How to Measure Sound System Frequency Response
TL;DR
Transfer function measurement reveals how your sound system modifies the audio signal — showing every peak, dip, and phase shift between the electrical input and the acoustic output. Unlike RTA which only shows what the mic hears, transfer function compares input and output, isolating the system's contribution from the room's background noise.
Transfer Function vs RTA
An RTA shows the spectrum of whatever the microphone captures — useful for a quick look, but it includes ambient noise and does not distinguish between the speaker's response and the room's influence. Transfer function (TF) compares the electrical reference signal to the acoustic measurement, revealing the system's true response. Coherence indicates measurement quality at each frequency.
Step-by-Step Procedure
- Connect your reference signal. You need a reference (what goes into the speaker) and a measurement (what comes out). In SonaVyx's single-device loopback mode, the tool generates pink noise, plays it through the speaker, and captures the mic simultaneously. For two-channel measurement, use a separate source device paired via WebSocket.
- Open Transfer Function mode at /measure?mode=tf. Select your FFT size (4096 for a good balance of frequency and time resolution at 48 kHz, giving ~12 Hz resolution).
- Position the measurement microphone. Start at the primary listening position (mix position, front-of-house, or design seat). Point the mic at the speaker array. Height: 1.2 m seated, 1.5 m standing.
- Start measurement. Tap Start. SonaVyx plays pink noise and begins computing the transfer function using the H1 estimator. Allow at least 8-10 averages for a stable reading. Watch the coherence — it should be above 0.8 across the measurement band.
- Interpret the magnitude plot. A flat line at 0 dB means the system reproduces the input signal perfectly. Peaks above +6 dB indicate resonances or feedback-prone frequencies. Dips below -10 dB indicate cancellations (often from reflections or destructive interference between speakers).
- Check coherence. Low coherence (below 0.5) at specific frequencies means the measurement is unreliable there — caused by insufficient signal-to-noise ratio, reflections, or nonlinearities. Increase the source level or move the mic closer to improve coherence.
- Check phase. Smooth, gradually descending phase indicates proper time alignment. Sudden phase wraps may indicate crossover regions or time alignment issues between speaker components.
- Store and compare. Store the trace (press S), move to the next measurement position, and measure again. Overlay traces to see how the response varies across the coverage area.
Common Mistakes
- Measuring RTA instead of TF. RTA shows the room spectrum, not the system response. Always use transfer function for system measurement.
- Insufficient averaging. A single FFT snapshot is noisy. Use at least 8 linear averages for a usable trace.
- Ignoring coherence. A beautiful-looking magnitude trace means nothing if coherence is below 0.5. Low coherence frequencies are noise, not signal.
- Wrong smoothing. 1/3 octave smoothing is appropriate for system EQ decisions. Use less smoothing (1/12 or 1/24) when looking for narrow-band problems like comb filtering.
Tool Bridge
Open SonaVyx Transfer Function mode to measure your system's frequency response. Use the trace store to compare multiple positions and before/after EQ adjustments.
Standard Reference
AES-2id:
Try It Now
Open this measurement tool in your browser — free, no download required.
Last updated: March 19, 2026