CS4624 Spring 1998 Project Idea - Multimedia Preservation (PRSRV)

A major problem in ensuring the longevity of items in digital libraries is posed not by potential obsolescence of hardware, but by the obsolescence of the format in which the data is stored. Multimedia, in particular, requires an enormous amount of supporting software (both operating system and application software) to effect display. If some, or all, of the prerequisite software is unavailable, the multimedia data object will effectively be "lost" to subsequent generations of users, since they have no means of viewing its contents.

This problem is particularly relevant to the Networked Digital Library of Theses and Dissertations (NDLTD). Theses and dissertations are, by their very nature, long term archival documents, and, in their electronic form, encourage the use of multimedia. Unfortunately, most students submitting electronic theses and dissertations (ETDs) focus on the short-term task of creating and submitting an ETD so they can graduate, and, usually, are unaware of the long-term archival considerations: what is pragmatic, rather than what is best, is the rule of the day. Often, this is compounded by a justifiable lack of knowledge on the issues involved, and the availability of suitable tools to create multimedia in a desirable format.

The NDLTD is currently flexible in accepting multimedia data. However, we wish to migrate to accepting only a small subset of multimedia formats: ones that adhere to accepted international standards, provide high quality archival encoding, and have wide cross-platform support. MPEG and JPEG are two standards meeting these criteria. We wish to encourage students to create multimedia only in acceptable standard formats, rather than whatever format happens to be convenient on the machine they use to author their ETD.

The proposed project would involve facilitating the creation of multimedia objects in the "preferred" formats, and also conversion of existing multimedia to these formats. Conversion should be as lossless as possible. The project would, thus, require the identification of freely-available tools for a range of platforms (MacOS, Windows 95 and NT, Unix, etc.), both in terms of tools allowing the creation of multimedia and the conversion from more proprietary formats to those less so. Tools should be located and catalogued for handling still images, audio, and digital video. Accompanying these tools should be instructions and caveats for their effective use, i.e. suggested settings for minimising loss of subjective fidelity.

(Writeup by Paul Mather, paul@csgrad.cs.vt.edu, the contact person for this effort, along with staff in the Library.)