The agenda was approved as written. The report of the previous meeting was amended with minor editorial corrections.
This meeting reviewed the work of task group SC-02-02-G on project AES-X111 (UMID insertion into the AES3 stream). This task group has completed its work and its output document is close to completion. From a metadata point of view, this work will produce a key link between live audio streams and the metadata associated with these streams until such metadata is "wrapped" with the audio in file-based structures like Broadcast Wave, AAF or MXF. A discussion occurred on how the Unique Material Identifier (UMID) would be inserted and how its accuracy would be affected by changes in the audio stream. It was also explained that the method identified for the insertion of a SMPTE UMID could also be used to insert other types of identifier that may be required. This would require a separate standardising document to be produced for each of these types so the mapping between these different identifiers could also be standardised.
D. Ackerman, chair of the SC-03-06 working group on Working Group on Digital Library and Archive Systems, reported on progress of the administrator and audio metadata schemas for libraries and archives. T. Sheldon discussed the progress of the document describing administration metadata which is in its final stages.
There were two other documents forming the SC-03-06 audio metadata set: one dealing with core audio metadata, the other with process history metadata. The core audio metadata contains the technical data describing the audio itself while the process history metadata describes the "who, what, where and why" of the processes that have been applied to the audio during its evolution and which contains both event and timeline based data.
This key work effectively producing an audio metadata set to which all other audio metadata requirements could be mapped. It would document the processes used in planning, recording and producing professional audio and would clearly lead to the desirable goals of publishing, transmitting and archiving of the audio recording with a consistent set of metadata. As the archiving process is most likely to contain the metadata required by the other two, this area could provide the best point to describe the audio metadata "super-set" covering the requirements of the other audio delivery mechanisms. Metadata could be generated by processes further down the process chain that would not be required by any final audio delivery point, but this area should be covered elsewhere.
All the work being carried out in SC-03-06 is using XML to describe and carry the metadata.
There was a short discussion on the difference between metadata wrapped with the audio and that which may be held in databases within an origination. It is more likely that database structures may be used during the production process within organisations. The nearer the audio "delivery point" the process gets, the more metadata may be also wrapped with the audio in file structures like Broadcast Wave, AAF and MXF. Clearly, database structures of some kind will still be employed, even when all the required metadata is delivered wrapped in a file, to enable rapid searches.
The documents that had been distributed to the group were discussed. The key document following our last meeting has been provided by S. Aoki and is titled "The metadata for digital broadcasting in Tokyo FM". Aoki explained the reasoning behind the Tokyo FM document and how they have been discussing this with the TV Anytime group. He encouraged the AES to work with W3C and TV Anytime. A question was asked about the degree of detailed metadata that would be broadcast. Concerns were expressed that, if too much detail were distributed along with the programme material, it would become easy for automated receiving equipment to recover elements of material which some rights owners have deprecated. It was felt this was not a matter for standards but more a matter of policy for individual broadcasters. However, it was agreed to note this issue.
AES-X134: Core audio metadata XML definitions.
The goal of this task group is to produce a number of metadata sub-sets for specific tasks, and to map these sets to the administrator and audio schemas being produced in SC-03-06 that could be considered as defining the metadata "super-set".
The meeting discussed possible sub-sets. Some of the possibilities were:
The meeting also discussed the standards to be used when exchanging metadata. The Key-Length-Value (KLV) convention is used by the SMPTE; however, Extensible Markup Language (XML) seems to be growing worldwide as a preferred method of expressing and exchanging metadata. The meeting agreed that AES-X134 should concentrate on the use of XML documents and schemas to develop and define professional audio metadata sub-sets and map these to the work coming from SC-03-06.
It was hoped that members could be encouraged to join this work and put their ideas and proposals on the SC-06-06 reflector.
Information on AES liaisons with EBU P/AGA, SMPTE and TV Anytime was requested. Chambers said he was a member of the EBU P/AGA group and he would also seek a contact with TV Anytime. An action point was picked up by R. Caine to find suitable contacts with W25 in the SMPTE.
The next meeting is scheduled to be held in conjunction with the AES 116th Convention in Berlin, Germany, 2004-05.