[addaon]

One thing that still leaks is the exact spacing between words adjacent to the black boxes. This is a particular concern one only one or two words are blacked out -- for example a name. With a known font (with known character widths and kerning tables) and a known text engine applying those kerning rules, there's often very few combinations of letters that give the exact pixel (or subpixel) width blank space.

[lelandfe]

Clever. Monospaced fonts improve things, with the tradeoff of making it trivial to know the number of characters redacted.

[8note]

For a first name + last name combination, you still have a lot of potential combinations based on the total length

[addaon]

Yes, but being able to rule out "Firstname Lastname" can itself be interesting. Depending on context, the space of names is not all people, but can be very restricted.

[toomanydoubts]

Has someone made a demo of that?

[lelandfe]

https://cryptome.org/cia-decrypt.htm

[toomanydoubts]

Thanks.

[bobsmooth]

So blurring can be reversed and black bars can be cracked, is there any way to securely redact a document?

[pabs3]

Convert to ASCII, normalise whitespace, normalise spelling, normalise word choice hope, there are no word or paragraph based watermarks present.

Or just summarise the content yourself and hope there are no event/story/narrative based watermarks present.

[lelandfe]

I wonder if there are redaction tools that randomly stretch or warp the blacked out areas so as to make character guessing impossible.