Audio Fingerprinting
Audio fingerprinting is a technique that uses the Web Audio API to generate and process a sound signal, then measures the numerical output to derive a device-specific identifier. Differences in audio hardware, digital signal processing, and floating-point behavior across devices cause the same synthesized audio to produce subtly different results.
How Audio Fingerprinting works
A script builds an audio processing graph, typically using an oscillator to generate a waveform routed through nodes such as a compressor or analyser. Crucially the audio is usually processed in an offline context, so no sound is played aloud; the goal is to compute the resulting sample values rather than to make noise.
The script reads back the processed samples and reduces them to a compact hash or summary statistic. Because the computation traverses the browser's audio stack and the device's floating-point implementation, tiny variations accumulate into a value that is consistent for one device and differs across devices.
Audio fingerprints are moderately stable and add a dimension that is independent of graphics-based signals, which strengthens the overall fingerprint. Like other rendering probes, they can be perturbed by privacy tools that inject noise into audio output, and that perturbation can itself be detected.
Why Audio Fingerprinting matters for fraud prevention
Audio fingerprinting adds independent entropy that complements canvas and WebGL signals, making the combined device profile harder to fake convincingly. For fraud prevention this diversity matters because spoofing tools may neutralize one class of signal while leaving others intact, so cross-checking audio helps expose inconsistent or manipulated environments. It also works silently in the background without user-visible effects.
How TRACIO handles it
TRACIO can incorporate audio-stack behavior as one of its many device signals, valued for being orthogonal to graphics-based probes. It is weighted by observed stability and used to corroborate other signals rather than as a standalone identifier. When audio output shows signs of deliberate noise injection, TRACIO factors that inconsistency into its assessment of the environment.
Related terms
Explore further
Frequently asked questions
Identify every device with confidence
Start with a free plan of 2,500 API calls per month. No credit card required.