These settings are tied to the hardware of your device. Normally, we recommend keeping the default values. That's because, by default, Multilogin provides you with an optimized fingerprint, meaning all aspects of it naturally work well together.

If you do decide to make changes here, remember to keep the values consistent by following the below recommendations. 

Screen resolution

Analyzing the screen resolution is a common method for browser fingerprinting. These tips will help you avoid red flags:

• Don't use resolutions larger than your native one: Websites might check for differences between the declared resolution and the available screen area
• Be consistent in a team: Keep profile resolutions within the smallest screen used in your team (if your device is 4K and a colleague has Full HD, set the resolution to 1920 x 1080 or smaller)
• Don't minimize the window: When Multilogin launches a profile, it's maximized according to the resolution in profile settings; it's best not to minimize the profile, since most users have their browsers maximized

Media devices

WebRTC is a browser plugin that enables audio and video communication on a web page. To make this happen, it taps into your media devices like microphones, cameras, and headphones. But this process can get risky – keep reading to learn why!

Device enumeration

Websites can use WebRTC to check out all your devices, making a list of what you've got. While just this list might not pinpoint you specifically, it can still contribute to creating your digital portrait.

In Multilogin, you can customize the number of different devices within these limits:

• 📷 Video inputs (number of web cameras): 0–1
• 🎤 Audio inputs (number of microphones): 0–4
• 🔊 Audio outputs (number of speakers or headphones): 0–4

ID tracking

Although browsers protect the full model names of your media devices, websites can still track their unique IDs. In Multilogin, they are masked by default for every device.

WebGL + WebGPU metadata

WebGL and WebGPU metadata is like your browser profile's artistic signature. It reveals details about how your device creates graphics on the web, such as the brand of the graphics tool and its specific model. Websites use this signature to recognize and remember you online.

WebGL graphics

WebGL is a JavaScript technology that lets websites create 3D graphics on a webpage. At the same time, websites can use it to identify you in two ways:

• WebGL report: They inspect the entire WebGL browser report table, sometimes turning it into a hash for faster analysis
• WebGL image: They request for a hidden 3D image to be drawn and hashed, as the final result depends on your device's hardware and its calculations

Noise

In this mode, when websites request a graphics read-out, our masking algorithm intercepts it and adds a random but consistent noise. Websites may perceive such fingerprint as 100% unique if statistical analysis is applied.

It's crucial to note that even though the noise is persistent, the read-outs will still differ if a profile is launched on different devices. Check out the example below, illustrating the hash from the same profile launched on two different devices.

 

 

Canvas graphics

The Canvas API is designed for drawing 2D graphics and animations on a webpage via JavaScript and HTML. But just like WebGL, Canvas can add extra uniqueness to your browser fingerprint. According to a study by Princeton University, over 5% of websites use it for fingerprinting.

Here's the process: websites ask your browser to draw a hidden canvas image. This image appears slightly different on various devices, but remains the same for identical setups. Then the image gets transformed into a hash, serving as additional entropy in identification.

Noise

In this mode, when websites request a graphics read-out, our masking algorithm intercepts it and adds a random but consistent noise. Websites may perceive such fingerprint as 100% unique if statistical analysis is applied.

It's crucial to note that even though the noise is persistent, the read-outs will still differ if a profile is launched on different devices. Check out the example below, illustrating the hash from the same profile launched on two different devices.

 

 

AudioContext

AudioContext fingerprint is a hashed derivative of your device's audio setup.

Here's the process: websites ask your browser to simulate a sinusoidal function, mimicking how it plays audio files according to your settings and hardware. This sinusoid is then turned into a hash, which becomes an extra detail in your browser fingerprint.

Noise

Multilogin will modify the audio stack on a browser level by applying persistent noise. Websites may perceive such fingerprint as 100% unique if statistical analysis is applied.

It's crucial to note that even though the noise is persistent, the read-outs will still differ if a profile is launched on different devices. Check out the example below, illustrating the hash from the same profile launched on two different devices.