Schneier on Security

J.D. • November 1, 2012 1:00 PM

All I can say is the guy giving the talk is known about in some crypto circles due to developing various types of compression algorithm that are bijective and thus unlike other more common compression systems don’t suffer from “known plaintext” and other key finding short cuts. Further his other written stuff on his site is fairly readable, so I’m assuming it was an experiment and he got “camera nerves” whilst recording it.

I am fascinated, even hypnotized, by how he over-articulates each and every syllable. If you were deaf you ought to have no problem reading his lips, even though his face is almost totally obscured. Is that an aspect of his accent, do you think? Or is it more idiomatic?

In terms of the crypto, his website definitely doesn’t look like the work of a crank or snake-oil salesman. However, I am a little concerned with his “BIAES” file encryption tool. He says it uses AES CBC, in two modes: with, and without an IV (where the role of IV is replaced with a SHA-1 hash of the entire message to be encrypted).

In either case, the key is changed every 32 blocks via some key evolution algorithm. He doesn’t specify exactly how the key is evolved, except to say that for the mode with an IV it “depends on the IV”. Given the power of related key attacks against AES, I think he might be better off without the key evolution aspect.