[11205] in cryptography@c2.net mail archive
Re: building a true RNG
daemon@ATHENA.MIT.EDU (Eugen Leitl)
Wed Jul 24 14:00:51 2002
Date: Wed, 24 Jul 2002 19:17:12 +0200 (CEST)
From: Eugen Leitl <eugen@leitl.org>
To: Paul Crowley <paul@ciphergoth.org>
Cc: "John S. Denker" <jsd@monmouth.com>,
<cryptography@wasabisystems.com>
In-Reply-To: <87r8htmqyl.fsf@saltationism.subnet.hedonism.cluefactory.org.uk>
On 24 Jul 2002, Paul Crowley wrote:
> I can't believe any compression software could be as fast as just
> feeding the signal straight into SHA-1.
I haven't tried this, but assuming I'm digitizing dark video and only get
noise in the lower significant bits I can just mask out the constant
(zero) ones and paste them together to destill the entropy with a very low
computational cost before feeding it into a cryptohash.
As an aside to what constitutes physical entropy of a system it is indeed
depending on context of the measurement. A good source of information on
entropy in all contexts is
http://www.math.psu.edu/gunesch/entropy.html
---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo@wasabisystems.com
