PBCore 2.0 can be used to express “abstract,” or metadata-only assets.
These assets simply do not have any instantiations, and might serve to represent a high level idea such as a television for which there is no single physical or digital instantiation.
The descriptive metadata within such an assets may be referenced from within other assets. For example, programs within a series can reference the series record and vice versa. This can save the effort of creating redundant metadata.
When PBCore is used to list related items, it can also express the nature of their interrelationships.
For example, when expressing an asset within pbcoreDescriptionDocument, use the sub-element structure of pbcorePart to define segments, stories or episodes.
Then utilize the element structure within ‘pbcoreRelation’ to articulate any associations between assets.
<?xml version="1.0" encoding="UTF-8"?>
<pbcoreDescriptionDocument xmlns="http://www.pbcore.org/PBCore/PBCoreNamespace.html"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.pbcore.org/PBCore/PBCoreNamespace.html>
<pbcoreIdentifier source="MARS" annotation="this value is a unique identifier to which episode records will point">XYZ123</pbcoreIdentifier>
<pbcoreTitle>Homeowners: The Series</pbcoreTitle>
<pbcoreDescription>This series is dedicated to inspire homeowners to invest wisely in their homes.</pbcoreDescription>
<!-- section removed for clarity -->
<pbcorePart>
<pbcoreIdentifier source="MARS">XYZ123-A</pbcoreIdentifier>
<pbcoreTitle>XYZ123: Program One Title</pbcoreTitle>
<pbcoreDescription>This is the pilot episode of the series.</pbcoreDescription>
<pbcoreRelation>
<pbcoreRelationType>Is Part Of</instantiationRelationType>
<pbcoreRelationIdentifier source="MARS">XYZ123</instantiationRelationIdentifier>
</pbcoreRelation>
</pbcorePart>
<pbcorePart>
<pbcoreIdentifier source="MARS">XYZ123-B</pbcoreIdentifier>
<pbcoreTitle>XYZ123: Program Two Title</pbcoreTitle>
<pbcoreDescription>This episode shows how to safely shovel snow.</pbcoreDescription>
<pbcoreRelation>
<pbcoreRelationType>Is Part Of</instantiationRelationType>
<pbcoreRelationIdentifier source="MARS">XYZ123</instantiationRelationIdentifier>
</pbcoreRelation>
</pbcorePart>
<!-- section removed for clarity -->
</pbcoreDescriptionDocument>
PBCore does not limit the amount of data contained in its description-oriented fields, so it’s possible to includet whole transcripts. Transcripts and other lengthy descriptions are likely to contain characters that are problematic to include in XML, however.
There are two solutions:
1) wrap or surround it within a CDATA structure:
<pbcoreDescription> <![CDATA[This is how to surround data that would otherwise cause problems to an XML parser; such data types can range from characters otherwise used in XML, to binary data types]]> </pbcoreDescription>
2) use html escape characters within the text. For example: for & (ampersand), substitute
&
for ” (quotation mark) substitute
"
WNET.org is New York City’s premier public television station, serving the largest market in the country. The station is also one of the major producers of national programming on PBS, originating such signature series as NATURE; Live from Lincoln Center; American Masters; Religion & Ethics Newsweekly; Cyberchase; and Need to Know. The WNET Tape Archives has more than 35,000 tapes and other materials and its holdings go back more than 50 years. WNET is also deeply involved in teacher education and adapting our productions for use in the K-12 classroom.
When the station acquired a new Digital Asset Management (DAM) system made by DALET, the metadata fields were configured from the start to be PBCore compliant. As DALET is rapidly becoming the primary in-house system for digital editing and storage, PBCore records are created at the time new files are ingested, and then supplemented with additional information as the production moves along.
The DAM allows multiple departments each to have access to the same information about any given asset, and PBCore is key to creating a standard structure and vocabulary for content, regardless of format or origination. It has enabled the Archive to get access to program information immediately, at the same time it facilitates program files being used by various departments with different requirements, such Interactive (on-line content) and Creative (interstitials and promos) as well as individual production units.
One of the more important applications has been using PBCore to help describe segments used in WNET’s VITAL (Video in Teaching and Learning) service, which packages public television program materials into segments geared to classroom curricula. VITAL educational materials are shared with other PTV stations through Teacher’s Domain and PBS’s Digital Learning Library, and PBCore assures standardization of information across these and similar program exchanges.
Jonathan Marmor
WNET.org
450 W. 33rd St
NYC 10001
212-560-3151
MarmorJ@thirteen.org
PBCore was first used in the Democracy Now! archive by our former archivist, David Rice, but we continue to use it because it fits our needs. We have limited resources and rely on interns to catalog our production material, so we need a metadata standard with which one can familiarize oneself quickly, but can also accommodate details including content description, and instantiation information such as media format and specifications. Although our archive contains a great deal of physical media, our production team now works in a “tapeless” environment. Keeping track of duplicates or redundant copies of file-based media would be difficult without PBCore’s instantiation elements.
We are using PBCore to archive raw footage that is shot in the Democracy Now! studio or in the field, along with edit masters and other footage that is incorporated into our daily broadcast. Democracy Now! interns watch all incoming production footage and catalog it, improve records from media assets that are no longer in production and have been prioritized for preservation, and use PBCore to catalog unprocessed materials that may or may not have been used for production. We also subscribe to Reuters’ live and packaged feeds, which are delivered with .xml records that we can map into our PBCore FileMaker Pro database. We have also spoken with other PBCore users about footage swaps, but have not yet acquired any new material or corresponding records.
Before adopting PBCore, our archive was not using a formal cataloging standard, so implementing a new system did not involve a migration of a large body of data. Spreadsheet information and metadata, for example, from archive.org, have been incorporated into our PBCore database. Our FileMaker Pro database was designed by David Rice, and a considerable amount of time was invested in making the database compliant and able to output records that pass PBCore validation. Making this FileMaker database or one like it available for free to the A/V archives community might be a great step for PBCore, because although FileMaker Pro has many limitations, it is also one of the only database software products available that is easily edited and reconfigured by users. Our Filemaker/PBCore database suits our needs very well with few exceptions, allowing us to access our content in meaningful ways in our busy, newsroom environment.
Nicole Martin
Archivist
212-431-9090
nmartin@democracynow.org
Sharing content between broadcasters used to mean shipping tapes back and forth, which inhibited access and discovery, raised costs, and increased the likelihood of damage. Digital technologies have made file sharing more practical as an access point, but most file sharing on the Internet has been geared towards consumer applications or one-size-fits-all approaches. Previous attempts at program sharing systems also lacked file and metadata standards and compliancy checks required for a seamless experience across vendor equipment. When you download a program, you need to know it will play at your station and fit in with the rest of your workflow.
The Alliance for Community Media, a nonprofit national membership organization that represents over 3,000 Public, Educational and Government (PEG) access organizations across the country, wanted to launch the Community Media Distribution Network (CMDN), an Internet-based broadcast content sharing and distribution system to help make quality video programming available to PEG stations. The ACM created a Standards Working Group to define the program sharing system. Core goals included; a common file format that would be compatible on the greatest variety of playback systems, an extensible metadata standard that met the needs of community media, and seamless integration with popular Content Management Systems (CMS). The vision was a centralized repository of content that could be browsed online, and with a single click, content would be downloaded into a station’s local CMS or playback server along with the metadata.
The ACM Standards Working Group decided early on that it would be best to try to leverage an existing, well thought out metadata standard rather than reinvent the wheel. PBCore quickly rose to the top as a great fit due to the many parallel needs between PEG and Public Broadcasting organizations.
The ACM Standards Working Group selected PBCore as the metadata standard for the Community Media Distribution Network. A subset of the PBCore fields was designated as required fields, and the PBCore extensions functionality was leveraged to include Tags, Producer Contact Information, and Indemnification information. A front-end user interface and a set of APIs were outlined to allow fully automated program upload/download workflow where PBCore metadata follows the program to eliminate double entry. Additional features of the CMDN include; search, online video previews, and download reports.
After an RFP process, TelVue Corporation was selected to develop the CMDN due to TelVue’s prior experience developing web-based broadcast automation and content management applications, and other broadcast file sharing systems. The CMDN is now fully operational. TelVue hosts the centralized system, developed as a Ruby on Rails web application with back end storage, in its National Data Center. Content management and playback systems can auto publish to the CMDN using the APIs, or contributors can use the web-based interface to upload content and fill out the required metadata fields. Users can browse programming online, and either download directly form their web browser, or add program to their download queue for their integrated CMS to fetch in the background. For example, TelVue Princeton broadcast video servers support the ACM PBCore metadata natively, and can both publish to and download from the CMDN via a single click.
Currently there are over 150 users of the CMDN, sharing hundreds of hours of broadcast quality programming, and the numbers are growing. Content producers including Democracy Now! and Perils for Pedestrians distribute video programming on the CMDN. TelVue is ready to support PBCore 2.0 in the CMDN if the ACM Standard Working Group determines the need, for example, to phase out the custom extensions, simplify rights management, and provide new features.
Jesse Lerman
President & CEO
TelVue Corporation
16000 Horizon Way, Suite 500
Mt. Laurel, NJ 08054
800.885.8886 ext. 101 (Phone)
856.866.7411 (Fax)
jlerman@telvue.com
http://www.telvue.com
When WHYY Philadelphia wanted to create a searchable, expandable digital archive of over 30 years of one of public radio’s most popular and successful programs, Fresh Air, they chose PBCore for the metadata format. With its robust field set and integrated support for multiple instantiations, PBCore was a strong fit to take what was previously disparate data and notes, mostly in documents and spreadsheets, and create a single record set.
After working with AudioVisual Preservation Solutions to create 5,000+ PBCore 1.2.1 records, WHYY turned to broadcast technology innovator TelVue Corporation to create a digital archive system that could store all of the records, data, and audio programs. WHYY chose TelVue because TelVue had previously worked with PBCore on several projects, and already offered digital media servers for audio and video archive and retrieval that could store terabytes of data. TelVue expanded their content management application to support the complete PBCore 1.2.1 dictionary, including full-text searching of all records. The resultant TelVue digital archive server can ingest existing PBCore records, create new records, search across all metadata fields, preview media, and link digital content as instantiations to each record. All features are accessible through a browser-based interface developed in the Ruby on Rails web application framework.
WHYY uses the TelVue system today to archive all of Fresh Air making the library more accessible for reuse, collaboration, Internet distribution, and as niche content for specific community, service, and institutional needs. As it continues to standardize on PBCore, WHYY will soon expand its use of the TelVue PBCore archive server to other programs produced at WHYY. TelVue is exploring full support for the forthcoming PBCore 2.0, based on WHYY’s needs.
Jesse Lerman
President & CEO
TelVue Corporation
16000 Horizon Way, Suite 500
Mt. Laurel, NJ 08054
800.885.8886 ext. 101
jlerman@telvue.com
http://www.telvue.com
WHYY uses PBCore in its audio archiving solution for radio. Working with consultant David Rice and hardware vendor TelVue, WHYY was able to develop a searchable digital library for Fresh Air — an NPR program that has been produced on a daily basis since 1987.
There are over 6000 episodes of Fresh Air to date. Up to July of 2007, these programs were edited and archived on analog tape. Since then, Fresh Air has been produced using ProTools Digital Audio Workstations. Pre-July 2007 programs existed in a 2 track edited format on a combination of analog ¼” tape, and since the mid-1990’s also on DAT and more recently, CD-R media. All these were collected, restored when necessary, and digitized into a common .wav format. Since July of 2007, all original source material has also been preserved.
Using PBCore, WHYY has been able to create metadata, documenting and cross-referencing all media types. In addition, using the description fields, all show descriptions and available transcripts have been made searchable using the familiar “keyword” technique.
WHYY chose PBCore for the Fresh Air archive, as well as enterprise wide audio and video archiving because of its open source origin, previous use in the Public Media environment and its flexibility. Using the Fresh Air PBCore model, we are currently expanding the WHYY digital archive, creating metadata for our other audio, and video and web assets.
Julian Herzfeld
Media Technology Manager
WHYY, Inc.
215-351-1278
jherzfeld@whyy.org
Development funding for PBCore provided by the Corporation for Public Broadcasting 