The Performing Arts Data Service (PADS) at the University of Glasgow is one of five distributed service providers of the Arts and Humanities Data Service. The PADS has a dual remit -- to collect, preserve and promote re-use of digital data resources in the areas of dance, theatre, music and film and broadcast media and to specialise in the networked delivery of time-based data resources such as music and moving images. During 1997-99, the PADS played a key role as one of two pilot sites in the Imagination/Universities Network Pilot Project, a joint initiative funded by the Joint Information Systems Committee (JISC); the British Film Institute (BFI) and the British Universities Film and Video Council (BUFVC). Now, building on the success of that project, the PADS is working with a number of UK film archives to host on-line access to their catalogues as well as providing support and practical advice to institutions wishing to embark on full-scale digitisation projects.
This article does not outline the technical challenges inherent in delivering high-quality moving images over networks, as these have already been documented in detail in the final report of the Imagination Universities Network Pilot Project (http://www.pads.ahds.ac.uk/ImaginationPilotProjectCollection) and elsewhere. This article seeks to present the challenges facing library professionals and those working in film archives in building on-line collections of data for use in film research and teaching.
To digitise or not to digitise?
A brief search of the world wide web will throw up an increasing number of film collections available on-line, but these are mainly commercial organisations based in the USA which include a number of low-demand, freely available films streamed on demand as "tasters" for pay-per-view resources which can typically attract premiums of 10-20 dollars. Educational and publicly-funded archive sites are generally limited to on-line supply of relatively obscure, rights-cleared materials and usually offer a less technologically advanced service than their commercial counterparts.
Whilst it still remains rather a thrill for the casual user to access unfamiliar digital film over the web, Ian Christie, speaking at the Computer-Assisted Film and Drama Studies Conference in Oxford in 1997, drew attention to "... a very strong coherent purism built into film studies which believes in the sanctity of the 35mm print, so much so that many in the field feel their culture being threatened by digitisation." Digitisation is not necessarily, therefore, the holy grail of film studies that many in the digital library world assume. The enormous costs involved (commercial providers can charge in excess of 50 pounds per minute to create digital film), fears of obsolesence, and the demands of rights-holders also make large-scale digitisation projects daunting and, in many cases, wholly impractical.
However, the electronic world can offer much to film research. Digital film collections offer unprecedented access to often rare and hard-to-source materials. They can aid in the battle to preserve fragile materials by creating easily-duplicated surrogates that should be migrateable as new technologies emerge. Even if large-scale digitisation projects are impractical, carefully-focussed ones targetting specific interest groups are achievable. If digitisation of archive material is not possible, on-line access to catalogue data is within the reach of most organisations and can revolutionise the way researchers approach materials. Digital collections can be extremely flexible and should not rely wholly on moving images alone. "Added value" resources such as scripts, reviews and stills can be crucial in promoting understanding of both individual resources and the work of an archival organisation itself.
Added value in digital collections
The PADS' role in the Imagination/Universities Network Pilot project was a "proof-of-concept" presentation of a fully-integrated site for research and teaching, including both the highest-quality moving images and a representative selection of supporting materials including OCR extracts of scholarly monographs, digital images of working scripts and HTML versions of contemporary criticism and review. Whilst purists may still wish to seek out the original 35mm print, the access afforded by the Imagination/Universities approach created a revolutionary tool for teaching which was piloted by academics in Glasgow University's Department of Theatre, Film and Television and evaluated by colleagues from a number of other leading teaching institutions.
Lecturers in Glasgow were able, for example, to present the two different versions of "Blackmail", an early Hitchcock production that was the first British film to have a soundtrack, and to discuss in class the techniques used to convey the meaning of the piece in both its silent and sound versions. Students could then access the PADS site in computer labs on-campus (the system was accessible only through the SuperJANET network), watch the films on-demand and read supporting documentation.
The Cinemedia Project (http://www.cinemedia.net/) in Australia has provided a useful model for the PADS in developing on-line resources for academic use. Their SWIFT (http://www.cinemedia.net/SWIFT/) database provides educational users with online access to film and video titles, on demand, via commercial broadband delivery networks and is currently working as a pilot with 300 titles available on a pay-per-view basis. Examples of added value resources to enhance on-line video delivery suggested by Cinemedia include:
However, as Sarah Porter and Michael Fraser pointed out in their report from the 1997 Computer-Assisted Film and Drama Studies Conference, added value resources often come at a cost, "... their use in a teaching and learning context needs to be considered with care; the substantial expert input needed at the authoring stage of a complex resource should not be underestimated."
Enquiry Management and Charging Policies
It is the PADS' view that online resources (and especially those presented in an added value context) have the potential to pay for themselves. In hard currency terms, a pay-per-view system such as Cinemedia's SWIFT project allows users to operate a simple web interface to access titles and pay a per view usage charge. Copyright owners are provided with flexible control over access to, and charges made, for individual titles, full usage records and automatic disbursement of returns with access rights monitored by SWIFT's Copyright Management Software.
However, operational gains are also feasible. Staff time taken up with searches on behalf of clients can be re-allocated if the catalogue is available on-line. Screening facilities and time can be saved if key films are available in digital form. Many funding bodies, for whom access is a primary concept, are likely to look favourably on organisations which have sought to improve access through new technologies. On-line facilities also allow archives to market their services to new customers and to make links with commercial organisations who may wish to exploit their collections.
For operational gains to be made, it is necessary to automate activities such as enquiry management. The North-West Film Archive (http://www.nwfa.mmu.ac.uk/) at Manchester Metropolitan University has mounted an online enquiry form along with a fully-searchable version of its catalogue on its website and found that enquiries have become more focussed and relate to specific films rather than general questions about the content of the archive as a whole, which can often be the first of a string of queries, taking up valuable staff time.
Documentation and Film Cataloguing
A key principle of the Arts and Humanities Data Service since its inception has been the promotion both of interoperability and of documentation standards appropriate to digital media, and the PADS has been keen to apply these principles to film and broadcast media datasets. The AHDS is committed to the development of the Dublin Core Metadata Set as a standard for description, and a Dublin Core record must be available for data to be accessible through the AHDS portal.
Three levels of description are routinely used to allow interoperable gateway access to resources: a collection level Dublin Core record where data is extremely complex and dependent on its proprietary format (for example historical census records stored in a database); a series of newly-created records relating to digital objects such as digitised photographs or musical scores; and collections of "mapped" catalogue records where the original database (for example, the catalogue of holdings at a film archive) is filtered through a Dublin Core template, and the original records are made available in an interoperable environment.
The PADS is currently offering Dublin Core access to two different types of records. The Imagination/Universities Network Pilot Project data was created to describe discrete digital objects -- films and textual data. Projects under way with the Scottish Film and Television Archive (http://www.scottishscreen.com/) and with The Bill Douglas Film Centre at the University of Exeter (http://www.ex.ac.uk/bill.douglas/) concentrate on creating on-line access to the catalogues of these important research bodies. Data from these organisations is "mapped" to the Dublin Core template, with additional contextual data added where appropriate.
THE DUBLIN CORE ELEMENT SET
The Dublin Core is a simple set of 15 basic elements which necessarily precludes the kind of detailed description practices that subject-specialists in the performing arts usually require. It was clear to the PADS that it would be necessary to use a system of qualifiers to "tag" the data in such a way that it was meaningful to film researchers. Most obviously, the term "creator" does not adequately express the myriad tasks involved in the creation of a moving image resource. A film directed by Alfred Hitchcock is something rather different from one in which Hitchcock's contribution was in a lesser capacity. The PADS was loathe to create a large number of subject-specific qualifiers such as DC_creator_director = Alfred Hitchcock, as it was evident that these could proliferate at an alarming rate across the performing arts (DC_creator_composer; DC_creator_choreographer; DC_creator_keygrip.... ). Another profound difficulty is the lack of universal description standards in the film archive community, with the result that any proposed subject-specific qualifiers were unlikely to be applicable to all collections.
Our solution has been to create a partnered field structure which numbers each creator (or any other element where there are likely to be multiple values) and assigns an additional role qualifier for insertion of supporting contextual detail:
DC_creator_1 = Alfred Hitchcock
DC_date_1 = 1936
Another problem endemic to film resources is the sheer weight of potential individual creators. A full-length feature film such as "Titanic" can depend on the creative contribution of hundreds, perhaps thousands of individuals. Clearly, in consideration of the sanity of the unfortunate cataloguer, if nothing else, to delineate each individual would be inappropriate. We were also concerned by the ambiguity of the term "role" in creative film-making. Our solution was to identify the primary intellectual contributors to the production of the film (in most cases these included director, producer, author of the screenplay, composer of the score, etc.) and delineate their contributions individually. For all other contributors, we pooled data into "cast" and "credits" fields as below:
DC_creator_6 = PC George Dixon = Jack Warner, Andy Crawford = Peter Byrne, PC "Tubb" Barrell = Neil Wilson, Sgt. Flint = Arthur Rigby, Sgt. Grace Millard = Moira Mannion, Alderman Mayhew = Geoffrey Wincott, Tom Carr = Paul Eddington, "Captain" = Geoffrey King, Mr Collings = Christopher Hodge, Mary Dixon = Jeannette Hutchinson, Inspector Cherry CID = Robert Cawdron
Added value in catalogue records
Librarians and archivists have historically favoured different approaches to documentation which can be crudely categorised as "hands on" and "hands off". The basic library catalogue record facilitates simple identification and location of an article -- its primary role is access. The archive approach seeks to limit access to rare and often fragile items by providing surrogate material which will satisfy the needs of all but the most demanding of researchers -- it seeks to limit access by providing as much detail as possible.
There is currently much debate within the electronic documentation community about the concept of "added value" records which include information (for example, tracking the history of a digital object through its analogue life) that transcends the notion of a discrete one-object, one-record structure. Dublin Core records from the Imagination/Universities Network Pilot Project include information about the analogue version of the digitised films because this is regarded as crucial information for understanding the development of film and television history -- it had no value as an access tool because the digital films were immediately accessible from the server during the life of the project. Traditional added value tools for film archivists, such as shot-lists and textual descriptions of the intellectual content of the resource, are a key element in on-line catalogues, allowing users to refine searches and pinpoint individual resources, but they are also valuable in the promotion of collections. With these added value elements, the catalogue itself becomes entertaining and informative and can be marketed to a wider audience.
Crucially, most film archives and depositories do not control the rights to the majority of items preserved in their collections, and existing deposit agreements detailing the dissemination of analogue copies of archived film are unlikely to cover the kind of widespread access afforded by network technologies. Increasingly, film archives are being asked to negotiate licences for new usages (for example, supplying Beta SP tapes to website owners keen to digitise moving images for use on the Internet) which were never envisaged when original deposit agreements were drawn up. It may prove difficult to communicate the implications of this kind of data exploitation to rights holders if film archives themselves are unclear about the technologies used (for example, the crucial difference between downloadable files and streaming technologies).
Managing the rights for a large-scale digitisation project may prove significantly more expensive than undertaking the digitisation process itself. Many film resources will be subject to a number of multiple and cascading rights: permissions must be sought for mechanical, performative and underlying rights along with the primary intellectual property rights enjoyed by the film-maker(s). A common difficulty resides in the underlying rights attached to musical soundtracks. At present, music licensing agencies such as the MCPS (Mechanical Copyright Protection Society) are only tentatively exploring the opportunities and threats created by network technologies, and whilst commercial licences for limited use of digitised music resources are readily available, both educational use and underlying rights usages remain a grey area. Therefore it is necessary to clear all music rights by a laborious and expensive process of individual application to publishers, record companies and other rights-owning parties. A long-term goal of the PADS is to influence national policies on rights management in digitised music and moving images through the creation of working partnerships with licensing agencies such as the MCPS and the Performing Right Society (PRS) and through working groups at JISC and other key educational bodies.
Much of the history and culture of the Twentieth Century is reflected in moving image and audio material. Loss of these kinds of archived resources would result in the loss of substantial amounts of our heritage and loss of crucial materials for research, learning and teaching. Yet preserving these collections is expensive and fraught with great conflict over rights management. Projects such as the Performing Arts Data Service are providing the catalyst to address digitisation, information management and access and retrieval issues. They also facilitate the collaborative framework for addressing rights negotiation and clearance, which may prove the greatest obstacle to long-term preservation and access to our moving image heritage.
The Performing Arts Data Service can advise on any (or all!) of the issues raised in this article. Please contact Catherine Owen (firstname.lastname@example.org), PADS Manager, for more details. Findings and metadata records from the Imagination/Universities Network Pilot Project are available from the PADS website (http://www.pads.ahds.ac.uk), along with a series of links pages of interest to film scholars and data holders. The on-line catalogue of the Scottish Film and Television Archive will be officially launched from the PADS site on 28th March 2000.
Performing Arts Data Service
Imagination/Universities Network Pilot Project
Scottish Film and Television Archive
North-West Film Archive, Manchester Metropolitan University
Bill Douglas Film Centre, University of Exeter
Duffy, Celia and Owen, Catherine: Resource Discovery Workshops: Moving Image Resources - Report from the Performing Arts Data Service
Pearson, Tony: The PADS Project at the University of Glasgow, paper for the INFOG99 Conference, Melbourne, Australia, Screening the Past, November 1999
Hiley, Nicholas: Archiving Issues in Film and Broadcast Media: a report for the Research Support Libraries Programme
Ballantyne, Jim: Resources for Film and Media Studies in the United Kingdom: a report for the Research Support Libraries Programme
Schlesinger, Philip: Academic Researchers and Film and Media Studies Resources: a report for the Research Support Libraries Programme
Duffy, Celia: Networking Moving Images, Computers and Texts no 16/17, Winter 1998
Puttnam, David: Screen Literacy and Cultural Morality in the Digital Era, The 1999 Grierson Lecture, Delivered at Cinema 2 Village Centre Melbourne, 23rd November, 1999
Christie, Ian: Old Films, New Media (reported in
Fraser, Michael and Porter, Sarah: Computer-Assisted Film and Drama Studies Conference, 17th March 1997, Oxford, CTI Textual Studies), Computers and Texts no 14, April 1997
Copyright © Catherine Owen, Tony Pearson, and Stephen Arnold
|Top | Contents
Search | Author Index | Title Index | Monthly Issues
Previous story | Next story
Home | E-mail the Editor
D-Lib Magazine Access Terms and Conditions