[cryptography] Which encryption chips are compromised?

James Cloos cloos at jhcloos.com
Sun Nov 10 19:40:14 EST 2013


>>>>> "JY" == John Young <jya at pipeline.com> writes:

JY> Two types of PDF programs were used. Thus the redactions made
JY> by likely different means and therefore possibly different means
JY> might be used to lift the redactions.

A quick look shows that both files use DCT-compressed 300-dpi images
for each page; they may be scans.

Attempts to unredact should treat the pages as the jpegs they are.

I suspect matching the font and testing pixel lengths at 300 dpi
may be the only productive attack, but there is variation in the
intensity of the pixels in the redaction; perhaps something can
be found by analyzing it as pixel data?  Or perhaps the variation
is just a combination of quantum noise from the scan and jpeg/DCT
artifacts?

-JimC
-- 
James Cloos <cloos at jhcloos.com>         OpenPGP: 1024D/ED7DAEA6


More information about the cryptography mailing list