The Digital Media Project  
Source L. Chiariglione
Title Television and Digital Rights Management No. 050613chiariglione01

 

Television and Digital Rights Management
 

Executive summary

The paper considers the difficult relationship of television with digital technologies and advocates a flexible use of Digital Rights Management (DRM) technologies as they are being developed by the Digital Media Project (DMP). The main features are: interoperability, toolkit approach and accommodation of Traditional Rights and Usages (TRU).

Table of contents

Executive summary

Table of contents.

1      Introduction.

2      Analogue television

3      Digital television.

4      Behind the figures of success

4.1       Audio distribution

4.2       Movie distribution

4.3       Pay TV.

4.4       TV in the home

4.5       IP TV.

5      The solutions.

5.1       Conditional Access

5.2       Multi-CA set top boxes.

5.3       Authorised Domains

5.4       New ways of managing content

5.5       Media for all

6      Many views.

7      Can we reconcile all these views?

8      Interoperable

9      Making DRM interoperable.

10        The DMP results so far

11        The DMP specifications Phase I

11.1     Use Cases.

11.2     Architecture

11.3     Interoperable DRM Platform.

11.4     Value-Chains

11.5     Registration Authorities

11.6     Terminology

12        Beyond technical specifications

12.1     Reference software

12.2     Conformance

13        Adapting to the new environment.

14        Interoperable DRM Platform, Phase II.

15        How can DMP help television?

16        Conclusions.

17        References

18        Acronyms.

1          Introduction

Ever since technology was introduced to help creation, distribution and consumption of media there has been a tension between those who own “rights to content” – usually called “rights holders” –, intermediaries in the media value-chain and end-users. Still the technologies in use until some ten years ago, that we will call “analogue” were usually so unwieldy that for each new, usually destabilising, technology a new equilibrium point could be reached, sometimes with the help of legislation and the courts.

The explosive combination of Digital Signal Processing, Personal Computers and Digital Networks – the “digital technologies” – have empowered intermediaries and end-users to do more with media than they could ever have imagined before. The new situation has exacerbated the contrast between the three classes of “users” of the media value-chain. On the one hand rights holders enforce the rights that have been granted to them by successive legislations, including those granted by recent legislation stemming from the World Intellectual Property Organization (WIPO) Copyright Treaty [1], introduced to cover the use of digital technologies. On the other, intermediaries and end-users try to cling on rights and exceptions that were traditionally assigned to them, whose applicability to the new digital space is often disputed by other parties affected.

The result is that in the last ten years the great promises of digital media have not been delivered. What we are witnessing is on the one hand the continuous looting of rights holders’ assets that seriously affect their economic viability and on the other the ongoing deployment of media distribution that is based on Digital Rights Management (DRM) technologies that limit, in some cases substantially and always unilaterally, the Traditional Rights and Usages (TRU) of media users.

In this situation we see public authorities either shying away from an issue that is admittedly thorny but critical for the future of society, or boldly getting into it only to retreat after declaring victory but actually without even scratching the surface of a solution. In this vacuum other public authorities – the courts – are busy creating the legal framework of digital media by applying laws or court cases that were the result of just another age.

The field of civil liberties advocates is obviously very involved in this issue but in most cases these organisations are fighting a rearguard battle. As soon as there is major clash between the use of digital technologies or regulation with TRUs, these organisation side with those affected by the technology or regulation and stop there. There is no doubt that this role is important in a democratic society, but it is a far cry from the what is really needed, i.e. actively proposing a way out of this stalemate.

Proposals do come from some quarters but most of them are hopelessly idealistic. Indeed it would be great if a gifted musician could sustain himself by giving away his music because he can make money from his real time performances. The problem is that bringing a couple of cases where this has happened does not prove that this is a universal solution to a complex problem

Two years ago the author of this paper launched the idea of the Digital Media Manifesto (DMM), a grass root initiative to discuss the digital media situation outlined above. In less than three months the movement published the Manifesto [2] where a solution to the stalemate was proposed with the establishment of a not-for-profit organisation called the Digital Media Project (DMP) [ REF _Ref105735918 \r \h 3]. The goal of DMP should be the twin development of technical specifications of interoperable DRM and the practical means to support TRUs in an interoperable DRM framework.

Today the DMP can claim to be on its way to achieve the goals that were assigned to it by the DMM. In April 2005, less than 18 months since its establishment, the first interoperable DRM specification was approved, the first working draft of the second interoperable DRM specification was in development and a Call for Contributions on “Mapping of TRUs to the digital space” was issued. The intention is to publish the second interoperable DRM specification in October 2005 and a Recommended Action on TRUs in January 2006.

This paper is about the problems faced by television in its drive towards incorporating digital technologies and how the standards developed by DMP can be used to alleviate and in some cases resolve the clash of views of the different players brought about by digital technologies thanks to the availability of interoperable DRM specifications which support Traditional Rights and Usages.

2          Analogue television

Television is one of the most successful forms of media distribution. First deployed 70 years ago using the Very High Frequency (VHF) bands it began expanding rapidly in the USA in the late 1940s and in Europe and Japan in the 1950s occupying portions of the nearby Ultra High Frequency (UHF) band. The VHF and UHF bands are also called “terrestrial”.

The use of the new medium has always been under the influence of public authorities. One stated reason was the need for each television channel to use a large chunk of the most obvious example of limited real estate, the radio spectrum. The other was the ability on the part of the television station to influence TV viewers’ opinions by exposing them to suitably selected information sources.

For these reasons in many countries television was originally run by the state or by a state agency while in the others running a television station required, as a minimum, a license from the state. Starting from the USA, cables were progressively used as a means to extend the patchy coverage of VHF/UHF television, as shown by the original Community Antenna Television (CATV) name that eventually became in many countries an audio-visual distribution system in its own right. Note how CATV still required a franchise to operate, even though the “scarce spectrum” justification did no longer apply.

The commercial exploitation of the huge spectrum assigned to television in the 7 GHz (upstream) and 11 GHz (downstream) bands opened the way to satellite television. The extended bandwidth lent credibility to the ideas of those who advocated “higher picture and audio quality” where the West (Europe) and the East (Japan) trod diverging paths: better 625 lines in Europe with Multiplexed Analogue Components (MAC) and higher resolution (1125 lines) in Japan. Both attempts never achieved significant market interests.

Satellite television broadcasting happened at about the same time as others experimented with a different type of “higher picture and audio quality”, i.e. better content provided through a selected form of television distribution in the terrestrial bands called pay TV, as opposed to public or commercial television broadcasting . Pay TV was enabled by the “scrambling” technology that only allowed users with a special device to receive the scrambled television programs. The same technology was also used by CATV.

Pay TV was also prompted by the much larger availability of television content compared with the early days of television. One such source of content was movies that attracted more and more viewers away from theatres. The keen consumer interest in this type of content also prompted the consumer electronics industry to develop an alternative distribution channel of television content – the video cassette recorder (VCR). In spite of the understandable fears that premium content would leak out because of the duplication capability of the VCR, today the movie industry earns more revenues from the home video distribution channel than from movie theatres.

3          Digital television

The business of television was, and is still being, shaped by spectrum scarcity, for good or for worse. Therefore it is no surprise that the different facets of the television business were fancied by digital television because of its promise to allow a much more efficient – at least 4-5 times at that time – spectrum utilisation.

Digital television required two main technologies: the first the reduction of the number of bits required to represent the video information, from 216 Mbit/s of a standard television signal to something in the order of 5 Mbit/s, including digital stereo if not multi-channel audio. The second technology was modulation, i.e. a system to adapt the stream of bits to the specific features of the physical medium and frequency spectrum.

The different initiatives that various countries and industries had in place for the former – video compression – were wiped away by the ISO/IEC MPEG-2 standardisation initiative. The latter – modulation – remained a patchwork of solutions that depended on the type of physical medium (terrestrial, satellite, cable etc.), industries and countries/regions.

In its 10 years of existence digital television has changed the television distribution landscape beyond recognition, but the impact has not been uniform.

  1. The Digital Versatile Disc (DVD), a system built by extending the Compact Disc (CD) technology and adding the MPEG-2 video compression technology, has been hugely successful, both in term of hardware and software devices and amount of content released through the medium.
  2. The CATV industry in some countries has converted its infrastructure to digital thanks to the availability of a standard digital television system.
  3. The pay TV model has been boosted by digital television and the number of receivers and programs offered worldwide is staggering.
  4. The number of digital television programs in the clear that are broadcast from satellite amounts to several hundreds using tens of languages.
  5. Digital terrestrial broadcasting (DTT) has started in quite a few countries
  6. The television production landscape has also been changed with the massive introduction of digital television technologies.

MPEG-2 was the enabling technology for digital television, but the high bitrate (at least 3-4 Mbit/s for reasonable quality) made it unsuitable for the new form of delivery made possible, especially in the early days, by the Internet. The compression and transmission systems of Real Networks, Microsoft, Apple and MPEG-4 made off-line and real-time delivery of audio-visual content to the desktop PC a reality that has grown to millions of users.

The maturity of the digital video delivery technology on the Internet has given new impetus to the old dream of telecommunication companies to extend the use of their infrastructure to the carriage of video content by providing “on demand” services. Several attempts have been made, starting from the early MPEG-2 days, but today the inexpensive Internet Protocol (IP) network technology, coupled with the maturity of low bitrate digital video systems, is promising to make an old dream real.

4          Behind the figures of success

Digital television has been the most innovative factor in the television business of the last ten years. It has brought undoubted advantages to some industries, but also new challenges, especially because the use of digital technologies is shared by many industries.

4.1        Audio distribution

Audio has been the first to bear the brunt of the explosive technology mix of digital technologies.

The ability to use a PC to rip a CD and create an MP3 (MPEG-1 Audio Layer III) file more than one order of magnitude less than the original, create compilations at will, share music with others in small groups and with the entire world using Peer-to-Peer (P2P) networks is undermining the established business model of record companies. Using the arm of the law to curb the phenomenon has proven so far a Sisyphean task.

4.2        Movie distribution

The predicaments of the music industry are now extending to the movie industry. The increased bitrates of digital networks, both in the local access and in the core network, make the time to move a DivX file on the Internet as short as it was moving an MP3 file 5 years ago. Even though sharing movie files is still not as widespread as sharing music files, the implications on the business model of the movie industry is clear: potentially anybody can watch the latest movie as soon as it is released somewhere on the face of the Earth.

4.3        Pay TV

In some countries Pay TV has been a successful business model at the time of analogue television, but its implementation as digital pay TV has been fraught with difficulties.

In the UK DTT first started as two competing pay TV operators, then merged to one then bankrupted. The current UK DTT offers a selection of TV programs in clear.

Most satellite Pay TV ventures turned sour because it was impossible to sustain two competing operators in most markets, such as in Europe, because of the high cost of content rights and the pay TV set top boxes.

4.4        TV in the home

In the analogue age distribution of television in the home was always a nightmare, but today different forms of digital distribution technologies, via radio and cable, promise to make the dream of seamless distribution of all types of content in the home and consumption on any device real.

This “end-user’s” dream, however, may turn out to be the content and service provider’s nightmare. Indeed, the common availability of PCs in a growing number of homes in developed countries offers the possibility to mix, mash and share all sort of content distributed via different broadcast channels. But with today’s business model such content that retain value as long as its use is within the scope envisaged by the content or service provider.

4.5        IP TV

Many telcos hope that now is the right time to make inroads in the long-coveted business of content services on top of their traditional infrastructure services. Indeed the provision of a bitrate of a few Mbit/s required by compressed video (MPEG-2, MPEG-4 Visual and AVC) to a sizeable number of their subscribers is becoming a matter of fact in many countries.

But this is the easy, technical part. Unfortunately an IP-friendly content delivery solution plays havoc with the “TV in the home” concerns expressed above, even more than with other content delivery systems.

5          The solutions

Given the existence of such high-profile content “health problems” it can only be expected that “doctors” abound. Here we would like to see the medicines that have been offered to treat the set of problems highlighted above.

5.1        Conditional Access

Conditional Access (CA) has been the principal enabler of the majority of business models exploiting digital television. It is based on an infrastructure of devices that are capable of decrypting an encrypted digital bitstream thanks to the keys embedded in set top boxes or smart cards.

CA provides an easy way of setting up the means to securely deliver content to one’s subscribers but entails a number of problems

  1. It requires a high level of expenditure to create one’s subscribers base because of the need to put in place an independent technology infrastructure for secure content delivery (set top boxes and smart cards) in addition to the usual content rights and subscriber acquisition and management costs
  2. The break even point can only be achieved with very high numbers and virtually inhibits competition between operators in small-to-medium size markets as shown by the UK and Italian cases.
  3. There are concerns about the discrimination felt by subscribers to one pay TV service when high-profile (e.g. football matches) content is only available from another service.

5.2        Multi-CA set top boxes

In this solution a service provider selects a CA system but is not saddled with the task of creating his own independent technology infrastructure for secure content delivery (set top boxes), only the much less onerous smart card delivery infrastructure.

From the end-user view point this solution is a great progress compared to the single-CA system solution. However, its value in the context of a growing number of content sources through different delivery systems, e.g. the Internet, depends on the number of services that end users can access using their multi-CA set top box.

5.3        Authorised Domains

To allay the fears of misuse of television content on the part of end users there have been proposals to extend the sealing of the television distribution channel inside the home by creating groups of devices affiliated to a domain. The domain would be managed by an external entity that would oversee the flow through and use of content by the different devices belonging to the domain.

This is a solution that would certainly satisfy the desire of content and service providers, even though it would not solve all problems, e.g. the so-called “analogue hole”. Whether end users are ready to accept en masse this solution is a matter of speculation. For example the technology is prone to enabling a further segmentation of distribution by delivery systems and by service providers within delivery systems.

5.4        New ways of managing content

The world of media has developed on the basis of increasingly tight management policies, e.g. those represented by the Berne Convention [4] and the Universal Copyright Convention [5] and lately on the basis of increasingly tight protection policies, e.g. those represented by the WIPO Copyright Treaty.

The massive introduction of digital technologies in media content has brought in aspects of the libertarian culture that is part of the information technology world, emblematically represented by Linux, the Open Source Software (OSS) [6] that is being increasingly supported by different companies from different industries. The idea is that there are certain components of any business that should be shared, because they enable the business of the industry, while other parts can or should remain proprietary.

A proposal stemming from this environment is Creative Commons (CC) [7]. This is based on a set of licenses that provide the means for authors to let others use their works, without losing control of them, typically for non-commercial uses. While CC have several positive sides, it is not clear how it can coexist and link to more traditional content management and control approaches to media.

5.5        Media for all

In the spectrum of possibilities that range from tight management and protection to progressively less constraining systems it is important to mention other ways of delivering content that are much less controlled (from the rights holder’s view point) and accessing content that are much more liberal (from the end user’s view point). In this case the proposal is that content should basically be available to anybody, say, within a given jurisdiction, because creators would be paid for their works on the basis of some scheme usually called “Alternative Compensation Schemes (ACS)”. See [8] for one proposal.

Clearly it would be great if people could freely create and freely donate their creations and users could freely access them. This could indeed work if everybody had a fat bank account or were a tenured professor. Unfortunately this is not the world we know. Many artists struggle to make ends meet and work hard to get to success, if ever they get to it.

6          Many views

The world of television summarily described in the pages above is undergoing profound changes under the influence of three classes of value-chain users: content rights holders, distributors (i.e. service providers) and end users.

A simplified classification of content rights holders could be

  1. Movie companies. They are very conservative and advocate very tight management and protection of their content throughout the value chain
  2. Content creation businesses of television companies. They are obviously concerned by what further use is made of their content but they are more flexible in the type of management of content they need
  3. Small-to-medium content creation companies. They are in search of an outlet for their creations and can be very flexible in their use if this is a means for their content to achieve notoriety
  4. Private content creators. People who produce video to distribute private information, e.g. in the form of video blogs ("vlogs").

These are the principal classes of distributors:

  1. Public television. In many countries they are the original providers of the television service and they are stuck between their “public service” legacy that includes promotion of national culture and the need to manage the value of their assets (see BBC’s Creative Archive project)
  2. Commercial television. They have less restraints in letting end users play with their content but are still squeezed by contrasting needs: making their viewers happy, their advertisers happy and their content providers happy
  3. Pay television. They tend to be more conservative in the use of their services as a result of them beings providers of service with premium content
  4. Community Antenna Television. Their feet are in today’s business of providers of television services and their mind in what they see (and in part already is) their future role of Internet Service Providers (ISP) and gateway to a world of interesting content
  5. IP television. Their feet are in today’s ISP business and their mind in what they see their future role of providers of television services on demand
  6. Private publishers. People who exchange and trade video via P2P networks (e.g. BitTorrent)..

These are the principal classes of end users:

  1. Couch potato. With so much talk about free this and free that, we should not forget that there are still many people who just watch television as their fathers did
  2. Interactive. Television is progressively becoming less a “lean-backward” experience and more a “lean-forward” experience
  3. Mix, mash and share television. These are the people who would like to have a free hand to do play with a great source of content: television
  4. “Free” content. This is a more extreme version of the preceding class.

7          Can we reconcile all these views?

It is a hard job to put in the same game those who want to tightly manage other users’ sphere of action and those who want free access to content, as part of their citizenship rights. There is a technology satisfying the requirements of the former group that is called Digital Rights Management (DRM) but the latter group is vehemently opposed to it.

The problem is that, as for everything that is new, and on top of it, technically hard to grasp, there is a lot of confusion with this term, particularly when those who could don not do anything to clear the air. Therefore it helps to introduce a definition. The one quoted here has been developed by the National Institute of Science and Technology (NIST) in the USA:

DRM is a system of IT components and services, corresponding law, policies and business models which strive to distribute and control Intellectual Property and its rights.

The definition is not perfect because it bundles together the two separate functions of “management” and “protection” that is one of the roots of confusion. This separation is spelled out in the term “Intellectual Property Management and Protection” (IPMP), which is used within the MPEG group [9]. Even though more confusing, in the following we will use the more common term DRM and NIST’s definition, but it is important to stress again that DRM does not necessarily mean protection, which does not mean that protection is excluded.

Actually the only thing that the NIST definition says is that there are “rights” to “Intellectual Property” (true that some dispute the existence of any such “right”) that is “distributed” and the rights holder wants to retain “control” of it. That this is a pretty neutral definition can be seen from the fact that Creative Commons, which also has a form of license that is machine readable, utilises a form of DRM because it “distributes and controls Intellectual Property and its rights”. In the following we will called “governed content” the one that has some form of “control” whatever is the practical form the control takes.

The definition of DRM given above implies that in general there is a chain of users to which the rights holder entrusts his IP. Arguably all types of business entail the creation of a value-chain, but value-chains are particularly relevant in the case of media because they are created to “connect” those who produce with those who consume intellectual creations. In this paper we will call media value-chains those designed to achieve this goal.

A value-chain is a special type of communication system linking different users of the value-chain and DRM is a technology that is added to the system to enable management and control of the flow of information through it. As for all communication systems, if the specification of the technology is known and open, anybody can communicate with anybody else, possibly with the payment of a fee to a third party. If the specification is not known or, if it is, is not open, and the communication system is owned by a third party, this party becomes a gatekeeper.

If a value-chain or a portion of it employs a DRM system that is closed or cannot be independently implemented (that from now on will be called “non-interoperable”), it is very easy to keep the barrier to access high and the control of distribution tight. But if a value-chain or a portion of it employs a DRM system that is based on open specifications that can be independently implemented (that from now on will be called “interoperable”) the advantages of a low entry barrier to creators and low access barrier to users that was demonstrated by such social phenomena as MP3 and DivX can be combined with the possibility to control the use of digital media, a feature that is known to work to preserve the creators’ incentive to continue creating.

8          Interoperable DRM

Lack of interoperability is one of the reasons for users’ reluctance to accept DRM systems and is, or has been, of concern for public authorities, at least in Europe. After the adoption of the European Copyright Directive, the European Commission asked the Comité Européen de Normalisation (CEN) to look into DRM issues, with priority to interoperability. A group of major stakeholder representatives was set up and a report issued in September 2003 [10]. The report can hardly be taken as a model for practical recommendations as can be seen from the following sentence drawn from the conclusions:

The solution suggested is regular and informal discussions between all the stakeholders in order to hear the concerns of each party and progress in the debate, whilst examining the market evolution.

The document is, however, useful to get a sentiment of how interoperability is seen by different market players. In the following some of the opinions expressed are reported assigning them to possible sources.

Rights holders

  1. The ability for content and rights usage rules to be supported, unambiguously interpreted and enforced across multiple proprietary DRM systems and end user devices.
  2. The portability decisions should rest with the content provider who should have the right to determine the range of devices and domains on which the content can be accessed. As the content provider makes choices in terms of his licensing models, then the marketplace (through the end-user) will reward or penalise accordingly.
  3. At present interoperability is not the primary barrier to the uptake of DRM. Interoperability must be guaranteed but market forces should resolve this issue at a later point via global, open, voluntary technologies. The development and proof of different business models is more of an issue.
  4. A certain level of interoperability is important to prevent the establishment of new gate keepers for content delivery.

Intermediaries

  1. The ability to use datasets from different origins as though they were built to a common standard (e.g. for using metadata from different communities).
  2. It should be possible to have a consistent and predictable interface when accessing content from different suppliers in the same sector.
  3. There is a need to listen to what the end-user is saying and to study how the market is reacting to what it is being told because the future of the industry lies with its ability to reach its end-users. So far the lack of interoperability has been a major obstacle encountered by those who have used DRM in their solutions offered to end-users because of the confusion created by multiple choices, system complexity, novelty of dealing with possibly radically different systems and impossibility to move content and enjoy it on multiple platforms

End users

  1. It should be possible to use the same software or hardware for content from different suppliers.
  2. It should be possible to purchase content that is not tied to a specific device, i.e. content that can be used across devices, such as to enjoy a song purchased via a cell phone on a PC or a car stereo etc.
  3. If the end-user cannot access/use/shift the content he has bought with the same usage rights then he will either prefer to get it through other more traditional or less legitimate distribution channels

9          Making DRM interoperable

There is growing awareness that interoperable DRM is needed, but how near are we to having one? The answer is very simple: it’s being done by the Digital Media Project.

As stated in the DMP statutes [11], the basic DMP position is that digital technologies are an asset of mankind that should be used so that creators, intermediaries and end-users all benefit from:

…DMP promotes the development, deployment, and use of digital media that safeguard the rights of creators to exploit their works, the wish of end-users to fully enjoy the benefits of digital media and the commercial interests of value-chain players to provide products and services.

The means to achieve this goal is to provide an open DRM specification, i.e. to achieve standardisation of appropriate protocols between value-chain users supporting the functions they perform, as per DMP definition of DRM interoperability:

The technical ability of value-chain users to perform functions through interfaces and using protocols of open specification that can be independently implemented and provide predictable results.

Converting the general DMP objectives into a work plan that can be practically implemented is not simple. The first problem is that there is no such thing as a “universal DRM system” to develop a standard for simply because there is no such thing as a “universal way of doing business with media”. We can only think of a range of implementations of DRM systems that satisfy the needs of specific value-chain users.

The second problem is that digital technologies have forced changes on media value-chains and are likely to keep on doing so. Therefore it is impossible to standardise functions performed in existing value-chains as we do not know how today’s value-chains will evolve in the future. It is even more difficult to standardise functions that will be performed in future value-chains, as the shape they will take is simply anybody’s guess.

What it is possible to do, however, is to standardise protocols for functions at a more atomic level – which DMP calls primitive functions – which are executed between value-chain users. Examples are offered by identification of content and devices, expression of rights to content, authentication of devices, access to content etc. Functions performed by value-chain users are obtained through combinations of primitive functions.

Standards are useful because they provide interoperability to users today, but in developing a standard one must make sure that it will be possible to continuously inject innovation into the system. The ability to combine primitive functions to create new functions ensures that the system remains open both to the introduction of new primitive functions and therefore to new roles in the value-chain. The DMP approach ensures that new technologies can be developed and standardised to enable new functions to be performed in an interoperable fashion.

Currently DMP is developing a series of toolkit standards. These specify component technologies - that DMP calls tools – which support the implementation of interoperable primitive functions. User of the standards can build solutions that suit their requirements by picking the tools to support the functions of their interest.

There are several benefits with a toolkit standard. The first is that value-chains based on this kind of standards are interoperable, still users can tailor the technologies to their needs, instead of being forced to implement a “one size fits all” type of standard. The second is reduced implementation cost because technologies can be obtained from multiple competing suppliers and re-used in multiple instances.

Toolkit standards have also disadvantages. The most relevant is that building value-chains may require an additional effort in designing the technical solution and selecting the tools. We will see later how DMP tries to mitigate this problem.

10     The DMP results so far

In its 18 months of activity DMP has proceeded along the well-established process of other standards organisations by

  1. Identifying the target of its standard (i.e. primitive functions)
  2. Developing the corresponding requirements
  3. Publishing Calls for Proposals
  4. Reviewing submissions received
  5. Selecting the most promising technologies
  6. Integrating technologies
  7. Publishing the standard.

DMP has identified the users of a generic media value-chain, made a thorough analysis of the functions performed between them and derived a comprehensive list of primitive functions [12]. The table below lists some of the most relevant primitive functions organised by categories.

Table 1 – DMP Primitive Functions

Category

Primitive Function

Access

Content, License, Tool

Assign

Description, Identifier

Authenticate

Device, Tool, User

Certify

Author, Device, User, Work

Identify

Content Format, Content, Device, Device Capability, Domain, License, Tool, Use Context, User

Manage

Device Capability, Domain, Tool, Use Data

Package

Content, Tool

Process

Backup, Binarise XML, Bind, Copy, Encrypt, Export, Import, Move, Play, Restore, Render, Store

Represent

Content, Key, Metadata, Rights Expression, Tool, Use Data

Revoke

Device, Domain

Verify

Data Integrity, Device Integrity

 

Requirements for primitive functions have been developed through a process that has seen the involvement of a range of industry representatives: collective management societies, consumer electronics manufacturers, end-users (particularly people with disabilities), public service broadcasters, sheet music publishers, telecommunication companies etc.

The table below is an example of how each primitive function is handled in [12], in the specific case of the “Assign Descriptors” primitive function. The 3rd column indicates the source of the requirement.

Table 2 – Requirements for Assign Descriptors

Definition

The function performed by an Authority to assign a Descriptor to a Work, a Resource or a piece of Content

AHG3

Objective

To facilitate search and find Works, Resources or pieces of  Content

AHG3

Requirements

·        To include the following mandatory fields

  • Author

  • Title

  • Genre of Authorship

  • Date of Creation of Work …

CMS

 

·        Assign Descriptors that facilitate cataloguing Content for B2B distribution

SMP

Benefits

Easy and accurate retrieval of Works, Resources or pieces of  Content

AHG3

     

Legend

AHG3

DMP Ad hoc group that provided the entry

 

CMS

Collective Management Societies

 

SMP

Sheet Music Publishers

So far DMP has produced two Calls for Proposals. The first Call has been for “Portable Audio and Video (PAV) Devices” [13]. The Call was the result of focussing the general DMP requirements to the specific case of devices capable of playing governed content but without the ability to connect directly to a network or broadcast channel [14]. Governed content can reach the device only through another device (e.g. a PC) that has a network connection.

11     The DMP specifications Phase I

In just six months DMP has been able to convert a large number of submissions covering all areas called for by the PAV Call into a set of six specifications (that DMP calls Approved Documents – AD). The table below gives the title of the six Approved Documents. In the table “Type” refers to the fact that a given AD must be implemented as specified in order to be able to interoperate with other implementations (normative) or used as a guide (informative).

Table 3 – DMP Approved Documents, Phase I

AD

Title

Type

Ref.

#1

Use Cases

Informative

[15]

#2

Architecture

Informative

[16]

#3

Interoperable DRM Platform

Normative

[17]

#4

Value Chains

Normative

[18]

#5

Registration Authorities

Normative

[19]

#6

Terminology

Informative

[20]

In the following a short overview of the content of the six ADs will be given.

11.1    Use Cases

This informative AD considers 7 Use Cases. These are examples of creation, distribution and use of governed content that can be implemented with the DMP ADs. Here is short description of some of the Use Cases.

Open Release

Imagine that I create what I consider an outstanding piece of work, but no one is willing to publish it. With DMP I can resort to Open Release, i.e. I create a package with my content, metadata about the content and usage rules, and then I post the package on my web site[1]. Note that, at this stage, there is no need to encrypt my content because, if my work becomes famous, I will be able to exploit it economically as a new release later on, possibly using more robust content control technologies.

Open Search

Imagine now that I run a commercial web site in a world where there is plenty of valuable Open Release content around. I start indexing the Open Release content items using their metadata and usage rules. Then I put in place a business whereby I give away basic authoring and browsing tools, I offer free hosting of Open Release content on my web site and then I engage in intermediation between sellers and buyers of Open Release content.

Internet distribution

Imagine a garage band that, after having achieved notoriety by Open Releasing their music, tries to actually sell it. Today the only way is by printing and distributing CDs, a hopeless task without a contract with label. With non-interoperable DRM selling music is an even more impossible task as it requires DRM-specific players that the garage band can obviously not afford. On the other hand an interoperable DRM standard would allow the garage band to purchase the necessary authoring and e-commerce tools from competing suppliers – therefore at a low price – and rely on a population of devices capable of playing governed content.

Personal photography

If today I take pictures and I pass them on to a friend or business partner I do not know how they will end up being used. But DMP tools allow me to encrypt my pictures, add usage rules to them and create a file that I can send to my friend or business partner. The advantage is that only my friend will be able to see the pictures because he is the only one who will get the decryption key from me or from a service provider.

This short sequence of examples proves once more the fact that DRM is a neutral word and should be kept such without loading it with emotional tones. Indeed the ability to release content by retaining “control” of it is not just a goal for incumbent media companies but often for small companies and individuals as well. “Control” technologies can be used to preserve the status quo because they enable the establishment of a high entry barrier to content distribution and seal the distribution against competition even better than it was ever possible before. On the other hand “control” technologies can also help break the status quo because they can lower the barrier to entry content distribution and create a competitive market, can support business models based on “long tails” [21], help the unknown artist to make himself known because of the quality of his work and not because some unknown entity controlling the distribution has decided to back him.

What discriminates the two versions of DRM is a simple word: “interoperability”. Therefore, before talking about DRM people should make sure whether the DRM they talk about is interoperable or not, because it can make a lot of difference.

11.2    Architecture

This informative AD provides a high-level description of the technologies required to support digital media value-chains. It shows how governed content is generated, packaged, passed on to other value-chain users and eventually consumed. The architecture presented is capable of supporting value-chains that are essentially digital extensions of today’s analogue value-chains, even though smart use of the standard technologies can lead to media value-chains that are vastly different from today’s.

A brief description of the architecture is given below. Note that terms beginning with a capital letter have the meaning of the DMP Terminology [20].

The process starts at the moment a Work is generated by a Creator in the form of a Manifestation that needs to be Instantiated before it can become an Instance carried in a Resource.

11.3    Interoperable DRM Platform

This normative AD is called Interoperable DRM Platform (IDP). It provides specifications of basic standard technologies (Tools) that are required to build Value-Chains. The word Platform has been selected to indicate that interoperable DRM applications sit on top of the collection of DRM Tools and use them as required. As this is the first version (Phase I), the Interoperable DRM Platform is currently called IDP-1.

In the table below (the same as Table 1) Tools for the primitive functions specified in IDP-1 are marked in bold.

Table 4 – IDP-1 primitive functions (in bold)

Category

Primitive Function

Access

Content, License, Tool

Assign

Description, Identifier

Authenticate

Device, Tool, User

Certify

Author, Device, User, Work

Identify

Content Format, Content, Device, Device Capability, Domain, License, Tool, Use Context, User

Manage

Device Capability, Domain, Tool, Use Data

Package

Content (file), Tool

Process

Backup, Binarise XML, Bind, Copy, Encrypt, Export, Import, Move, Play, Restore, Render, Store

Represent

Content, Key, Metadata, Rights Expression, Tool, Use Data

Revoke

Device, Domain

Verify

Data Integrity, Device Integrity

The following table provides a summary description of each IDP-1 Tool:

Table 5 – Description of DMP Phase I Tools

Access Content

Protocol that is employed when the Content Item to be Accessed has the License Bundled within it

Access License

Protocol that is employed in order to Access a License for a Content Item that does not have the License Bundled within it

Authenticate Device

Protocols to Authenticate three classes of Device

  1. Devices having unique certificates
  2. Devices that are uniquely identified by data
  3. Devices without a unique data with certificate proxy

Identify Content

Identification according to the Uniform Resource Names (URN) scheme

Identify Device

Two kinds of Device Identification (comprising the Device Identifier’s format, generation scheme, generation protocol and exchange protocol):

  1. ‘Device info-based identification’ generated by a ‘Device Identification Server’ based on Device information
  2. ‘Certificate-based identification’ based on a X.509 certificate

Identify Domain

(Done by a Domain Manager in Manage Domain)

Identify License

Same as Identify Content

Manage Domain

Comprises Protocols to:

  1. Set up a Device Domain Context
  2. Control the Use of Content within the Domain.
  3. Manage Device Domain membership (join/leave)

Package Content

File format containing the DCI with some or all of its ancillary Resources, potentially in a single package, It uses a DMP-defined subset of the MPEG-21 File Format

Process: Binarise XML

Tool to Represent a DCI in binary form. It is based on the MPEG BiM technology

Process: Encrypt

Tools to Encrypt/Decrypt (128-bit Key AES in CBC and ECB modes) and RSA with a variable Key length

Process: Play

(No specification required)

Process: Store

(No specification required)

Represent Content

Tool to:

  1. Convey Identifiers of Content and Resources/Metadata
  2. Associate DMP-specific information and Metadata with Content and Resources/Metadata
  3. Associate information with Governed Content.

It uses a DMP-defined combination of subsets of MPEG-21

  1. Digital Item Declaration (DID)
  2. Digital Item Identification (DID)

3.      IPMP Components

Represent Key

Tool to Represent Keys. It uses the W3C’s name spaces dsig and xenc.

Represent Rights Expression

Tool to Represent Rights and Conditions associated with the Use of Content. It uses a subset of MPEG-21 Rights Expression Language

Represent Use Data

Tool to represent the data generated by Users Using Governed Content

11.4    Value-Chains

This normative AD specifies how the informative Use Cases described in AD No. 1 can normatively be implemented using the Tools specified in AD No. 3.

By giving a normative value to this AD DMP does not imply that Use Cases of AD No. 1 can only be implemented as specified in the Value-Chains AD, but only that the AD provides example normative implementations so that Users assembling Tools as specified in the Value-Chain AD will be able to interoperate with other Users who will assemble the Tools in a similar way.

The steps in Value-Chain No. 1 Open Release can be implemented as follows

Creation:

Create

Resources

Obtain

Resource Identifiers from Resource Registration Agency (R-Agency)

Create

Metadata for all Resources

Obtain

Metadata Identifiers from Metadata R-Agency

Create

Human-readable license (H-license)

Create

Machine-readable License (M-License) corresponding to H-license

Obtain

License Identifiers from H-license and M-License R-Agency

Create

DCI with Resources, Metadata, H-license and M-License

Obtain

Content Identifier from Content R-Agency

Create

DCF from DCI

Release

DCF

Access and Use

Access

DCF

Play

DCF

 

The steps in Value-Chain No. 5 Internet Distribution can be implemented as follows:

Creation:

Create

MP3 and other Resources

Create

Metadata for all Resources in each recording

Obtain

Resource and Metadata Identifiers from R-Agency

Create

H-license

Create

Corresponding M-License

Store

M-License in a License server

Obtain

H-license and M-License Identifiers from R-Agency

Encrypt

Resources

Create

DCI with Encrypted Resources, Metadata, H-license and M-License

Create

DCF

Post

DCF on web site

Access and Use

Access

DCF

Play

DCF

 

The steps in Value Chain No. 7 Personal Photography can be implemented as follows

Creation:

Create

Photos and Metadata

Obtain

Resource and Metadata Identifiers from R-Agency

Create

H-license

Create

Corresponding M-License

Store

M-License in the License server

Obtain

H-license and M-License Identifiers from R-Agency

Encrypt

Resources

Create

DCI with Encrypted Resources, Metadata and H-license

Create

DCF

Send

DCF to friend

 

Access and Use

Access

License

Create

New DCF with License Bundled within it

Transfer

DCF to PAV Device

Play

DCF

11.5    Registration Authorities

This normative AD collects roles, qualification requirements, appointment procedures and operation rules of DMP-appointed Registration Authorities.

For any type of Entity for which identification is required, DMP appoints the “root” element (Registration Authority). Registration Authorities allocate namespaces and appoint Registration Agencies. Registration Agencies Assign Identifiers.

11.6    Terminology

This informative AD collects definitions of terms used in DMP ADs.

12     Beyond technical specifications

12.1    Reference software

DMP is developing reference software for its Phase I specification. There are three main reasons for this development:

  1. Check the text of the specification by verifying that the software implementation performs the functions that are expected
  2. Promote the use of DMP specifications because users find it easier to adopt the technology
  3. Provide tools for conformance testing (see below)

To the extent possible the modules of the DMP reference software will have an open source license (OSS).

When this is not possible because there is Intellectual Property (IP) that is required for the implementation, the module will have a license that allows modifications of the software and use in products with the condition that the software or its derivative conforms to DMP specification (MUSS).

The following table describes the type of license adopted for DMP reference software.

Table 6 – Licenses for DMP reference software

Source technology or specification

License of reference software

Technology is unencumbered by IP

OSS

Technology is encumbered by IP with a RAND license

MUSS

Specification from another body provided with reference software

Existing license, if OSS- or MUSS-compatible

Specification from another body provided without reference software

OSS, if possible, otherwise MUSS

     

Legend

OSS

Open Source Software license

 

MUSS

License to modify and use the source if implementation conforms to DMP specification. No license for patent is granted

The DMP reference software for Phase I will be published in October 2005 as a normative Approved Document No. 7. It will contain reference software for all Tools and for some selected Value-Chains as well.

12.2    Conformance

Value-chains are the result of business agreements and are supported by a set of technologies that Value-Chain Users decide to adopt. As DMP specifications can be implemented by anybody Value-Chain Users can get solutions from multiple sources. However, not every implementer can be equally trusted; therefore Users must have the means to make sure that the other Users “play by the rules” i.e. employ conforming products.

DMP will develop Recommended Practices for End-to-End Conformance to be published in January 2006 as Approved Document No. 8. When this will be achieved, Value-Chain Users will be able to reference the document in their business agreements.

13     Adapting to the new environment

For DMP DRM interoperability is a necessary condition for its acceptance but any DRM solution, even an interoperable one, has the potential to substantially alter the balance between users that existed in the analogue-world, in particular when the user is the end-user. If the imbalance is not remedied the scope of Traditional Rights and Usages (TRU) of media users, be they rights and exceptions sanctioned by law or “liberties” that users have taken with media, will be reduced. This reduction may lead to the outright rejection of DRM by some users, in particular end-users.

DMP does not claim that an established TRU necessarily implies a right of a user to a particular use of digital media or an exception or something else. DMP only observes that, if users have found a particular TRU advantageous in the analogue domain they are likely interested to continue exercising that use in the digital domain as well. This does not necessarily mean that such usage will be for “free” as the exercise of a TRU in the digital space may have substantially different implications. Leveraging on this interest may offer opportunities for new “Digital Media Business Models” that are attractive to users but also respectful of rights holders rights. Such Digital Media Business Models can then be extended to cover usages that were not possible or not even considered in the analogue age.

During 2004 a large number of TRUs have been collected and analysed [21] and an exercise has been carried out to study the effect of a range of scenarios of TRU support [22]. Finally in April 2005 a Call for Contributions on “Mapping of TRUs to the digital space [23] with an attached document “Issues in mapping TRUs to the digital space” [24] have been published.

The latter document contains an analysis of 14 (out of 88) TRUs: 

Table 7 – TRUs analysed in the TRU Call for Proposals

#

TRU name

1

Quote

2

Make personal copy

3

Space shift content

4

Time shift content

5

Make playback device  

6

Choose playback device

7

Use content whose copyright has expired

14

Edit for personal use    

18

Apply a rating to a piece of content

19

Continued access

30

Freedom from monitoring

55

Access content in libraries

67

Make content creation device

69

Access content of one's choice

For each analysed TRU the following is provided

  1. Nature of TRU
  2. Digital support
  3. User roles and benefits

Let’s see the case of the following table:

Table 8 – Analysis of TRU #55 “Access content in libraries”

Nature

A user can draw a book from a public library at nominal or no cost and read it at will for a short, possibly renewable, period of time

Digital support

Scenario 1: Digital support of copyright exceptions for libraries

Repositories can “lend” Governed Content by using copy-controlled check-in/check-out mechanisms for the borrowed Contents Items

 

Scenario 2: Focussed on-line access to Content

A Repository offers the following services

  1. Content search services not driven by commercial criteria
  2. Pointers to content offered by other (possibly commercial) sources
  3. Governed Content when no other sources are available
  4. Content when copyright has expired (see TRU #7)
  5. Content when the repositories hold the copyright

User roles & benefits

Public Authorities:

Enact appropriate legislation to enhance the social role of public libraries in the digital media era

The TRU Call requests contributions addressing the following topics:

  1. Comments on the methodology utilised for studying TRU support in the digital space
  2. Identification and description of new TRUs
  3. Comments on the choice of scenarios, the type of measures to support the scenarios and the effects on the main Value-Chain Users
  4. Identification and analysis of new scenarios of TRU support

DMP expects to receive contributions by 15 July 2005 so as to start the development of a “Recommended Action on Mapping of Traditional Rights and Usages to the Digital Space” to be published in January 2006 as Approved Document No. 9.

The process to technically support TRUs within the IDP is already being discussed, and the following process is being considered:

  1. Selected a set of TRUs/scenarios
  2. Document each TRUs/scenarios as a Use Case in AD #1
  3. Develop add to AD #3 (IDP) the Tools needed to support the selected set of TRUs/scenarios
  4. Document in AD #4 (Value-Chains) how TRUs/scenarios are actually supported
  5. Add to AD #7 new software modules corresponding to new Tools or Value-Chains
  6. Develop AD #9 (Mapping of TRUs) by:
    1. Providing references to other ADs, viz: Use Cases, Tools, Reference Software etc.
    2. Identifying actions to recommend for legislative support
    3. Documenting results of legislative actions supporting specific TRUs/scenarios where known

The Recommended Action is expected to be of use to national legislative or regulatory bodies when defining the scope and support of TRUs in their jurisdictions.

14     Interoperable DRM Platform, Phase II

In January 2005 DMP issued a new Call for Proposals targeted to “Stationary Audio and Video (SAV) Devices”, i.e. Devices capable of Playing Governed Content obtained from a network or a broadcast channel.

In April 2005 submissions were received and working drafts of Approved Documents No. 1, 2 and 3 were produced. These drafts add new Tools to existing specifications according to the following philosophy:

  1. Re-use Phase I Tools, when these continue to be useful
  2. Extend Phase I Tools, when this is needed and can actually be done
  3. Add new Tools when 1. and 2. are not possible.

This is the list of Tools that are being considered for IDP-2 (Tools underlined are those extended from IDP-1, Tools in bold the new ones):

Table 9 – IDP-2 primitive functions

Category

Primitive Function

Access

Content, License, Tool

Assign

Description, Identifier

Authenticate

Device, Tool, User

Certify

Author, Device, User, Work

Identify

Content Format, Content, Device, Device Capability, Domain, License, Tool, Use Context, User

Manage

Device Capability, Domain, Tool, Use Data

Package

Content (file), Content (stream), Tool

Process

Backup, Binarise XML, Bind, Copy, Encrypt, Export, Import, Move, Play, Restore, Render, Store

Represent

Content, Key, Metadata, Rights Expression, Tool, Use Data

Revoke

Device, Domain

Verify

Data Integrity, Device Integrity

Publication of Phase II specifications is planned for October 2005.

15     How can DMP help television?

The DMP specifications contain many elements that can help television to complete a successful transition from the analogue to the digital world. Some examples

  1. Rights management. Content comes into production from different sources and different licenses. The Rights Expression Language allows the management of complex rights configurations, including the release of different versions of the content with different licenses.
  2. Multi-channel distribution. By defining the bundle of content, metadata and licenses DMP governed content is ready for distribution on multiple delivery systems.
  3. Open set top box market. By defining a content protection infrastructure the DMP specifications enable the birth of an open market of set top boxes for broadcast governed content that are service-provider and possibly delivery-system agnostic and are open to innovation.
  4. Content re-use in the home. By using a standard rights expression language DMP allows broadcasters to set permissions for content re-use in the home and beyond and deploy multiple distribution strategies.

16     Conclusions

Television is still trying to adapt to the new context created by digital technologies. Digital Rights Management is a technology that is often touted as the magic wand that will save television, but the acceptance of a possibly very invasive technology in a media distribution system for the masses is far from clear.

The Digital Media Project, a not-for-profit organisation established in December 2003, is developing specifications with a set of desirable features:

  1. Possibility to implement different grades of DRM, from the very light to the very tight, depending on the type of markets, content, delivery system or viewers
  2.   Applicability to different delivery systems
  3. Flexibility to deploy different content governance strategies
  4. Balance between the requirements of content governance and the traditional end-users’ liberties

17     References

  1. WIPO Copyright Treaty, 20 December 1996, http://www.wipo.int/treaties/en/ip/wct/trtdocs_wo033.html
  2. The Digital Media Manifesto, http://manifesto.chiariglione.org/
  3. The Digital Media Project, http://www.digital-media-project.org/
  4. Berne Convention, http://www.wipo.int/treaties/en/ip/berne/trtdocs_wo001.html
  5. Universal Copyright Convention, http://www.unesco.org/culture/laws/copyright/html_eng/page1.shtml
  6. Open Source Initiative, http://www.opensource.org/
  7. Creative Commons, http://creativecommons.org
  8. William Fisher, An Alternative Compensation System, Chapter 6 of Promises to Keep: Technology, Law, and the Future of Entertainment. http://cyber.law.harvard.edu/people/tfisher/PTKChapter6.pdf
  9. The Moving Picture Experts Group, http://mpeg.chiariglione.org/
  10. CEN/ISSS, Digital Rights Management Final Report http://europa.eu.int/comm/enterprise/ict/policy/doc/drm.pdf
  11. The Statutes of the Digital Media Project, http://www.dmpf.org/project/statutes.htm
  12. Primitive functions and requirements, http://www.dmpf.org/open/dmp0277.zip
  13. Call for Proposals on Portable Audio and Video Devices, http://www.dmpf.org/open/dmp0145.doc
  14. Requirements for Portable Audio and Video Devices http://www.dmpf.org/open/dmp0146.doc
  15. Digital Media Project; Approved Document No. 1 – Technical Reference: Use Cases; 2005/04, http://www.dmpf.org/open/dmp0401.doc
  16. Digital Media Project; Approved Document No. 2 – Technical Reference: Architecture; 2005/04, http://www.dmpf.org/open/dmp0402.doc
  17. Digital Media Project; Approved Document No. 3 – Technical Specification: Interoperable DRM Platform; 2005/04, http://www.dmpf.org/open/dmp0403.zip
  18. Digital Media Project; Approved Document No. 4 – Technical Specification: Value-Chains; 2005/04, http://www.dmpf.org/open/dmp0404.doc
  19. Digital Media Project; Approved Document No. 5 – Technical Reference: Registration Authorities; 2005/04, http://www.dmpf.org/open/dmp0405.doc
  20. Digital Media Project; Approved Document No. 6 – Technical Reference: Terminology; 2005/04, http://www.dmpf.org/open/dmp0406.doc
  21. Chris Anderson, The Long Tail, Wired, October 2004 http://www.wired.com/wired/archive/12.10/tail.html
  22. Collection of TRU templates, http://www.dmpf.org/open/dmp0270.zip
  23. TRU scenario analysis, http://www.dmpf.org/open/dmp0280.zip
  24. Call for Contributions on “Mapping of Traditional Rights and Usages to the digital space”, http://www.dmpf.org/open/dmp0409.doc
  25.   Issues in Mapping of Traditional Rights and Usages to the digital space, http://www.dmpf.org/open/dmp0410.doc

18     Acronyms

This is a collection of the most important acronyms used throughout this paper

Acronym

Meaning

ACS

Alternative Compensation Schemes

AD

Approved Document

CA

Conditional Access

DCF

DMP Content Format

DCI

DMP Content Information

DMM

Digital Media Manifesto

DMP

Digital Media Project

DRM

Digital Rights Management

IDP

Interoperable DRM Platform

IP

Intellectual Property

IP

Internet Protocol

IPMP

Intellectual Property Management and Protection

MPEG

Moving Picture Experts Group

MUSS

Modify and Use Source Software

OSS

Open Source Software

REL

Rights Expression Language

TRU

Traditional Rights and Usages

URN

Uniform Resource Names

VCR

Video Cassette Recorder

WIPO

World Intellectual Property Organization

XML

eXtensible Markup Language


 

[1] Note the similarity of this case with Creative Commons. The difference is that the licenses can be one of those of CC, but they can be other licenses as well.

[2] A User can deliver Content to another User for free Use, such as when the Content has been put in the public domain, but DMP specification applies only to the case in which a User delivers Content to another User in the form of Governed Content.

[3] Note that the License of the Governed Content Item can be Bundled within the Governed Content or not Bundled within the Governed Content. In the latter case it must be obtained separately from the Content.