AudioContext Fingerprinting: Tracking That Does Not Use Your Microphone

Q: What is audio fingerprinting?

Audio fingerprinting is a tracking method that uses the browser Web Audio API to generate a sound signal internally, process it, and measure tiny numeric differences in how your device handles it. Those differences come from your hardware and software and form an identifier. No sound is ever recorded or played out loud.

Q: Does audio fingerprinting use my microphone?

No. This is the most common misunderstanding. Audio fingerprinting never accesses the microphone, never asks for microphone permission, and captures no real audio. It only measures how your device performs internal audio math on a synthetic signal.

Q: Can I hear the sound used in audio fingerprinting?

No. Most implementations use an offline audio context that computes the signal without ever sending it to your speakers. The entire process is silent and typically completes in a fraction of a second. You would have no way of noticing it.

Q: Is audio fingerprinting stronger than canvas fingerprinting?

Generally it carries less identifying information on its own than canvas fingerprinting, but it adds independent signal. Trackers combine it with canvas, fonts, and other vectors to build a stronger overall fingerprint. Its main value is that it is hard to evade and stable over time.

Q: How do I block audio fingerprinting?

Brave randomizes Web Audio output per site, and Tor Browser and the Firefox resistFingerprinting setting restrict or standardize the API. Some extensions can add noise to the audio context as well. None of these are perfect, and blocking the API entirely can break legitimate audio on websites.

~/sheets/audio-fingerprinting.md

The Fingerprint Your Speakers Never Play
Audio fingerprinting sounds alarming, and the most common belief about it is wrong. It does not listen to you. It uses the browser Web Audio API to generate a signal internally, run it through your audio processing stack, and measure tiny numeric differences in the result. Those differences come from your hardware and software, and hashing them produces an identifier. Check what your browser reveals with our browser fingerprint tool, or start with the broader browser fingerprinting overview.
It Does Not Use Your Microphone
This is the single most important point. Audio fingerprinting never accesses the microphone, never asks for microphone permission, and records no real sound. It generates a synthetic tone in software and measures how your device performs the math on it. Nothing is captured from the room you are in.
How It Works
A common implementation chains an OscillatorNode as the signal source into a DynamicsCompressorNode, whose non-linear processing amplifies tiny floating-point differences, and reads the result, often through an OfflineAudioContext that renders the buffer without ever playing a sound. The whole process is silent and usually finishes in a fraction of a second.
Why It Identifies You
The output varies because it depends on how your particular combination of processor, operating system audio implementation, and browser audio engine performs digital signal processing. Like canvas fingerprinting, it is stateless, so it survives clearing cookies and private browsing, and it tends to be stable because your audio stack rarely changes. It was documented across the web in the 2016 Princeton study Online Tracking: A 1-million-site Measurement and Analysis.
How Strong Is It
On its own, audio fingerprinting generally carries less identifying information than canvas. Its value to trackers is that it adds an independent signal that is hard to evade, which strengthens a combined fingerprint when paired with canvas, fonts, and other vectors. You can see your overall exposure in our privacy checkup.
How to Reduce It
Brave adds randomized noise to Web Audio output per site. Tor Browser and the Firefox resistFingerprinting setting standardize or restrict the API. Some extensions can add noise to the audio context as well. Disabling Web Audio entirely is possible in some setups but breaks legitimate audio features and can itself become a distinguishing signal.
Frequently Asked Questions
What is audio fingerprinting?
Audio fingerprinting is a tracking method that uses the browser Web Audio API to generate a sound signal internally, process it, and measure tiny numeric differences in how your device handles it. Those differences come from your hardware and software and form an identifier. No sound is ever recorded or played out loud.
Does audio fingerprinting use my microphone?
No. This is the most common misunderstanding. Audio fingerprinting never accesses the microphone, never asks for microphone permission, and captures no real audio. It only measures how your device performs internal audio math on a synthetic signal.
Can I hear the sound used in audio fingerprinting?
No. Most implementations use an offline audio context that computes the signal without ever sending it to your speakers. The entire process is silent and typically completes in a fraction of a second. You would have no way of noticing it.
Is audio fingerprinting stronger than canvas fingerprinting?
Generally it carries less identifying information on its own than canvas fingerprinting, but it adds independent signal. Trackers combine it with canvas, fonts, and other vectors to build a stronger overall fingerprint. Its main value is that it is hard to evade and stable over time.
How do I block audio fingerprinting?
Brave randomizes Web Audio output per site, and Tor Browser and the Firefox resistFingerprinting setting restrict or standardize the API. Some extensions can add noise to the audio context as well. None of these are perfect, and blocking the API entirely can break legitimate audio on websites.
Last updated: April 2026