[19888] in cryptography@c2.net mail archive
Re: Hiding data on 3.5" using "40 track mode"
daemon@ATHENA.MIT.EDU (Ben Laurie)
Wed Feb 8 11:19:56 2006
X-Original-To: cryptography@metzdowd.com
X-Original-To: cryptography@metzdowd.com
Date: Sun, 05 Feb 2006 18:01:13 +0000
From: Ben Laurie <ben@algroup.co.uk>
To: Dave Howe <DaveHowe@gmx.co.uk>
Cc: cryptography@metzdowd.com
In-Reply-To: <43E52A29.9090304@gmx.co.uk>
Dave Howe wrote:
> Oh - before I forget, I was thinking about covert channels and cds a few days
> ago and realised there is already one - CDs support a special mode called "CD+G"
> - this is used making "karaoke" cds to support the video data stream; the vast
> majority of pc drives cannot read this data - there are exceptions of course.
> however, karaoke players (and many low-end dvd players) CAN, and by design
> display them on the screen of the playback device. This is pretty much STO, but
> could conceal a message trivially that normal examination of the cd would not
> reveal, but which the recipient could display (again, trivially) using nothing
> more than a tv set and cheap mass-produced DVD player.
I used to write CD device drivers (ancient history, I wrote, with my
brother, one of the first CD rippers ever: CD-GRAB). As I remember it,
most drives _can_ read this extra data, but drivers often don't support it.
In fact, mixed in to the CD data stream are 8 extra single-bit streams,
known as A, B, C... H. CD+G probably (I forget) used the G stream, hence
the name. One of the other streams was normally used for timestamps,
IIRC, and the rest are free.
Hopefully someone will remember better than me so I don't have to dig
out my old documentation (I still have several yards of SCSI manuals for
CD-ROM drives as well as the CD standards, somewhere).
> Needless to say, you could always write or read data from the low bits of the
> audio too, provided you got a reliable read of that data... the software to do
> that could be considered suspicous though, while a cd that has a short text
> message imbedded in track #12 of a 20 track audio collection would be harder to
> detect (but of course for even vague security would have to be treated as a steg
> channel and encrypted in addition, with something decodable by hand like a book
> code)
Note that CD data is actually unreliable when used for audio - I
definitely saw the same CD show bitwise differences on successive reads,
much to my surprise, so if you want to do this, remember to include
error correction :-)
Cheers,
Ben.
--
http://www.apache-ssl.org/ben.html http://www.thebunker.net/
"There is no limit to what a man can do or how far he can go if he
doesn't mind who gets the credit." - Robert Woodruff
---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo@metzdowd.com
