The discs in question are all from major recording company EMI or its various imprints and labels distributed by it; these include labels like Mute, Rough Trade, Modular and Ministry of Sound. Long beset by poor sales, EMI's new management have decided to do what Universal promised to do but failed to follow through; that is, to "copy-protect" all their new releases. The Napster kiddies, unable to pirate them, would be forced to shell out their pocket money and buy them, profits would go up, investors would be happy and EMI's future would start looking a lot brighter. Or so the theory went.

The copy-protected CD fallacy

Of course, it's not quite as easy as all that. The goal of "copy-protecting" CDs in a way that users can play them on CD players but cannot make digital copies is a tricky one. The audio CD format is specified in a document known as the Red Book, written in 1980 before the advent of personal computers and CD-ROM drives, and is an extremely simple format. An audio CD is essentially a stream of 16-bit samples with a table of contents, intended to be read as is. As such, any copy-denial method which attempts to prevent the discs from being read on PCs whilst maintaining compatibility with simple CD players has to rely on idiosyncracies in the ways computers and CD players access non-compliant discs. By definition, a "copy controlled" disc is a CD whose internal structure is defective in a way intended to confuse most CD-ROM drives whilst playing perfectly well in virtually all CD players. As you can probably guess, it's easier said than done.

Copy Control on test

For the purpose of this article, we acquired a number of discs bearing the "Copy Controlled" logo, and proceeded to test them with various equipment, hardware and software.

Firstly, not all such labelled discs were copy-protected. One disc (West Papua: Sound of the Morning Star) bore the logo but was an ordinary Red Book-compliant CD, with no data section. Another single had the special EMI Windows Media player on its data portion but was otherwise a normal CD. The other discs contained various permutations of EMI's copy-denial technology. This involved corrupted low-level format data as well as a data section set to automatically run on Windows PCs, launching some sort of player application. The visible files on the data sections of several different discs were identical; however, it is believed that the section contains hidden data in Windows Media format.

The scheme used by EMI appears to involve corrupting the CD's table of contents and/or the low-level structure of the disc. Dumb CD players, in theory, ignore the corruption, just playing the next part physically on the disc, while the smarter units, designed to deal with non-audio CDs, fail to read the disc properly. That way, legitimate consumers can play the discs on CD players, but can't rip them on their computers. So, how well does it work?

We attempted to play and rip the discs using various CD-ROM drives and software. Read on a PC running Linux with a Lite-On LTN486S 48x CD-ROM drive, the discs looked like perfectly ordinary mixed-mode CDs. Looking at them with CDfs, the only visible anomaly was that the data portion was 650Mb in size, and appeared to encompass the entire disc. In the Grip CD ripping application, the anomaly disappeared; the data track merely took up the entire rest of the disc after the audio tracks.

We attempted to extract the audio tracks from the discs using ordinary CD ripping software, such as grip and cdparanoia, and we were successful.

The Empire strikes back

We then attempted to access the discs on different hardware: a Pioneer DVD-115 DVD-ROM drive, and hit a snag. The copy-denial technology which had lain dormant reared its head and bit us. Using the same software, the disc's table of contents was garbled; some tracks appeared normal, others were unusually short or impossibly long, or started after their end. What's more, the randomness changed each time the disc was mounted. It was impossible to rip the disc using the DVD-115 drive.

We tried accessing it on a Sony CRX175A1 CDRW drive as well, and had different problems. The drive would fail to recognise some discs as valid; when it did, the table of contents was uncorrupted. However, audio extracted from the CD tended to jump all over the disc, consisting of chunks from both tracks. (Mind you, given that the CD in question was a Lisa Marie Presley single, one could argue that that was an aesthetic improvement. I wonder whether Radiohead fans will complain when their newly-ripped MP3s sound even more glitchy than the original, but I digress.)

We then attempted ripping the CD in question on an Apple Macintosh G3, using Toast Audio Extractor on MacOS 9.1. Both tracks ripped and played perfectly, without any sign of hindrance.

The drive's the trick

Finally, we decided to perform a further experiment. We removed the CD-ROM drive from the Macintosh and placed it in the PC, replacing the DVD drive with it. (The Macintosh CD-ROM drive is a Matsushita CR-585, an extremely common generic 24X IDE CD-ROM drive from a few years ago; if you bought a beige-box PC in 1997 or so, it probably contained one of these.) We then attempted to rip the CDs under Linux, as before. This time, the process worked perfectly.

Non-computer devices

We also tested the discs with several consumer electronics devices. We attempted to play a disc (David Bridie's Hotel Radio) on a Hiteker AD-660 DVD player (believed to use a DVD-ROM-type drive); the disc played without problems. We also tested the disc with a Digitor CD/MP3 player, a portable CD player capable of reading and playing CD-ROMs containing MP3 files. The unit played the disc perfectly, but only when anti-shock protection was switched off; when it was switched on, it would stop after the first 10 seconds of each track; presumably this is due to the difference between the ways the unit handles CDs in the two modes, though this will provide little comfort to joggers.

Conclusions

The main conclusions of this experiment are: firstly, that EMI's copy-restriction scheme involves corrupting the low-level data on their discs as to fail to work with some players. Secondly, that the ability to play or rip the discs depends wholly on the player or CD-ROM drive used, and not on software or computer type. And thirdly, that the newer or better your CD-ROM drive is, the more likely it is to fail to handle these discs; for now, at least.

So, the question you've undoubtedly been asking yourself: how do you rip your brand new EMI discs to MP3s? The short answer is: the way you would any normal CD. If it works, you're in luck; if not, go out and buy a cheap CD-ROM drive or dig one out of an obsolete computer. There's no guarantee that it'll work better than your shiny new unit, but if it doesn't, you've just lost $20, which is probably less than you paid for that Coldplay album.

Every Mistake Imaginable: unintended consequences of Copy Controlled discs

One thing that's apparent is that copy-restricted CDs don't solve any technical problem. By the very nature of the CD technology, it is impossible to devise a CD copy-denial scheme that can be played by the vast majority of "legitimate" consumers and yet block digital copying in all but a negligible number of cases. EMI's scheme is strong enough to be seriously annoying if you use MP3 files and have a CD-ROM drive susceptible to it -- but utterly ineffective otherwise. There are enough gaps that MP3s will get out there, and not only a few times.

Pushing an ultimately ineffective means of copy-denial on the public is fraught with the hazards of unintended consequences. The obvious hazard is antagonising consumers, eroding what little sympathy the recording industry has left to call on from the public. And consumers who feel motivated to "stick it to the man" will go the extra mile to do so, Secondly, a more subtle effect could occur when some people can rip copy-protected albums. If Alice and Bob are friends and both want the new Blur album but only Alice's computer can rip it, Bob can take his copy to Alice's place and ask her to rip it for him. If such a situation comes about, it would be naïve to assume that both Alice and Bob will necessarily do the right thing by the much-reviled recording company and each shell out $27 to buy their own copy of the album.

One could argue that the problem will go away when devices capable of ripping CDs are taken off the market. EMI (or the RIAA/ARIA or IFPI) could launch a campaign to buy up and destroy all CD-ROM drives known to rip copy-controlled discs; soon the drives would become scarce, fetching prices of thousands of dollars on Ebay. Eventually, the remaining ones would die of natural causes and the New World Order would be complete, with consumers having to accept the company's rules or go without music. Or so the theory goes, anyway. After all, there are virtually no consumer-grade VCRs which are immune to Macrovision out of the box.

The fatal flaw with this is that the VCR analogy doesn't hold. Certainly, EMI could deplete their dwindling cash reserves destroying 24x CD-ROMs, but it wouldn't do them any good if the hardware companies start making new drives immune to copy-denial techniques. And there is no reason why they wouldn't; unlike VCR manufacturers (who are bound by the VHS patent licensing agreement to comply with Macrovision), CD-ROM manufacturers have no obligation to heed copy-protection techniques, especially since Philips (licensors of the CD standard) have publicly stated that non-Red-Book-compliant discs are not legitimate CDs. And market competition will ensure that manufacturers will ship CD-ripping drives, one way or another; for proof of this, one only needs to look at the DVD player market, and the number of players that can be easily adapted to bypass region coding.

In short, whichever way you cut it, EMI's "copy control" scheme appears to be a mistake. It annoys legitimate consumers, does little to reduce the supply of unencrypted MP3s, and, in discouraging people from buying the discs, is unlikely to boost album sales. If EMI have a clue, they'll scrap the whole idea. Personally, I'm hoping they do so before the new Radiohead album comes out.