4 ms·

I wonder whether introducing slight per-pixel-per-draw noise into the values could be used here, to mask this sort of detail? Or would you be able to average it

I wonder whether introducing slight per-pixel-per-draw noise into the values could be used here, to mask this sort of detail? Or would you be able to average it out somehow

gruez 4y ago
>Or would you be able to average it out somehow Isn't it pretty obvious? Generate the same image 1000 times and take the median pixel value for each pixel to get the "real" image.
- leodriesch 4y ago
  If the noise would not change between individual calls but only on page reloads it would be pretty hard to get 1000 samples.
  - gruez 4y ago
    how do you ensure the noise doesn't change between calls? if I want to fingerprint how the letter "a" is rendered, how do you ensure that I can't try drawing "a1", "a2", "a3", etc. which are "different" draw calls, but still allow me to build a composite image of what "a" looks like?
tomxor 4y ago
I think you're right on both... adding noise would reduce the accuracy of the fingerprint, but boarder biases would continue to be detectable. For instance i'm aware of some quite significant biases between firefox and chrome in antialiasing fillRect() for subipixel values as the dimensions drop bellow 1x1 pixel that only a distracting amount of noise would be able to conceal for one sample. I know that's already detectable via the useragent string, but I suspect those who are focused on collecting those biases will find strong ones for other data points.
fn-mote 4y ago
In general noise is never the answer to fingerprinting. It makes gathering "accurate" data a little harder; it doesn't stop statistical analysis from revealing the truth.
antonok 4y ago
This totally works as an approach! Brave does it for canvas data and a bunch of other web APIs as well. The "farbling" noise is deterministic per profile/origin combo to prevent being able to average it out across multiple page loads, but otherwise random. https://brave.com/privacy-updates/4-fingerprinting-defenses-2.0/ https://brave.com/privacy-updates/4-fingerprinting-defenses-...
- codedokode 4y ago
  This allows to run fingerprinting script on hundred of domains and filter out the noise though.
  - antonok 4y ago
    If you own hundreds of domains, you're welcome to try, but there's no good reason to. Fingerprinting scripts rely on a canvas being identical between sessions to identify that the same user is even there in the first place. If you can already single out one user's activity across a hundred domains, you don't need to fingerprint them any further.

I wonder whether introducing slight per-pixel-per-draw noise into the values could be used here, to mask this sort of detail? Or would you be able to average it out somehow

gruez 4y ago

>Or would you be able to average it out somehow Isn't it pretty obvious? Generate the same image 1000 times and take the median pixel value for each pixel to get the "real" image.

leodriesch 4y ago
If the noise would not change between individual calls but only on page reloads it would be pretty hard to get 1000 samples.
- gruez 4y ago
  how do you ensure the noise doesn't change between calls? if I want to fingerprint how the letter "a" is rendered, how do you ensure that I can't try drawing "a1", "a2", "a3", etc. which are "different" draw calls, but still allow me to build a composite image of what "a" looks like?

tomxor 4y ago

I think you're right on both... adding noise would reduce the accuracy of the fingerprint, but boarder biases would continue to be detectable. For instance i'm aware of some quite significant biases between firefox and chrome in antialiasing fillRect() for subipixel values as the dimensions drop bellow 1x1 pixel that only a distracting amount of noise would be able to conceal for one sample. I know that's already detectable via the useragent string, but I suspect those who are focused on collecting those biases will find strong ones for other data points.

fn-mote 4y ago

In general noise is never the answer to fingerprinting. It makes gathering "accurate" data a little harder; it doesn't stop statistical analysis from revealing the truth.

antonok 4y ago

This totally works as an approach! Brave does it for canvas data and a bunch of other web APIs as well. The "farbling" noise is deterministic per profile/origin combo to prevent being able to average it out across multiple page loads, but otherwise random. https://brave.com/privacy-updates/4-fingerprinting-defenses-2.0/ https://brave.com/privacy-updates/4-fingerprinting-defenses-...

codedokode 4y ago
This allows to run fingerprinting script on hundred of domains and filter out the noise though.
- antonok 4y ago
  If you own hundreds of domains, you're welcome to try, but there's no good reason to. Fingerprinting scripts rely on a canvas being identical between sessions to identify that the same user is even there in the first place. If you can already single out one user's activity across a hundred domains, you don't need to fingerprint them any further.