Home

Project

Partners

Related

Archives

Search

Glossary

Metadata Watch Report #1

[ contents | section 1 | section 2 ]

Section 3 - Domain reports

3.1 Industrial sector
3.2 Publishing sector
3.3 Audio-visual sector
3.4 Educational sector
3.5 Academic sector
3.6 Research sector
3.7 Geographical information sector

3.3 Audio-visual sector

A description of the domain could be: "Production, distribution and archiving of digital audiovisual material". The application fields are: radio and television broadcasting; audio and video (post)production; audiovisual archives; multimedia library systems; image banks; news agencies; WEB TV. Due to the convergence of television, computer and communication technologies, the separate area's within the digital media domain are approaching one another in many respects. They share a number of common focus points and interests :

• The increase of the number of digital documents in the organization.
• The need for standards for storage, exchange, cataloguing and indexing of digitized material.

• The necessity to fill multiple new channels and services.
• The need for migration scenarios from analogue to digital collections, procedures and systems.
• The need to integrate media-technology and business systems for efficiency savings, cost control and support for commercial transactions.
• The need of fast and solid delivery to increasing groups of (new) customers within and outside of the organization.
• The pressure on the preservation of cultural heritage via digitization.
• The need to guard, protect and track copyrights.

• The need to improve research possibilitities for digital audiovsiual material.

At the moment numerous differently scaled and shaped activities can be observed to go on in this area. The work is done on the one hand internationally, but on the other hand also locally and the developments are rather different in scope and effect. To name a few : the design of national infastructures for digital broadcasting; the distribution of radio and television programmas via the Internet; the work of the international standardization committees; the setting up of various kinds of online multimedia-catalogues; the development of educational and commercial Video on Demand Services. These activities are being executed in several project formats, as part of regular business developments, as government programmes, as European funded initiatives and so on. The variety in players is great. Involved are national and commercial broadcasters, publishers, producers, universities, national and private audiovisual archives, industrial partners, IT , broadcast engineering, governmental departments, national and international bodies and professional federations.

Important part of these activities and projects is the development of metadata standards. Metadata development might be the core objective of an activity (e.g. the standardization committees), in other cases it is (inevitable) part of a smaller or larger digital (private or public) media project. The dynamics and objectives of the developments show numerous differences and the level of synchronity is low. Some activities concern all forms of access to and (re)use of all digital multimedia on the web as well as for professional use, others again are dedicated to one single area. At the moment, in many cases, the lack of common standards forces the industry, the broadcast companies and the distributors of WEB TV alike, to develop proprietary methods and pragmatic solutions to be able to actually run their implementations and systems.

It can be observed that 'audiovisual' metadata still are subject to different interpretations. One the one hand IT would consider the concept as: data relevant to an information system including data models for technical appliances. In this sense metadata are really a type of information depending on form, characteristics, classification, storage and structuring; version management, integrity and performance are also parts of this comcepts. On the other hand 'broadcast IT' interpret metadata primarily as media systems, i.e. information that focuses on media content identification and analysis. Here, audiovisual production metadata are logical data and the manner of their classification, purely related to programmes on radio and television.

Broadcasters have therefore cast this appropriate definition : content = essence + metadata (essence, i.e. the audiovisual content itself). Information analystst in this area will often even think of an additional layer, metadata for them refers to information regarding the structure of metadata proper, that again refer to the document content.

Metadata in the digital media domain are rather more complex than metadata related to digitised textual information. Though the principles are identical, image and sound productions, more than text, contain much implied information that again constitutes separate groups of metadata. Finally, the digitised production process generates a quantity of technical metadata used to facilitate transfer, internal and external distribution or storage. It is clear that the increase of digital documents in this areas causes many problems, but it could also offer a possible solution: some of the identifying information can be embedded in the video or audio stream itself (the electronic metadata). To be able to maximize the benefits from this, the need for standardization is high.

The various types of 'audiovisual' metadata can be categorised as: signal related metadata, programme documentated related metadata and pointers. The standardization of the various groups and classes of metadata is linked to the global user requirements for digital media systems.

The main issues are:

• Interoperability between mediaformats and systems with automatic exchange of metadata.

• Standardization of data structures and data models.
• The development of unique identifiers to link the stored audiovisual material with the related documentation through the process of creation, delivery, use and re-use.
• The semi automatic and full automatic creation of embedded (wrapped) metadata within the content streams and files.
• Automated (semantic) indexing of important sequences or parts in video, sound recording or stills.
• Standards for automated indexing to be able to use image analysis, teletext information, speech recognition and sound analysis.
• The migration of the embedded metadata into unwrapped metadata, stored in databases, that can be managed and controlled over networks.
• Versionmanagement: the incorporation of different versions of the same document in different stages of the production process, as well as descriptions that cover different copies of the same document
• The development of methods of extraction , searching, evaluation an validation to assist content queries.
• Multilingual searching.

The work of the standardization committees and working groups may be summarized as follows:

• Regarding the definition of a central registration structure that implements the mapping of different metadata schemes.
• Regarding the development of tools with generic functionalities, to develop and use metadata repositories.
• Regarding metadata taxonomies, so as to structure the metadata discourse.
• Regarding ontologies related to the characteristics of metadata and to the definitions of data elements such as designation of fields, types, classification and semantics.

Amongst users and committees alike there is a growing assumption that metadata and metadata standardization will eventually solve all problems connected to the issues of interoperability. The questions surrounding the protection and guarding of the integrity of the metadata itself, seem somewhat neglected here.

One of the problems standardizing committees face, is that current technical equipment has become so inexpensive that the program makers and other users can generate near-broadcast quality themselves and thus not wait for "the next standard". Some committees operate closely with manufacturers. Therefore, industrial politics is involved, and this might be an obstacle. Other groups are very large and their list of required data is often so extensive that difficulties are foreseen. The work of a fair number of groups and committees is concentrating on practically every subject attached to the audiovisual domain. In these cases each workpackage could be a project of its own.

At the moment the broadcast world (production, broadcasting and archiving) is specifically looking at the work of SMPTE in this area (Metadictionary and Engineering Guidelines) and SMPTE-UMID. The UMID (Unique Material Identifier) has only very recently become a standard (SMPTE 330). The SMPTE Metadata Dictionary is out for publication in two or three weeks from now. It is expected that the Dictionary will be a standard in a few months. The broadcast world is also highly interested in the BBC Media Asset Management Data Model SMEF (Standard Media Exchange Framework) which only this week has been made widely available via free licence. The European Broadcasting Union Projectgroup P/Meta is aiming at developing common media exchange formats for broadcasters, publishers and archives and will incorporate the work of SMPTE, MPEG-7 and SMEF.

As for multimedia content in the professional environment, the activities of the MPEG-7 group are being monitored. The development of this standard is still at a very early stage. The MPEG-7 Draft International Standard is not to be published until medio 2001. Before that time the many different sections within MPEG-7 will have to be integrated and connected. Within the EBU community a initiative has recently started to establish a cooperation between the MPEG-7 Description Schema Work and the SMPTE Metadata Dictionary Taskforce. This initiative is aimed at providing interoperability between the two standards and to link the broadcast environment to the objectives of MPEG-7.

In general there is still a proliferation of activities that does unfortunately carry the danger of producing incompatible standards, even if those had not been -as they are, in different stages of development. At the moment there are many different contributions to standards developments for various areas and at different stages of growth, that stem from the various professional qualifications and commercal interest of the participants. For as long as this is found wanting, it has to make with ad-hoc solutions that cannot fail to contribute to the general confusion. As a consequence it is observed that while some standards have already a framework that only need to be completed, other- even though of comparable nature - are quite a distance from any completion. It might be concluded that-within the area of 'audiovisual' metadata - too many organizations are currently developing far too many 'standards', methods and procedures. This situation inhabits the danger of both overlaps and gaps.

MPEG-7 is aiming at solutions for the entire audiovisual content. The SMPTE work is dedicated to the production and distribution of digital television and radio material. Audiovisual material however, contains textual information including catalogues entries, production information, scripts etc. Offering access to this textual information is not within the objectives of MPEG-7 or SMPTE. Therefore they will have to tackle this issue by incorporating descriptive metadata as a seperate node within the Metadata Dictionary. This particular node still needs to be developed, i.a. with contributions froms other initiatives that specifically focus on descriptive metadata.

The membership of most commissions and workinggroups in the digital media domain is highly technically. Despite the fact that the convergence of documentation fields and technology is extremely manifest here, the technical expertise generally exceeds the archival and 'documentational' know-how. On the other hand it can be observed that the level of understanding among members of some committees, of the issues and techniques associated with media technology, is not very even. On top of this, the available resources are usually highly constrained. In addition, development of thinking in the field is proceeding fast while some clear priorities are emerging.

Audiovisual archives is not evenly represented in most of the workpackages even though subjects as migration scenario's, user requirements and datamodelling are higly documentation and archive fields. Matters documentalists and archivistst should more address are: storing metadata (embedded or separate); unique identifiers: how will these work on a day-to-day basis, significance and practical implications etc. Additionally, these professional groups could offer input as to matters covering quality control and information. In the end most committees will have to come up with both an archive format and a common data format. It is expected of broadcast archives to play a big role in developing technical formats, even though these archives have up till now not really done any serious work within this area.

For all committees and workingsgroups it will be important to stay aware of the work of on another in order to garantee compatibility, mappability and interoperability. The general question remains whether all these local, national and international efforts will in the end provide the audiovisual world with a solid, workable standard. A better option perhaps, would be for all the involved organisations and commissions to specifically focus on the metadata part they are experts in. An overall metadata infrastructure could then be designed, to facilitate for all media organizations to interoperate. For the broadcast environment the Advanced Authoring Format(AAF)-initiative moves in this direction.

>>Section 3.4 Educational sector

[ contents | section 1 | section 2 ]

Maintained by: UK Office for Library and Information Networking (UKOLN)
Last updated: 07 August 2001