travel from the speaker to the listener. Sound travels at approximately 343 meters per second at 20 degrees Celsius in dry air. At 5 meters distance, the delay is 5/343 = 14.6 milliseconds. This delay is significant for multi-speaker systems: if two speakers at different distances are not time-aligned, the listener hears the closer speaker first, followed by the farther speaker, creating an echo effect or comb filtering depending on the delay. For a single speaker, propagation delay is a constant offset that does not affect sound quality. For multiple speakers, the relative delay between them determines how they interact. SonaVyx delay finder measures the propagation delay from the impulse response peak to sub-sample accuracy.

Question 2

How fast does sound travel and what factors affect it?

Accepted Answer

Sound speed in air depends primarily on temperature: c = 331.3 + 0.606 × T (degrees Celsius), giving approximately 343 m/s at 20C, 349 m/s at 30C, and 337 m/s at 10C. Humidity has a smaller effect: moist air is slightly faster (about +0.5 m/s per 10% increase in relative humidity). Atmospheric pressure has negligible effect. Altitude has no direct effect (the temperature change at altitude is what matters). In practice, the relevant range is 335-355 m/s for typical indoor and outdoor conditions. A 10-degree temperature change shifts the speed by 6 m/s, which over a 20-meter path changes the delay by about 1 ms. For critical alignment work, SonaVyx measures the actual delay directly rather than calculating from distance, which automatically accounts for the current air conditions.

Question 3

Why is delay measurement important for sound system design?

Accepted Answer

Delay measurement is essential for multi-speaker system alignment. When multiple speakers cover the same audience area, their signals must arrive within a few milliseconds of each other to avoid: comb filtering (interference patterns that color the sound), perceived echoes (distinct repetitions for delays above 30-40 ms), and image shift (the audience perceives the closer speaker as the source). In delay speaker systems (under-balcony fills, distributed ceiling speakers, concert delay towers), each speaker must be delayed to match the arrival of the main system at its coverage area. Without proper delay, the fills arrive before the mains, pulling the perceived source away from the stage. SonaVyx delay finder measures the actual acoustic delay at each position for precise alignment.

Question 4

What is the relationship between delay, distance, and frequency?

Accepted Answer

Delay is directly proportional to distance: delay = distance / speed_of_sound. Distance and frequency relate through wavelength: wavelength = speed_of_sound / frequency. At 100 Hz, one wavelength is 3.43 m. A half-wavelength path difference (1.71 m, or 5 ms) causes destructive interference at 100 Hz (and odd multiples: 300, 500, 700 Hz). A full-wavelength path difference (3.43 m, 10 ms) causes constructive interference at 100 Hz (and all multiples). This means that for sub-main alignment at 100 Hz crossover, a distance difference of 1.7 m creates maximum cancellation. At 1 kHz crossover, only 17 cm difference causes the same problem. SonaVyx delay finder resolution is one sample (0.021 ms at 48 kHz), corresponding to 7 mm distance resolution — more than sufficient for all crossover frequencies.

Question 5

How does SonaVyx detect the propagation delay?

Accepted Answer

SonaVyx uses the impulse response to find the propagation delay. The impulse response is computed from the transfer function (inverse FFT of H(f)) or from sweep deconvolution. The first significant peak in the impulse response corresponds to the direct sound arrival. SonaVyx finds this peak using a threshold-based detector: it scans forward from the start until the IR level exceeds a threshold above the noise floor. The peak position, interpolated to sub-sample accuracy using parabolic interpolation, gives the delay. This method is robust because it measures the actual acoustic travel time, automatically accounting for: speaker processing latency, digital-to-analog converter delay, air temperature effects on sound speed, and any DSP chain latency. The result is the true acoustic delay, not a theoretical calculation from distance.

Question 6

What is the difference between acoustic delay and electronic delay?

Accepted Answer

Acoustic delay is the total time from when the electrical signal enters the amplifier to when the sound arrives at the microphone. It includes: amplifier processing time, digital-to-analog conversion, speaker motor response time, and air propagation. Electronic delay is an intentional delay added by a DSP processor to time-align speakers. When you set a speaker to "5 ms delay" on the DSP, you are adding 5 ms of electronic delay to its signal chain. The total acoustic delay for that speaker is then: electronic delay + all system latencies + propagation delay. To align two speakers, you need the total acoustic delays to match at the listening position. SonaVyx measures the total acoustic delay, so you do not need to know the individual latency components. Simply measure, calculate the difference, and set the electronic delay to compensate.

Question 7

What accuracy do I need for delay measurement?

Accepted Answer

Required accuracy depends on the highest frequency of overlap between the speakers being aligned. At the crossover frequency, a 30-degree phase error corresponds to: 30/(360 × frequency) seconds. At 100 Hz: ±0.83 ms (±28 cm). At 1 kHz: ±0.083 ms (±2.8 cm). At 3 kHz: ±0.028 ms (±0.95 cm). For sub-main alignment (80-120 Hz crossover), 0.5 ms accuracy is sufficient. For main-fill alignment (full bandwidth overlap), 0.1 ms accuracy is needed. For high-frequency driver alignment (crossover above 2 kHz), 0.05 ms accuracy is necessary. SonaVyx at 48 kHz sample rate provides 0.021 ms resolution per sample, with sub-sample interpolation giving approximately 0.005 ms accuracy — sufficient for all practical alignment scenarios.

Question 8

Can I measure delay without using an impulse response?

Accepted Answer

The impulse response is the most reliable method, but alternatives exist. Cross-correlation of the reference and measurement signals directly estimates the delay as the correlation peak time — this works with any broadband signal including music. Group delay at a specific frequency gives the local delay at that frequency, useful for crossover alignment. Manual method: play a click through one speaker at a time and measure the arrival time difference on an oscilloscope or SonaVyx waveform display. Physical distance measurement with a tape measure gives theoretical delay (distance / 343 m/s at 20C), but misses electronic latency. SonaVyx primarily uses the impulse response method because it is the most accurate and captures the complete system delay including all electronic processing.

Propagation Delay Basics: Speed of Sound, Distance, and Multi-Speaker Alignment

Frequently Asked Questions