Transfer Function Measurement Software
A transfer function describes the frequency-dependent relationship between a system input and output, expressed as H(f) = Y(f)/X(f) where Y is the output spectrum and X is the input spectrum. SonaVyx measures magnitude, phase, coherence, and group delay using the H1 estimator with Welch averaging, providing the same dual-channel FFT analysis used in Smaart and SysTune, entirely in your browser via Rust WASM processing.
Try It Now
Open the transfer function analyzer — dual-channel magnitude, phase, and coherence.
Technical Specifications
| Parameter | Value | Standard |
|---|---|---|
| Estimator | H1 = Gxy / Gxx | AES-2id:2023 |
| Coherence | gamma^2 = |Gxy|^2 / (Gxx * Gyy) | Welch method |
| FFT Sizes | 1024 / 2048 / 4096 / 8192 / 16384 | User selectable |
| Averaging | Off / Linear / Exponential / Peak Hold | 50% or 75% overlap |
| Smoothing | 1/1 to 1/24 octave | Log-frequency smoothing |
| Phase Display | Wrapped / unwrapped / group delay | Radians or degrees |
| Delay Finder | Cross-correlation peak | Sub-sample interpolation |
| Window | Blackman-Harris (default) | ISO 18431-2 |
How to Measure Transfer Function
Set Up Dual-Channel Input
Connect the system output signal as the reference channel (X) and a measurement microphone as the measurement channel (Y). In single-device loopback mode, SonaVyx generates the reference signal internally and captures the microphone return. In multi-device mode, use WebSocket pairing between two devices.
Configure Measurement Parameters
Select FFT size (4096 for general use), averaging mode (8-16 frames of exponential averaging for stable results), and smoothing (1/6 octave for system tuning, none for detailed analysis). The Blackman-Harris window provides the best dynamic range for transfer function measurements.
Start the Measurement
Begin playing the test signal through the system. SonaVyx computes the cross-spectral density (Gxy), auto-spectral densities (Gxx, Gyy), and derives the H1 transfer function estimate. Results appear as magnitude (dB), phase (degrees), and coherence (0 to 1) plots in real time.
Interpret the Results
Magnitude shows the system frequency response (flat is ideal). Phase shows timing relationships (flat or linear slope is ideal). Coherence above 0.8 indicates reliable measurements. Low coherence suggests noise interference, poor signal-to-noise ratio, or non-linear system behavior at those frequencies.
Extract Delay and Apply Corrections
Use the impulse response (inverse FFT of the transfer function) to find propagation delay. Use the magnitude response to guide EQ decisions. Use phase to verify speaker alignment. Store traces for before/after comparison when making system adjustments.
Understanding Transfer Function Measurement
The transfer function is the most powerful measurement in audio system analysis because it isolates the system response from the test signal and ambient noise. Unlike an RTA that shows whatever the microphone captures, the transfer function divides the output spectrum by the input spectrum, revealing only what the system does to the signal. This mathematical division removes the test signal characteristics, leaving the pure system response.
Cross-Spectral Density and Welch Averaging
SonaVyx computes the transfer function using Welch's method of averaged periodograms. The input and output are divided into overlapping segments, windowed, and transformed via FFT. The cross-spectral density Gxy = X*(f) * Y(f) captures the relationship between input and output. Auto-spectral densities Gxx and Gyy capture each channel's power. Averaging multiple segments reduces the variance of the estimate and improves coherence accuracy.
Phase Interpretation for System Alignment
Phase in the transfer function reveals timing relationships that are invisible in magnitude-only measurements. A linear phase slope indicates a constant time delay. A flat phase line means zero delay. Phase wraps from +180 to -180 degrees, but SonaVyx provides unwrapped phase for easier interpretation. When aligning two speakers, the goal is matching their phase responses at the crossover frequency so they sum constructively.
From Smaart to SonaVyx
SonaVyx implements the same H1 transfer function estimator used in Smaart, SysTune, and other professional measurement systems. The key difference is that SonaVyx runs entirely in the browser using Rust compiled to WebAssembly, achieving comparable DSP performance without requiring a desktop application or audio interface drivers. For basic single-device measurements, your laptop microphone and speakers are sufficient to start.
Transfer Function Software Comparison
| Feature | SonaVyx | Smaart v9 | REW | OSM |
|---|---|---|---|---|
| H1 transfer function | Yes (WASM) | Yes (native) | Yes (Java) | Yes (native) |
| Live coherence | Yes (real-time) | Yes | Post-capture | Yes |
| Phase + group delay | Yes | Yes | Yes | Phase only |
| Multi-trace overlay | 10 traces | Unlimited | Unlimited | 2 traces |
| Browser-based | Yes | No | No | No |
| Internal signal gen | Yes (loopback) | External | Yes | External |
| Delay finder | Yes (sub-sample) | Yes | Yes | Yes |
| Price | Free | $898 | Free | Free |
Frequently Asked Questions
Related Tools & Resources
Standards References
- AES-2id:2023 — AES information document for room acoustics and sound reinforcement systems: Measurement of loudspeaker-room impulse response
- IEC 60268-5:2003 — Sound system equipment: Loudspeakers (frequency response measurement)
- ISO 18431-2:2004 — Mechanical vibration and shock: Signal processing: Time domain windows for Fourier Transform analysis
- AES17-2020 — AES standard method for digital audio engineering: Measurement of digital audio equipment