Feature Request: Audiocontext Fingerprint Protection

[andrew-ma]

Description

Audiocontext Fingerprint Protection

Who's implementing?

• I'm willing to implement this feature myself

The problem

From the results of "https://coveryourtracks.eff.org/", there is a "Audiocontext Fingerprint" option that can be used to fingerprint users/browsers.

Safari has implemented some measures to protect from Audio Fingerprinting ("https://fingerprint.com/blog/bypassing-safari-17-audio-fingerprinting-protection/").

It would be wonderful if this could be implemented and added as an Ungoogled-Chromium feature/switch/flag option.

Possible solutions

On the linked fingerprint.com blog, this is the audio fingerprinting protection code implemented in Safari:

void applyNoise(float* values, size_t numberOfElementsToProcess, float magnitude)
{
    WeakRandom generator;
    for (size_t i = 0; i < numberOfElementsToProcess; ++i)
        values[i] *= 1 + magnitude * (2 * generator.get() - 1);
}

The description says this:

• An audio signal produced with the Audio API is an array of numbers representing the signal amplitude at each moment of time (also called “audio samples”). When fingerprinting protection is on, Safari adds a random noise to every sample individually. A noised sample lies between sample*(1-magnitude) and sample*(1+magnitude), and the distribution is uniform(https://en.wikipedia.org/wiki/Continuous_uniform_distribution). This is how it’s implemented in Safari (https://github.com/WebKit/WebKit/blob/167dc5118a3f6228a19df40e673ea0a6d03b9bec/Source/WebCore/platform/audio/AudioUtilities.cpp#L80):

Alternatives

No response

Additional context

No response

[PF4Public]

By any chance, have you checked whether the solution in Safary works correctly?