Distributing and Storing Media

Universal Preservation Format: User Survey

2. How should material that describes an archives collection be distributed and/or stored?

a. Continuously with content : for example, at any given point, one might freeze a video frame and access information about:
+ the video as a complete entity (example: copyright material, original format and running time...)

+ the specific video clip or sequence that includes this frame (example: number of frames in this version versus number of frames in original, which numbered frame within the sequence)

+ the individual frame's descriptive information (what objects are contained in the image) 19

b. As header information only: information describing "essence" stored as a Table of Contents at the beginning of the video. 13

c. As a separate document: information about the "essence" would load and be stored as a document, accessible via retrieval or "parsing" software. 18

d. Do not know. 13

I am concerned mostly with audio and so I have no opinion on some of these moving image issues. I believe most metadata about audio can be captured in descriptive data entry fields such as: key words, participants... - Baker

[Describing continuously with content] ... might be a realistic undertaking within the budget of a major motion picture or for a uber-publisher, but I doubt individual artists (who can publish their own work on the net), small businesses, and other under-funded or under-staffed groups would be able to afford the time or money to catalog their work with this level of granularity. - Dreilinger

Frame-by-frame description seems extreme for most institutions. But, like all archival description, it should be possible to describe videos similarly at all levels--from a collection of 500 tapes down to a single frame.It's nice to have description attached to the material, but it should also be able to stand alone. Are these mutually exclusive options? -Hadley

Each of these are viable alternatives. Implementation would be dependent on institutions technical resources and abilitiies. - Hedges

"a" through "c" seem nearly impossible for most materials since someone has to oversee the input and value of the massive amount of descriptive information these methods would call for. - Messier

I really like [Continuously with content].
This would also be great for still images.
FlashPix suggests this but we are not there yet. - Ogle

I would have video and descriptive information in the same file, as well as stills. - Schwan

We are exploring the use of both header and separated metadata, and think we want both, with a preference for linked separate metadata files. Again, we are very involved in investigating this...
[Continuously with content] looks good, and we are after a similar level of retrievability of content and metadata for our sound collection, but it looks like a luxury at this stage. - Webb

For Euromedia, which also uses browser technology and is fully object-oriented, we have a hierarchical storyboard as a metadata descriptor, with keyframes, and with text associated at four levels: frame, shot, sequence, programme.
For security, this information could be stored with the digitised data as well as separately. At the object level this distinction is invisible. - Wright