Recently I had the pleasure of talking with the Director of the Library and Archives at Smithsonian Channel, Karma Foley. This summer Smithsonian Channel is migrating to a more robust database, and Karma shared how PBCore is informing this expansion and helping the channel preserve their private collection in a file-based, CollectiveAccess environment.
Smithsonian Channel formed in 2005 as a joint venture between Showtime Networks and the Smithsonian Institution. In 2007 they began broadcasting of their original content inspired by the museums and research facilities. The Library and Archives includes a mix of tape-based and file-based HD formats of raw footage for original programs as well as production elements and the finished products themselves. Karma joined in 2011 with the task of figuring out what the channel had and how to preserve it and make it accessible for re-use.
Karma first focused on establishing a basic database that would allow production staff to locate re-usable footage. Their in-house, unique database, SCOUT, is built on CollectiveAccess and is fairly simple. Currently, works and instantiations exist in one record and technical and descriptive metadata are tied together with PBCore acting as the consultant for controlled vocabulary and many elements. For their new model, PBCore 2.0 is being enlisted as a way of describing separate but related works and instantiations, which will be especially helpful for describing their older tapes that have been digitized.
The team picks and chooses specific PBCore elements and vocabulary and modifies others to suit their collection. Elements that help to describe pieces of the broadcast production process are especially helpful. Karma noted that, “One of the best things about PBCore is that it was created by people who know about broadcast and production, which our new data model will focus on, whereas our current model just focuses on individual video assets.” Film and analog video terms don’t apply to their collection, so those are stripped and replaced by additional terms for born digital and file-based formats.
Karma has also brought archival and PBCore thinking to the production end. Strict file naming starts before or right after material is shot so that a unique ID is associated with it from the beginning. Descriptive and technical metadata are gathered early on as well. This is done through a spreadsheet that all original productions must deliver when the project wraps. Specific metadata for each media item (such as a tape, disc, or card/folder of footage) must be supplied following a template that includes columns for date, location, description, media type, recorder brand/model, file format/codec, frame rate, and resolution. Having the production team enter this data ensures a greater level of accuracy and helps to keep track of media during production and edit. When the project wraps, the data is double-checked by Archives staff and then mapped into the database.
Smithsonian Channel’s collection is private and their database is for internal use only, so the data sharing capabilities of PBCore XML and the implementation of attributes are not primary benefits of the schema for their collection. However, Karma believes that incorporating standards is a good idea while working to develop a data model that best suits their unique collection.
What is most inspiring is how PBCore is helping to inform Smithsonian Channel’s production workflow that is both preservation minded and forward thinking. Karma noted that PBCore allows for more detailed technical metadata than they are currently describing, and she is especially anxious to explore automated ingest of metadata from media files that will improve the level of detail in technical metadata collected.
Best of luck during this exciting endeavor, Karma, and thank you for sharing your PBCore experiences with us!
Written by Bryce Roe, intern for the American Archive of Public Broadcasting. Interview with Karma Foley, Director of Library and Archives at Smithsonian Channel.