DVD Jon’s DoubleTwist
Jon Johansen (known for his work in defeating CSS in DVDs) has recently released an application known as DoubleTwist. DoubleTwist is designed to defeat Apple’s FairPlay DRM by playing the track’s in fast forward and capturing the analog signal as it leaves the sound card. In addition to stripping the DRM from songs, it allows users, through a Facebook application, to share songs with their friends.
Circumventing FairPlay in this manner arose due to the “Analog hole” that exists for most media in existence today. The bottom line is that for human senses to perceive the media in question, it has to be in an analog format with no protection at some point in the playback. Since there is no reasonable way to prevent this kind of attack on DRM, it seems it would be more beneficial to be able to detect if the attack had been performed and pursue the individual in question. If the media carried a watermarking code that was durable enough to withstand copying and most removal attacks while still identifying the source user, most people would be unwilling to share their media lest they be sued for copyright infringement.
Copyright law has been another issue that’s been complicating the use of DRM. In particular the idea of fair use is rather murky when it comes to implementation in DRM. In the US, it’s generally considered fair use to make back-up copies of media that you own however there is little to prevent people from passing on their backup copies to other people.
Apple’s long term response to this development will be to update their DRM scheme to prevent fast-forwarding of media to slow the copying of music protected by FairPlay. In the interim however they will likely try to find some legal recourse against the authors of DoubleTwist and possibly Facebook for allowing the application to be used. My hope is that Apple will do neither of these things and instead abandons DRM in favor of watermarking.
Source
http://www.telegraph.co.uk/connected/main.jhtml?xml=/connected/2008/02/21/dlhack121.xml
Comment by alpers
February 25, 2008 @ 12:24 pm
afaik, QTFairUse has been bypassing restrictions like this by also using analog playback, and also does it in an automated way like you stated. The playback speed is expedited however, in a way so that an entire 4 minute song will take around 30 seconds to capture the outgoing analog/digital audio signal and write it to disk. QTFairUse has been around a lot longer, looks like DVDJon’s just driving attention against DRM. 🙂
Of course, the main reason that consumers detest DRM is the inability to put their purchased media on any player possible, like putting purchased iTunes music onto any non-iPod media player. Watermarking would allow more compatibility in this case.
Comment by nekret
February 25, 2008 @ 5:42 pm
QTFairUse has been around longer however it’s method of operation is slightly different than DoubleTwist. From what I understand, QTRFairUse actually grabs the output AAC from the FairPlay decryption and is thus a lossless method. DoubleTwist grabs the audio after the AAC has already been decoded.
Despite the claimed 5% quality loss from DoubleTwist, I’d argue that it’s probably a superior solution since it likely won’t need an update every time Apple releases a new version of iTunes. I’m guessing this is the main reason why QTFairUse doesn’t support anything past iTunes 7.5.
Comment by Kris Plunkett
March 2, 2008 @ 11:36 pm
I wanted to add my input on the idea of watermarking digital media files with the intent of identifying the individuals involved in piracy. Although I haven’t read too much about it, I see two major flaws at first glance:
1) Watermarks can be removed from digital media for the same reason that media in that form is, and always will be, inherently copyable: At some point all DRM, watermarking, encryption, and/or encoding needs to be removed and the raw bits exposed for a media file to be played back. The trick is finding where those raw bits are produced and then capturing them, as QTFairUse and DoubleTwist do. This would effectively eliminate all dirt on the media file, including watermarks. Perhaps, however, if operating systems designers cooperated with media distributors, they could conceivably derive a system where the OS implements measures that prevent any user from accessing those bits when they appear in the system.
2) The second major problem is related to the issue regarding the RIAA using IP addresses to identify media pirates. Just as IP addresses could be spoofed, thus falsely making it appear that someone is downloading a file when they in fact are not, watermarks could potentially be tampered with to make it appear that someone other than the actual pirate is performing the illegal copy. This could inject sufficient doubt to make such watermarks not hold up in court.