Slashdot Mirror
Watermarking to Replace DRM?
An anonymous reader writes "News.com has an article on the announcement of Microsoft and Universal to introduce watermarking technology into audio files. The technology could serve several purposes including tracking file sharing statistics and inserting advertisements into audio tracks. The article goes on to suggest that watermarking could possibly replace DRM in the near future."
And there's certainly no watermarking that isn't noticeable that can't deliberately be removed.
I assume "They" want to catch the sophisticated pirates distributing tons of material, not the unsophisticated share-1-song with a friend people. Oh wait... that would make sense in a sane world.
If you use the watermark to trigger a banner ad on a player, it could convince the RIAA that there is an ad-revenue stream and cause them to drop the DRM and lawsuits.
Unfortunately it will be used to connect specific downloads to individuals allowing the RIAA to target their lawsuits more accurately. It will still be as impossible to prove in court but will drive an even deeper wedge between the RIAA and reality.
The only way the RIAA will stop suing is when someone wins a countersuit big enough to affect the bottom line of the corporations supporting them.
I have mod points. The reign of terror begins now.
The DMCA makes it illegal (or legally difficult) to remove DRM. But any watermarking and advertising is fair game...
In Soviet Washington the swamp drains you.
"You could even be legally responsible if a virus happened to alter the file"
I love the smell of FUD in the morning...
Seriously. Watermarks are progress. You disagree, that's fine, but lets debate it on its merits and not base our opinions on fear-mongering and FUD.
I don't whether this has ever been tried with audio files, but the techniques used in CDMA radio communications might work here. Essentially, you would need to add a small amount of noise to the audio signal, however it's not true random noise and can be decoded to reveal a signature, or watermark. If you combine two files with different 'noise' signatures, then both signatures can still be extracted with a high probability of a correct result. Only as you combine a large number of similar files does the probability of correctly decoding the signatures of the components decrease. However by that time, you've added a large amount of noise to the audio file and it will probably sound bad anyway, so no-one will want to download it.
The downside is that by definition the noise you add has to be audible. Note that for a long time audio cassettes sold very well despite their awful noise characteristics, so this may be acceptable to all but the strictest of audiophiles.
43 - For those who require slightly more than the answer to life, the universe and everything.
i think the hole point of watermarking is that you can not detect it, meaning you can not remove it by rewritting it. And neither by converting to analog and recording it again.
Should you succed in detecting one watermark, who says that they have not many watermarks in the sound? Can you prove that you have removed ALL watermarks from a file?
A sound file might be small enough for just having a few watermarks, but a movie should be big enough for each file to have several watermarks in it. Happy sharing.
Hmmmm what if pirates get multiple copies of an mp3? they examine the differences between the files, minimally alter all the places where the differences occur (i.e. making the average of the sound when in phase), or even worse make an audio collage of the song from watermarked mp3s stolen from unprotected computers so it includes marks from multiple owners.
---- MISSING MISCELLANEOUS DATA SEGMENT --- [sigdash] trolololol