[Bush2000]

I completely disagree. You're looking at this the wrong way. The point isn't that steganography prevents pros from extracting information from a particular source.

The strength of steganography is that (1) somebody can easily embed information in dozens (or hundreds) of hiding places, (2) you don't know who that somebody is, (3) because you don't know who's hiding the information, you can't trace where the pieces are, (4) there is all likelihood a practical time limit in which to make use of the information. As long as the information is sufficiently distributed, steganography is extremely effective. The only thing that would help a pro is if he correctly guesses the identity of a terrorist and starts sniffing everything that individual (or individuals) communicates over the Internet. Otherwise, the pro doesn't know where to look. See what I mean?

[Southack]

"The strength of steganography is that (1) somebody can easily embed information in dozens (or hundreds) of hiding places, (2) you don't know who that somebody is, (3) because you don't know who's hiding the information, you can't trace where the pieces are, (4) there is all likelihood a practical time limit in which to make use of the information. As long as the information is sufficiently distributed, steganography is extremely effective."

You missed my point. Steg itself offers little. Breaking up a file, however that is done (with or without steg), offers something.

But none of that matters if the final destination is being monitored. For instance, the public web server where photos are posted is an ideal place to monitor. If all of those pieces come in to one final location, it then becomes a simple matter to examine where all of the traffic originated.