Volume 26, Number 5
Jul 2003

Preservation Reformatting Conference: Digital Technology vs. Analog Technology

About 180 representatives of large and small institutions attended the 18^th annual conference of the National Archives and Records Administration held March 27, 2003, in the auditorium of Archives II in College Park, Maryland. The theme of the conference was preservation reformatting, with an emphasis on digital technologies reflecting the current experience of the speakers. Speakers included Steven Puglia, Special Media Laboratory, NARA; Howard Besser, Moving Image Preservation Program, New York University; Carl Fleischhauer, National Digital Library, Library of Congress; Ed Zwaneveld, former member of National Film Board of Canada; Abby Smith, Director of Programs, Council on Library and Information Resources; Stephen Chapman, Weissman Preservation Center, Harvard University; and Ken Thibodeau, Electronic Records Archives (ERA) Program, NARA. NARA staff video-recorded the proceedings and will make available on their website, <http://www.archives.gov/preservation>, the text of some of the presentations. Other presentations not linked to the NARA website are available from sources indicated below.

Anyone who expected a review of standards or sought solutions for specific technical problems must have been disappointed, if not bewildered.

• The speakers arrived at few conclusions but emphasized the following:
• Digital conversion, when selected as a preservation solution, will require substantial sums of money.
• Most media chosen for preservation will require constant managed environments.
• All institutions small or large will eventually face the need to manage digital media.
• It is essential that institutions establish clear parameters for preservation and reformatting.
• There are no agreed upon standards, or best practices, for long-term preservation of digital media.
• Vendors are selling products, NOT solutions, to your problems.
• There is no archival medium that when properly stored can be expected to retain information forever so that such information can be retrieved without loss.
• The archival community may, or may not, be large enough to warrant the attention of manufacturers and vendors.
• Trends suggest that all types of researchers are seeking and demanding convenient access and often will prefer digital surrogates over the original artifact for every day work.
• Functionality of most electronic media is always compromised when reformatted in more permanent media.
• Obsolescence is created by industry as well as by user needs, expectations and demands.

Steve Puglia provided an overview of the variety of analog and digital media, which he categorized as static analog, dynamic analog and digital. He noted the increasing complexity and instability of newer media and the correspondingly increasing cost of preserving them. Although each medium has its own technical requirements, for dynamic media a constantly managed environment is essential to achieve the desired life expectancies. In an unmanaged environment, for example, digital media have lifetimes of about five years.

To maintain managed environments indefinitely will require substantial sums of money. For digital environments, maintenance costs over the life cycle of the entire supporting system are several times as high as their initial costs. Long-term strategy requires the use of each of the various technologies to its best advantage to minimize the disadvantage of the others. Approaches for reproducing tone with film negatives were compared with digital scanning, and merits of each discussed.

According to Puglia, a systems perspective is necessary for the management of digital artifacts so that there is no break in the necessary chain of migrations. Digital artifacts require more active and continual handling than physical records, as well as strategies for preserving equipment and software applications that quickly become obsolete.

The market continues to push image, sound and text records towards short-lived digital processes and products. Most of these cannot be preserved by creating more permanent surrogates on paper or microfilm without the loss of functionality. There is no single or simple answer.

Howard Besser spoke about the importance of moving images for the understanding of culture and the specific preservation needs of moving images. Because of the evolving nature of processes, moving images have always demanded the reformatting and handling of a wide variety of formats. However, digital preservation requires very different thinking and acceptance of a paradigm shift from managing artifacts to managing disembodied information over long periods. Management over time consists of periodic refreshing, emulation, migration and other strategies as they emerge.

Digital preservation must address old problems like equipment obsolescence, and need for technical expertise, as well as new ones like data loss due to compression, lack of persistent links for embedded works, lack of encoding and transmission standards, and exceptional custodial capacity. Institutions must attain an exceptional level of administrative and legal responsibility; achieve high levels of system security, organizational viability, financial sustainability, accountability and technological relevance. Besser addressed one important link in the creation of such institutions, the education and training of future moving image administrators.

An extensive list of works by Howard Besser on every aspect of moving image preservation is available at <http://www.tisch.nyu.edu/preservation>.

Carl Fleischhauer drew on his experience with audiovisual (AV) materials to focus attention on the proliferation of digital items in all collections, and the prevailing preoccupation with the upper limits of available technology. He highlighted shortcomings in the reformatting of AV media: limited life spans; quality loss; information loss from generation to generation; device and format obsolescence; proprietary nature of signal lay-down processes; and increasing complexity of processing requiring technical expertise.

Fleischhauer specifically noted the need to retain ancillary materials that accompany AV materials such as album covers, promotional literature, teaching and study guides and text of lyrics, and to consider preserving playback devices. He also recommends the preparation and preservation of extensive administrative, descriptive and rights bibliographic records or metadata when reformatting. Besides the traditional minimum of author, title date, producer and physical characteristics, this data should include technical information—frequencies, speeds, luminescence, tone references, etc.—about the source item as well its preservation treatment and needs.

Ed Zwaneveld expressed confidence and conviction (not shared by the others) that the archival community can and should exert pressure on manufacturers and vendors—who set industry standards—to provide archival media that meet necessary permanence and quality standards. He believes that new markets and services may grow from this demand.

Zwaneveld recommended that archival institutions refuse to accept poor quality materials and that they insist on stable masters and the highest quality of production for deposit items. The burden, including cost of producing stable masters, should fall on the producer, not the archive. This is especially important for materials like films, that have substantial residual value when reproduced and remarketed for profit. Plans for the production of permanent masters have to be made when production costs are allocated and money is available to the producer.

To Zwaneveld, perpetual migration is an option from hell, an unsustainable solution that leads to needless waste. Some technologies such as polyester-based panchromatic black and white film, etched glass DVDs and nickel stampers for color and sound data may provide longer lasting carriers.

Abby Smith argued for higher curatorial standards in the representation of artifacts online. Beyond technical proficiency, factors contributing to higher quality representations include editorial and subject matter expertise; knowledge of research methods and the use of supporting resources; and the provision of rich metadata that facilitates web searching of individual electronic files.

Digital enhancement of artifacts should seek to overcome the constraints of the source material and contribute to the density of information available to researchers. As an example, Smith described the digital "stitching" for computer screen viewing of a river that is impossible to portray on the source book folio-size page. Smith also emphasized the need for better stewardship of the originals, shared repositories specifically designed for long-term preservation, an accessible registry of surrogates, and expansion of access beyond academic and scholarly circles.

Smith illustrated her presentation on digital capture and enhancement with images from an 18^th century atlas of St. Petersburg, Russia, depth charts of San Francisco Bay, and historical maps of the imperial complexes in Tokyo, Japan. Software applications now available make possible layering and georectification of maps created under different mapping conventions and blending of topographic and bathymetric data. For more information on map digitization and mapping software, you are invited to visit <http://www.davidrumsey.com> and <http://www.esri.com>.

Steve Chapman presented specific criteria and costs of a mix of preservation reformatting services provided by Harvard University Library. Chapman, like other presenters, noted that preservation choices have to be tailored for each project but cautioned that there are as yet no standards for digital formats. For now, these can best be informed by comparison of vendor services, samples and price.

Chapman elaborated on production strategies for text and images. In addition, he made comparisons between technologies factoring quality and type of production, data storage, search and retrieval costs. Details of various services provided by the Weissman Preservation Center at Harvard are available online at <http://preserve.harvard.edu/services>.

Ken Thibodeau narrated the research progress of NARA's research partners in developing the Electronic Records Archives (ERA) program. The challenge presented to the partners is to authentically preserve and provide access to any kind of electronic record, free from dependency on any specific hardware or software, enabling NARA to carry out its charge. The charge is to preserve records for access now and for the life of the republic. This effort at a cost of $2 billion is to provide a building block for the proposed national e-government. ERA's partners now include Georgia Tech, San Diego Supercomputer Center, the InterPARES project, the Library of Congress, NASA, and Army Research Laboratory, among many others.

A wide variety of models have been proposed and investigated. From this investigation a complex three-by-three arrays of independently acting layers of data, services, and processes have been formulated. These formulations will serve as criteria for experimental archival systems to be tested at NARA, San Diego Supercomputer Center and at the University of Maryland. More details about the ERA program are available at <http://www.archives.gov/electronic_records_archives/>. Details about projects sponsored by ERA are linked to <http://www.archives.gov/electronic_records_archives/projects/era_sponsored_projects.html"

Several handouts were prepared for conference participants. For additional details about the conference proceedings, or to correct errors in this report, please contact Maria E. Gonzalez at <abbeypub@grandecom.net>.