<<< Photograph Database

One of the challenges of working in the archival profession is the apparent contradiction between two fundamental responsibilities: to preserve records of long-term value, and to make those records accessible to users. How are these positions contradictory? Consider this: providing access to records implies that they will be used, and such use normally involves physical handling of those records by both staff and patrons. In turn, even careful handling will result in wear-and-tear on the records, which will work against any conservation efforts.

Archivists endeavour to establish and follow procedures which strike the best possible balance between accessibility and preservation, but we are always looking for ways to improve that balance. The University of British Columbia Archives' Historical Photograph Database project is intended to take advantage of advances in digitization technology, in order to address the challenge of balancing archives accessibility and archives preservation.

The photographic holdings of UBC Library's Special Collections and University Archives Division are divided between the University Archives Photograph Collection, which documents UBC events, people, and places, and the B.C. Historical Photograph Collection, which documents the rest of the province. Together, they contain approximately 200,000 images, in the form of prints, negatives, and slides, and document the history of the University and of the province. It is a valuable historical resource for University faculty, staff, and students, as well as outside researchers.

As in many archives, the usual method for accessing archival photographs at UBC is rather labour-intensive, and also potentially damaging to the images. Users rely on brief descriptions of the images, either in our card catalogue or in one of our inventories, to choose which ones they want to look at. This can be difficult if several images are listed showing the same event, person, or location, with each photograph perhaps taken from a different angle or at a different point in time. In such cases the researcher will request many more images than he or she will ultimately need in order to find "the best one". They are then located and retrieved by staff. The researcher then has to physically remove each photograph from its envelope to view it. Afterwards, it is returned to its envelope and refiled.

This process is time-consuming for both staff and patrons, and can lead to unnecessary wear-and-tear or even damage to the images themselves.

In 1995, the University Archives proposed a project to scan photographic images, store them digitally on a computer database, link the digital images to short bibliographic records for easy retrieval, and make them accessible to patrons by computer. Using such a database, archives staff and patrons could consult the images on a computer screen without having to physically handle them, and determine which ones suit their needs before requesting the originals or ordering copies. Not only would this reduce the risk of damage to those photographs, it would also save time for both the researchers and our staff.

The University Archivist had long been interested in digitizing our photographic images, and had already consulted with Library Systems staff to determine the feasibility of such a project. There was initially some question about what medium to use, with the UBC Library on-line catalogue and CD-ROM both being proposed as options for storing the database. Another possibility would have been to make the images available to Internet users via FTP (File Transfer Protocol).

Another option presented itself in June 1995, when the Special Collections and University Archives home page made its debut on the World Wide Web. The rapidly-growing popularity of the Web as a means of accessing Internet resources, and the positive responses to our own Web site, made it an obvious choice for distributing our photographic images to researchers. The British Columbia Archives and Records Service's Visual Records Database also served as an example of what can be done with the new technology.

Using the Web also enhances the accessibility of our photographs by permitting remote access by users with Internet accounts. They no longer need to actually visit the Archives, but can view the photographs from their offices, or the comfort of their own homes - even if they are thousands of miles away. They can even ask for a high-quality digitized image, which we can re-scan from the original and provide for them as either an FTP file or an e-mail attachment.

In 1995 the University Archives obtained a Canadian Council of Archives preservation grant for a pilot project to scan approximately 1250 historical images from our collection. The grant paid for a flat-bed scanner (with software, running on an IBM PC); time required to design the database itself; and time required to scan and upload files.

The pilot project began in September 1995. Among the first photographs to be uploaded were a series of faculty portraits dating from the 1960's and 1970's; various building and construction photographs; and the Frank Buck Collection, documenting the work of the UBC campus' first landscape architect. The original goal of scanning 1250 images was soon reached, and the project was expanded and continued through February 1996, by which time it was obvious that the database was feasible and useful, and could be developed further.

The various tasks involved in the building of the database are distributed among several University Archives staff. One person enters bibliographic data for each image. Another does the majority of the photograph scanning and editing of the digital images, and uploads the images to the Internet server. Several part-time student assistants are also employed to do much of the scanning. On average, each image takes approximately five minutes to process.

Each image is scanned, saved, uploaded, and stored as a TIFF file. When a search is carried out on the database, the search result produces a bibliographic description along with a small "thumbnail" GIF image. The conversion from TIFF to GIF format is done automatically by the database program, and the GIF image exists only while it is being viewed. The user can view a full-sized GIF version of the image by clicking on the thumbnail. The user also has the option of downloading a copy of the original TIFF file (the "raw image") to his/her own terminal, and viewing it using an external graphics program.

It should be emphasized that, in developing our photograph database, it is not our intention to preserve photographic images by replacing the originals with electronic files. We still believe that there is no substitute for the original image, and that all steps with regard to storage facilities and environmental controls must be taken to safeguard it from harm. What the database does is help preserve the original image by reducing the physical handling involved in accessing it by making the image accessible in a non-physical (dare I say, "virtual") form.

One must also keep in mind that, if we did want to replace the original photographs with digital images, we would have to scan them at a sufficiently high resolution to record all the visual data included in the originals - a resolution as high as 600 dots per inch. The resulting image files would be too large to be practical for our purposes - they would be too big to store more than a handful on our Internet server, and too big for users to access without waiting several minutes for images to appear on their computer screens.

For our project, the scanning is done at the relatively low resolution of 75 d.p.i., which is not sufficient to use as a means of preservation. But the files are small enough to store on our server, and for users to view easily. At the same time, on-screen image resolution is adequate for research purposes; when they are printed out on a laser or bubble-jet printer, the image quality is comparable to that of a very good photocopy.

Since the database's public debut in January 1996, the reaction from the archival and research communities has been overwhelmingly positive. Archivists at several other institutions have asked for technical and other background information to use in designing their own databases. University Archives staff have given demonstrations to faculty, students, and other researchers, and the database has been used several times to answer reference questions. Nevertheless, some concerns have been raised, which should be addressed.

Some users have found the search engine interface difficult to use. We are looking for ways to improve it, although until now the priority has been to develop the search engine itself. We are regularly discussing with the database designer how to make the interface as user- friendly as possible.

Another search-related problem that has been pointed out is that a search sometimes produces a bibliographic entry with no image attached. This is because the bibliographic information is entered into the database before the image itself is uploaded, and there is a time- lag between the two steps. This will become less noticeable as the project progresses, since as the database grows the number of bibliographic entries without corresponding images will become proportionately smaller.

The other major concern that has been voiced is the possibility of losing custodial control over our photographic holdings, and the potential increase in copyright violations by users. While this is an issue worth noting, it is our belief that in this case increased accessibility does not lead to decreased control.

All of our rules and policies regarding copyright and assurance of appropriate use by patrons apply to the on-line images as much as to the originals, as is made clear throughout our Web site. In addition, keep in mind that the resolution of the scanned images is relatively low - low enough, in fact, to make on-line images unsuitable for publication. If this eventually proves inadequate to protect against unauthorized duplication, another option would be to superimpose text or "watermarks" over the images - they would not completely obscure the images, but they would be enough to make them unusable.

Building the database has not eliminated staff as intermediaries - patrons must still contact us to order prints made from the original photographs. In fact, it is less likely that a person could steal a publishable image from the on-line database than that same person could walk through our office area and into our vault to steal the original print or negative.

Digitization has allowed us to strike a new balance between preservation and access. It enhances the retrieval of information from images by making them remotely accessible over the Internet, while supplementing existing conservation procedures by reducing physical contact with the original photographs. Also, as I have pointed out, the database has already proved to be a valuable research and reference tool. When one considers that of the approximately 200,000 images in our photograph collection only about 8500, or less than 5%, have been scanned and uploaded to date, it is obvious that developing the database is a long-term project. However, it already has had a positive impact on our operations; as it grows, it can only become more useful to both our staff and our patrons.

-- Erwin Wodarczak

The author would like to acknowledge three people who have been most intimately involved in this particular project: University Archivist Chris Hives, UBC Library Systems Programmer/Analyst John Campbell, and Archives Assistant Leslie Field.

Originally published in the AABC Newsletter, Summer 1996 (Vol. 6 No. 4). This is a slightly revised version of a paper delivered at the 1996 AABC conference in Penticton, B.C.