Gary Handman, Director
Media Resources Center,
UC Berkeley
February 1998

I. Executive Summary

This proposal is for the development of a technical, administrative, and service model for identifying, selecting, cataloging, editing, digitizing, storing, and delivering a library of diverse sound and moving image files. Materials targeted for this project would include:

1. Campus-produced audio and video recordings (notable lectures, speeches, events, etc.) currently existing in the library or elsewhere on campus;

2. Archival materials or unique recordings currently existing in library special collections or elsewhere on campus;

3. Commercially-produced non-theatrical (non-fiction) works for which permission to digitize has been secured;

4. Teaching and study resources, such as lectures, field recordings, lab recordings; other course-reserve viewing materials.

If successful, this project would hold significant applications for a number of existing collections on campus, including The Bancroft Library and the Pacific Film Archives. A successful project may also provide the basis for future distance learning/distance information service operations and programs, including UC Extension, and the California Digital Library. Since the project would represent one of the few practical video-based digital library initiatives undertaken by a US academic library, it may also serve as a model for other institutions nationally.

II. Background and Rationale for Project

Electronic media such as film, video, television, and sound recording are arguable the most significant cultural shaping forces of the 20th Century. Not surprisingly, the use of these media in teaching and research has increased tremendously across discipines in the last decade at UCB. For some undergraduate programs, such as Ethnic Stuides, various courses taught for the American Cultures requirment, Film Studies, Rhetoric and English, and various area studies, this increase has been particularly dramatic.

Since its inception in 1979, the Media Resources Center (Moffitt Library) has evolved into one of the largest and strongest curated collections of commercially-produced non-theatrical (non-fiction) videos in a US research library--certainly the strongest such collection in the State. MRC is presently the primary facility on campus supporting curricular and research use of video. Like virtually all media libraries and archives around the world, MRC's holdings currently consist almost exclusively of analog media (in MRC's case, videotape and audiotape). There are, however, increasingly compelling reasons and expanding opportunities to begin thinking and experimenting with newer models and methods for storing, packaging, and delivering moving image and sound resources. The chief benefits of migrating media resources to a digital library environment include:

1. Expanded access to resources: access to resources from remote locations, including classrooms, residence units and off-campus sites; individual access by multiple users at one time; opportunities for bringing together and providing improved access to diverse resources on campus, both in the library and in other academic units.

2. Increased flexibility of use: on-demand access to resources; the ability to combine or associate various media in unique ways (i.e. integration of text, still and motion images); flexibility in editing or otherwise repurposing media works for particular academic uses.

4. Preservation of materials: tape-based media such as video tape typically have a shelf-life of well under 20 years, after which time the physical medium and the encoded image degrade substantially. This life-span decreases substantially under conditions of intensive use, or use on older equipment.

The multimedia server initiative would initially focus on several discrete test projects:

1) Cataloging and mounting selected audio recordings from the Berkeley Language Center's Speech Archives (Berkeley Language Center)

2) Developing multimedia capabilities in the existing Media Resource Center web site:

--Linking 5-10 minute clips of videos owned by MRC to the videographic record for the item in MRC's web site;

--Expanding MRC's web site to include relevant video-based course web pages (i.e., mounting on a library multimedia server course-related web sites which include digitized materials from the MRC collection).

3) Developing a collection of unique, locally-produced moving image materials drawn from library collections and from the uncataloged holdings of other campus units (teaching departments, etc.)

Over the past 25 years, the Berkeley Language Laboratory has developed a audiorecording collection of approximately 1200 notable lectures and speeches given on campus by cultural and academic luminaries ranging from Alfred Kroeber and Aldous Huxley to Carlos Fuentes and Malcolm X. The physical collection currently resides in the Language Lab (basement of Dwinelle Hall). A large portion of these unique recordings are on reel-to-reel tape and accessible only on-site in the Language Lab. Bibliographic access to the materials is limited to a local database management program/print listing maintained by the Lab.

The director of the Language Lab (Mark Kaiser) and Gary Handman (head of MRC) have developed a trial plan to digitize an initial 5-10 records (@ 1-2 hours each), to initially mount these tests on the Library's SunSite, and to make them accessible via streamed audio, using the techincal model currently in place in the Music Library (MusiLan Project). The digitized recordings will be fully cataloged by MRC's cataloger. Records in Pathfinder and MelWeb will include a URL for the audiofile (i.e., the complete recording will be accessible online via the bibliographic record for the work).

If, as is expected, this trial project is successful, MRC and the Language Lab will develop a plan to proceed with digitizing other items in the Speech Archives. This larger body of work would be migrated to the proposed library multimedia server.

The Speech Archives project is potentially important for a number of reasons: it represents a unique collaboration between the library and another instructional technology unit on campus; it will provide substantially improved access to a body of significant historical materials; it will provide a working model for similar projects involving spoken word collections (e.g., ROHO, other MRC holdings, campus collections outside of the library). The project will also represent the first attempt to provide access to full-length, widely-accessible audio "texts" via the library's online catalog.

B. Multimedia Web Site

Over the course of the last four years, MRC's web site (www.lib.berkeley.edu/MRC) has evolved into perhaps the largest web-based videographic resource for non-theatrical (non-fiction) videos in the world. The site has been designed to perform a number of inter-related functions:

1. Provide basic videographic access to MRC's collections

2. Provide related study and research information for specific MRC holdings (including reviews and article citations; links to full-text reviews and articles; links to related web sites and databases)

3. Provide links to general and specialized media resources on the web (including links to film scholarship sites; links to professional resources for librarians and educators; links to information about new and evolving media formats and uses).

4. Provide a virtual "community meeting place" for the international community of independent documentary film producers, distributors, media librarians, and media users. Serve as a clearinghouse for information about non-fiction film and film scholarship. Serve as a resource for video collection development. The MRC's web development has been principally guided by the fact that there is currently no effective national videographic apparatus (nothing vaguely resembling a Books in Print for video). Reviews for non-theatrical features are, furthermore, remarkably scarce; opportunities for previewing important new documentary features and other independently-produced works are few and far between.

The Multimedia Server project would attempt to extend the current features of MRC's web site in a number of ways:

1) Holdings Previews

Inclusion of streamed video clips (perhaps 5-10 minutes each) in the videographic records of selected MRC holdings. Users of MRC's web catalog would consequently be able to preview a short, fairly high-resolution segment from particular videos in the collection. One of the drawbacks of non-theatrical video as a teaching and research medium is the fact that these visual "texts" are not widely reviewed; they cannot, furthermore, be easily skimmed for style and content. Inclusion of video clips would be a step toward providing this kind of preview capability on line.

For a primitive example of how this might work, see http://www.lib.berkeley.edu/MRC/fight.html. In this case, the video clip is provided in QuickTime via a producer's web site; the MRC clip would be considerably larger, longer, and of higher visual resolution.

The head of MRC has been in contact with a number of Bay Area video distributors (California Newsreel, National Asian American Telecommunications Association, Flower Films) who have expressed interest in participating in this project.

2) Hosted Multimedia Course Sites

Over the last several years, MRC experimented with hosting web pages for film-related student research (see for e.g. http://www.lib.berkeley.edu/MRC/Amydoc.html and http://www.lib.berkeley.edu/MRC/Lyford.html). There has been great interest expressed by a number of MRC's primary faculty users to work with the Center on developing course-related film studies sites which would include segments from videos in the MRC collection. These sites would be used both for individual study and, possibly, for use in classroom teaching. Faculty from several film courses and American Cultures courses have expressed interest in participating in this project.

3) Digital Video Collection

While the majority of MRC's collections are commercially-produced and distributed titles, there is also a sizeable number of videos in the collection which are unique to campus, such as local recordings of notable lectures, speeches, and events. There is also a certain amount of primary source material owned by MRC that is either in the public domain, or copyright-cleared. Because of the relatively short life-span of tape-based media, much of this unique material is physically at risk. One aspect of the Multimedia Server Project would be to digitize and recatalog selected titles from this body of work (because this material is property of the university, it would most likely be possible to provide access to much of it via Pathfinder and MELWeb, as well as the MRC web site). A second phase would be to identify important locally-produced materials held by other units on campus, and to include them in the Digital Video Collection, if possible.

4) Digital Course Lectures

A number of teaching departments regularly tape large lectures and labs for student viewing and review. Because of MRC's limited space and the large number of students in these classes, the Center largely stopped providing individual access to these these lecture tapes a number of years ago (in some instances, the Office of Media Service holds group viewings of these tapes, but access to these tapes is generally insufficient to meet demand). A dedicated library multimedia server might allow digitization and on-demand viewing for some of this material. This aspect of the project would most likely require partnering with individual academic departments for staff funding.