Canvas, WebGL and Audio Fingerprints Explained: Why Every Device Leaves a Different Mark (2026)
Browser fingerprinting: passive and active
W3C divides browser fingerprinting into two types. Passive fingerprinting is based on characteristics observable in the contents of web requests, without any code executed on your device, such as cookies, HTTP headers, the IP address and other network-level information. Active fingerprinting is when a site runs JavaScript or other code on your device to observe additional characteristics of the browser, device or context. W3C lists active techniques that may include reading the window size, enumerating fonts or connected devices, evaluating performance characteristics, reading device sensors and rendering graphical patterns.
Canvas, WebGL and audio fingerprints are all active. For an overview, see what browser fingerprinting is.
Canvas fingerprint
Canvas is an HTML element that lets a page draw images with code. MDN describes the toDataURL() method as returning the canvas image as a data URL. For an ordinary page, that's just a way to export a drawing.
Used for fingerprinting, a page draws hidden text or a shape and reads the result back. How text, lines and colors render can differ slightly depending on fonts, operating system, graphics driver and browser. The difference is tiny to the human eye but measurable as numbers.
WebGL fingerprint
WebGL is a JavaScript API for rendering high-performance interactive 2D and 3D graphics in the browser without plug-ins. MDN states that the device must also have hardware that supports these features, so WebGL is tied closely to the machine's graphics system.
A page can ask about the graphics system and the capabilities it supports, and render a test scene and read the result. Because graphics cards, drivers and operating systems differ between machines, this information helps tell devices apart.
Audio fingerprint
The Web Audio API lets a page build an “audio-processing graph” of linked modules, with AudioContext as the starting point. Used for music or games, the graph plays sound through the speakers. Used for fingerprinting, a page can have the browser process a signal without playing any sound, then read the numeric result.
Very small differences in how devices and browsers process the signal can produce different results, and they are combined with other signals.
Why do results differ between devices?
The results of the three measurements depend on several factors at once:
- Graphics hardware and its drivers.
- Operating system and the fonts available.
- Browser version and configuration.
Two machines identical in these factors may give identical results; two different machines usually give different ones. That's what makes these signals useful to trackers, and it's what you need to understand when you run many profiles on one machine.
What does this mean for many browser profiles?
If several profiles run on the same machine and the same browser without their environments being separated, they may return the same canvas, WebGL and audio results. A website could then see that those profiles share a source, even when cookies and IP addresses differ.
That's why an antidetect browser gives each profile its own environment. More important than being “different” is being consistent: the characteristics within a profile must agree with each other (operating system, graphics, fonts, time zone, language, IP). An environment that contradicts itself draws attention. See time zone, language and IP must match. This article doesn't go into how a particular product builds that environment; ADSSPEED's public level of detail is on the Technical architecture page.
How to see your own fingerprint
- Open EFF's Cover Your Tracks and run the test; it shows the characteristics that make your browser stand out.
- Open ADSSPEED's Check IP and fingerprint tool to see your IP, time zone, language and browser characteristics.
- If you use several profiles, run the same test inside each and compare: profiles should differ from one another, and each profile should match itself.
Limits to know
- No signal stands alone. Trackers usually combine many signals, passive and active, plus IP and behavior.
- It's hard to remove completely. W3C notes that fully eliminating the ability to fingerprint against a determined adversary through widely deployed technical means alone is implausible.
- Blocking outright can have side effects. Turning canvas or WebGL off entirely can break pages that use graphics, and an unusual configuration can make a browser easier to recognize.
- Not every platform uses these methods, and usage changes over time.
Where ADSSPEED fits
ADSSPEED gives each browser profile and Android phone profile its own device setup and its own proxy, so profiles running on one machine don't blend together.
- Check it in the app and on the web. Use Check IP and fingerprint to see how each profile appears.
- Read more. What is an antidetect browser · What is a browser profile · WebRTC leak: what it is.
Frequently asked questions
What is a canvas fingerprint?
A page uses the canvas element to draw hidden text or shapes and then reads the result back, for example with toDataURL(), which returns image data. How fonts, lines and colors render can differ slightly by device, so the output can act as an identifying signal.
What is a WebGL fingerprint?
WebGL is an API for drawing 2D and 3D graphics in the browser, and the device needs hardware that supports it. A page can ask about the graphics system and render a test scene to see the result. Because graphics hardware and drivers differ between machines, that information contributes to a fingerprint.
What is an audio fingerprint?
The Web Audio API lets a page build an “audio-processing graph”. A page can have the browser process a signal without playing any sound through the speakers, then read the numeric result. Tiny differences between devices and browsers can contribute to a fingerprint.
Does clearing cookies remove a fingerprint?
No. Cookies are data stored on the device, so you can clear them; a fingerprint is computed from characteristics of the device and browser and needs nothing stored. W3C also notes that completely eliminating the ability to fingerprint against a determined adversary through widely deployed technical means alone is implausible.
Why do two different machines give different results?
Because the result depends on several things at once: graphics hardware, drivers, operating system, available fonts and browser version. Two machines identical in all of these may give identical results.
Where can I see my own browser fingerprint?
EFF's Cover Your Tracks shows how your browser looks to trackers. You can also use ADSSPEED's Check IP and fingerprint tool.
Sources
Run Android devices and browser profiles on your own PC
ADSSPEED manages many browser profiles and Android phone profiles from one app, each with its own device configuration, proxy and automation.

