What Is Audio Fingerprinting in the Browser?

Audio fingerprinting is a browser tracking technique that builds an identifier from the way your device processes audio signals. It uses the Web Audio API, a legitimate browser feature for music apps, games and audio effects. A script creates a short audio signal, processes it in the browser and measures the output. Different hardware and software configurations process audio slightly differently, so the result often differs between devices. For the overview of where audio fingerprinting fits, see our complete browser fingerprinting guide and the online privacy check hub.

Like canvas and WebGL fingerprinting, audio fingerprinting requires no storage on your device. Cookie deletion, private browsing mode and VPN usage do not affect it.

Is This the Same as Shazam-Style Audio Fingerprinting?

No. The term "audio fingerprint" has two meanings:

Browser audio fingerprintAcoustic fingerprint (music recognition)
GoalRecognize a device or browserRecognize a song or audio file
InputA test tone the script generates itselfA recording, stream or MP3
Used byTrackers and fraud detectionServices such as Shazam, ACRCloud or AcoustID

Music databases such as MusicBrainz use acoustic fingerprints for automatic content recognition: the system sends a fingerprint of the audio data to a server, which returns a matching ID. This article covers the browser type.

How Does Audio Fingerprinting Work?

The process creates an AudioContext object in JavaScript, generates a test oscillator signal, applies audio processing and measures the resulting waveform values. The steps:

  1. Create an audio context. Scripts often use an OfflineAudioContext, which renders audio into a buffer instead of playing it.
  2. Generate a signal. An OscillatorNode produces a test tone.
  3. Process it. The signal runs through audio processing such as compression or filtering.
  4. Read and hash the samples. The script reads the output values and turns them into a hash, the audio fingerprint.

These values differ between devices because of the browser's audio engine, the operating system and the math libraries that calculate the samples. No sound is played. Your speakers and microphone are not involved, and you hear nothing.

How Common Is Audio Fingerprinting?

Audio fingerprinting is rare compared with canvas fingerprinting. Princeton researchers discovered it in 2016 while crawling the top 1 million websites. They found AudioContext scripts on 67 sites, compared with 14,371 sites using canvas fingerprinting. They also noted that the technique may have some limitations for fingerprinting. Today it is mostly one signal among many in larger fingerprinting scripts.

How to Protect Against Audio Fingerprinting

Brave browser randomizes AudioContext output to prevent audio fingerprinting. Firefox with privacy.resistFingerprinting enabled reports modified audio values. The CanvasBlocker extension for Firefox provides additional audio fingerprint blocking options. In detail:

  • Brave: adds small random changes to Web Audio output. The seed changes per site and per session.
  • Firefox: since version 118 (2023), Web Audio uses one math library on all systems, which reduces differences between devices. Mozilla's developer notes that the reduction matters most with privacy.resistFingerprinting set to true.
  • Safari: Safari 26 (2025) prevents known fingerprinting scripts from reliably reading web audio output.
  • CanvasBlocker (Firefox): protects the audio API by default and alters the values in Fake mode.

How Can I Test My Audio Fingerprint?

The SpeedIQ checks do not include an audio test yet. You can still run the SpeedIQ test to see your canvas hash, hardware, screen, language, time zone and user agent. These values show how much your browser reveals in general. The results stay in your browser.

→ Try it free: You can check your browser fingerprint directly at SpeedIQ — no account, no ads.

FAQ

Can I hear audio fingerprinting?

No. The script renders the test tone into a buffer and never plays it. Your speakers and microphone are not used.

Does audio fingerprinting use my microphone?

No. Browser audio fingerprinting only processes a signal the script generates itself. Microphone access would need your explicit permission.

Is audio fingerprinting the same as Shazam?

No. Shazam-style acoustic fingerprinting recognizes songs from a recording. Browser audio fingerprinting recognizes a device from how it processes a test tone.

Does a VPN or incognito mode stop audio fingerprinting?

No. Neither changes how your browser computes audio. The fingerprint stays the same until your browser or system changes.

Which browser protects against audio fingerprinting?

Brave randomizes Web Audio output, Safari 26 limits known fingerprinting scripts, and Firefox reduces differences, most strongly with privacy.resistFingerprinting enabled.

Sources