Introduction
In the early times of digital electronic music copy protection was neglected: MP3 stormed the on-line world without DRM at all and people loved it. The main idea behind MP3, soon becoming the standard, was data compression (1:10) to help content providers and consumers to save bandwidth during music downloads. However this very feature also enabled illegal distribution. According to estimates, 3 million illegal MP3 files were downloaded every day in 1999 (Veridisc 2001). The deficiency of audio formats without content protection was soon realized, although it took quite a while for solutions to be developed. The main idea behind all these techniques is to encrypt the encoded (e.g. AAC) audio stream and store the decryption key in a sand-box (i.e. in a well-controlled environment) on the consumer's device. The music may only leave the hardware as sound waves or in the encrypted format. Decrypted data must not leave.

The current situation can be characterized by unprotected audio codecs (coder and decoder of audio signals from analogue to digital and vice versa often involving compression algorithms) on the one hand, which enable file sharing and easy distribution, and proprietary solutions on the other hand by e.g. Microsoft, Sony, Real and Apple, which are still not fully interoperable.

Before we enter into the debate about interoperability we will provide some background on audio formats in DRMs-protected and unprotected mode.

Audio formats overview
In order to better understand our categorization, first let's define two important terms:lossless and lossy compression. They both compare to the original CD audio quality. On a Compact Disc digital audio information is stored without any compression and therefore it consumes a large amount of storage space (1 minute of CD audio is about 10 MB of data), however CD audio offers superb sound quality. Lossless compression means that compression algorithms are used to reduce the storage space without any data (i.e. quality) loss. They typically reduce the size to 50%. On the contrary lossy techniques consider that the human ear has special characteristics that make the audio experience almost the same even if some parts of the sound are missing or are altered, this way a much higher compression ratio can be achieved, i.e. reduction to less than 10%. In the following we present the different popular formats, distinguishing between unprotected formats, DRMs-protected proprietary formats, and DRM-protected formats.

Unprotected formats
MP3 MPEG Layer 3, the pioneer in the field of audio compression, was developed by Fraunhofer Institute for Integrated Circuits (Fraunhofer IIS, Germany) more than 15 years ago. The main idea was to store audio information using "perceptual coding", a data reduction algorithm that is (almost) imperceptible to the human ear. The original solution achieved a compression of about 1:10. Virtually all music playing devices now support MP3.
AAC Advanced Audio Coding is the next generation audio compression algorithm, first introduced in MPEG-2 and now also incorporated in MPEG-4, the latest ISO/IEC standard of the Moving Pictures Expert Group. MPEG-4 is a complex specification defining a container for all kinds of media (i.e. audio and video), while AAC is the basis for natural audio encoding within MPEG-4. AAC was developed in order to give better performance over MP3 in compression while keeping or even improving sound quality (e.g. AAC fulfils the requirements for studio sound quality specified by the European Broadcast Union). AAC offers typically 1:16 compression ratio.
OGG Vorbis This is a compound solution developed by the Xiph.org Foundation, where OGG is the global container specification for containing any kind of multimedia data (just like MPEG-4), whereas Vorbis is the audio codec. The aim of Vorbis is the same as for AAC: to outperform MP3 by offering better compression ratio (i.e. over 1:10) while giving better sound quality. However, unlike AAC, which is commercially licensed, OGG Vorbis is free.
FLAC The Free Lossless Audio Codec is probably the newest contestant in this race of formats. The main rationale behind the sourceforge-hosted project is to provide lossless compression in a free product. The average compression ratio is about 1:2. DRM is not planned for this format by the developers.


DRM-protected formats
WMA Windows Media Audio is the proprietary solution from Microsoft for audio encoding. It is part of the Windows Media project (together with WMV, Windows Media Video). It supports several storage formats ranging from lossless compression to high-performance lossy compression and also voice encoding. The copy-protection of WMA is built on the Windows Media DRM architecture.
RealAudio It is the product offered by RealNetworks. The core focus of RealNetworks activity was traditionally on streaming media for which they achieve a compression ratio of about 1:16. The Helix DRM solution is part of the product.
ATRAC3 The Adaptive TRansform Acoustic Coding is the DRM-enabled sound encoding technology used by Sony and it is the successor of ATRAC. It achieves a compression of about 1:10, whereas its companion ATRAC3pro may go up to even 1:20. This format is used in Sony's MiniDisc or by the online shop Sony Connect.
FairPlay This is the DRM solution used by Apple's iTunes. The FairPlay offers protected AAC files in form of M4P (encrypted MPEG-4).
LWDRM The Light Weight Digital Rights Management is a new approach in the audio DRM field. Like MP3 it has been developed by Fraunhofer Institute. LWDRM currently supports MP3 and AAC, although in principle it could be applied to other formats too. The main idea of LWDRM is that there are LMFs (Local Media Files) to be used only locally, and SMFs (Signed Media Files) to be distributed. There are three levels of participation: level 1: you may only "read" SMFs, while you cannot create/modify anything; level 2: you may create LMFs. but these will be tied to your computer, and level 3: you may create SMFs but a signature (and watermarks) will be added to them that will identify you as the creator. The idea behind LWDRM is that the consumer may copy the content if he is willing to mark the media as his own. As long as the content does not leak out to the public, it is like using unprotected formats. But an illegal copy caught in the wild could be traced back to its originator. This approach is clearly an alternative to the existing encryption/key based solutions.


Discussion of interoperability
In this section we will put forward three arguments, why current approaches to interoperability are still deficient:

1) One might think any DRM solution could protect any kind of audio format, e.g. FairPlay could be used to encrypt, apart from AAC files (M4P), MP3 or OGG Vorbis files as well. This would be feasible in principle but would not solve the interoperability problem. Let's take OGG Vorbis, an open standard with published specifications, as example: Without DRM a compliant device simply decodes the data stream according to the definitions, and produces the sound output. However if some kind of DRMs was used, the result would not be OGG Vorbis any more and only devices fitted to understand the DRM solution would be able to play the content. Basically this is the main reason why currently only proprietary systems are used, where the chosen DRM solution can be enforced at the device level too. Finally, if the used DRM technology has to be enforced at this level, why bother about different formats? A single method is enough in a closed system environment.

2) While the approach of LWDRM is interesting and holds some promise to be applied to audio formats in a generalized way, we should not overlook one important issue. In tomorrow's world full of computer viruses, identity theft will be a key "black business". How can it be ensured that contents owned by someone won't be stolen when marked as their property and be held responsible for them (e.g. today's viruses are intelligent enough to send e-mails in the name of the infected computers owners, the next step is not that big)?

3) A third approach to interoperability of DRM-protected content could be interoperability of formats by conversion. The RealNetworks company recently introduced Real Harmony (Smith 2004a), which basically transforms its own copy-protected RealAudio files into other popular formats, this way allowing consumers to play their RealAudio songs also for instance on Apple's iPod, which was until now not possible. This can be seen as a step towards interoperability, but at the same time it can be interpreted as an effort to invade the domains of other companies (e.g. Apple-M4P, Microsoft-WMA). As this approach is very controversial, one may doubt that it is the best way to achieve interoperability. Notwithstanding, with this move Real has started the interoperability game, and we will see if others will follow (Smith 2004b).

Bottom line
It is safe to say that quality of audio formats is constantly improving - a clear benefit for the consumer. It is less clear how DRM-protected formats, which are backed by the record industry, will relate to free formats, which many people still prefer to use for the exchange of music files. The next big question is interoperability of DRM-protected formats. From the consumers' point of view, playing multimedia content on different devices (coming from different manufacturers) is an important requirement. Until now only hacker tools or nifty tricks allowed DRM protected content to be moved between devices from different vendors. Real Harmony is the first clear step in this direction by creating a solution for converting different DRM technologies, but it is not yet clear if this approach put forward by just a single company will be accepted by the entire industry concerned.

Sources

About the author
Gergely Tóth is a PhD student at the Budapest University of Technology and Economics in the SEARCH Laboratory where he currently co-ordinates a research project about mobile payment solutions in connection with DRM. Besides Digital Rights Management his core interests include security and privacy. You can contact him at gergely.toth@mit.bme.hu.

Status: first posted 24/08/04; included in INDICARE Monitor Vol. 1, No 3, 27 August 2004; licensed under Creative Commons
URL: http://www.indicare.org/tiki-read_article.php?articleId=34