| Library Home |
| Digital Library |
| Services |
| Conditions |
| Equipment |
| Resources |
A Digital Imaging Planning PrimerThe creation of digital imaging databases built on the scanning of primary source materials for the purposes of access to, and preservation of, these materials is a promising and exciting development in the library and archives field. That said, such projects are not necessarily simple to do and require a large amount of preparation, planning and resources. By assessing the level of access already in place and the physical condition of the materials being considered, collection managers can make the decision as to whether to proceed with a digital imaging project. This page is designed to help the curators and collections manager assess their collection to this end. It does not cover in detail issues that involve the implementation of such a system, rather it is hoped that it will provide the essential elements of deciding to take the leap. Planning and implementation requires the cooperation and coordination of several divisions, including cataloging, preservation and systems. The Library of Congress' National Digital Library Program home page has links to several technical papers that detail the digitization process: http://lcweb2.loc.gov/ammem/ftpfiles.html Planning for a Digital Imaging ProjectWhy digitize research materials?
Choosing a Collection: Access vs. PreservationDeciding what collections are good candidates for digital imaging should be based on two principle criteria:
Public and scholarly demand for content should be a major consideration when choosing a collection. Because of this demand, such collections are often already cataloged and indexed. If applicable, cataloging of the considered collection should be complete before a digital imaging project proceeds. As we all know, cataloging is labor intensive and the thus most expensive part of a digital imaging project. While not always economically feasible, it should be recognized that item level cataloging will make for a more robust, searchable database. The cataloging data should be exportable to ASCII delimited and/or MARC formats, which make imaging database construction much easier. Additionally, XML (Extensible Markup Language) will greatly enhance access points to the collection so this has to be considered as well. The delivery of reproduction grade prints from high-resolution files is now feasible and several institutions now deliver them in lieu of traditional silver based copy prints for such purposes. The Fujix Pictrography system offers excellent output and are said to be much more permanent than the dye-sublimation process, since they use photographic paper in their process. The preservation of quickly deteriorating or fragile materials may take precedent over more popular collections that are in stable condition. Some may require immediate attention since the deteriorating items may not be available at a later date to digitize. Such collections include nitrate and acetate negatives, glass plate negatives, color prints and transparencies, acidic scrapbooks, maps, prints and illustrations. In most cases, conservation treatment needs to be completed before scanning of the objects begin. In some instances, less costly treatments may be done after scanning. Preservation of materials after scanning should be as complete as possible since access to original materials will be limited once a digital surrogate is put into place. Criteria for allowing access to original materials by scholars and the public should be developed by the collection manager at this time. The Digital Lab in the AMNH Library will do some digital sampling of collections for the purposes of determining the viability of a planned digital imaging project. The Digital Lab will also assist in determining specifications for potential vendors to use as a guide to submitting their digital samples. The various physical and intellectual characteristics of your collections will greatly effect the outcome and size of the digital imaging project. Medium: Visual materials, such as photographs, line art, three dimensional objects, illustrations, audio and video, often take up a large amount of memory storage. Uncompressed TIFF files are used for archival files and usually average from 17-250 MB or more per image. Textual materials may require OCR (optical character recognition) and/or XML markup of original manuscripts, text/illus., microfilm, and finding aids. For photographic collections, if original negatives and/or transparencies exist, they should be scanned instead of prints for they hold far more information. (Caveat: In the rare cases where negatives for fine art photographs exists, one must consider the aesthetic considerations of the photographers original print.) Format: Like medium, the size of the original also determines image file size. Observe the item dimensions, and their level of detail. Assess the variety of dimensions and number of items that comprise the collection. Scanning of intermediates such as copy negatives and slides and transparencies need specialized equipment to adequately capture detail for reproduction purposes. While scanning original negatives and transparencies provide the greatest detail, copy intermediates often have a much lower dynamic range, which results in reduced sharpness and fidelity to the original. Oversized items such as maps and architectural drawings may require taking sectional scans of the items and then "tiling" them back together again in the computer. This will result in very large image files. Bound items can be difficult to scan if the book does not open at least 90 degrees. Quality: The importance of the detail in your collections effects the amount of resolution you will need to render that detail. Likewise, the fidelity of colors in items may also be essential. Color replication requires precise calibration of scanners, monitors and printers for best results. Finely detailed items like maps, engravings, musical scores or manuscripts may require very high resolution. Detail views also may be needed for close examination of the objects. Items that do not require fine detail or color fidelity, such as posters or line art, may not require high resolutions. Cost: Budgetary assessment and potential grant funding should be thoroughly explored. Costs can quickly accelerate if cataloging and/or preservation or conservation treatments are required before scanning. Many grant agencies favor collaborative projects with other institutions for digital projects. Shared Resources and Cooperative Projects: Many institutions have digital imaging projects under way. Care should be taken not to duplicate the efforts of other institutions. Several consortiums have been created to promote cooperative projects that share resources and are designed to create virtual collection on the internet. The Library of Congress, The Research Libraries Group, The Getty Center for Art Information, the Commission on Preservation Access, and the Digital Library Federation are active in this area. Integration: A successful digital imaging project requires close communication and cooperation between the curator or collection manager of the respective collection, and the preservation, cataloging, and systems divisions. The nature of the collections scanned, and the desires and needs of the staff and public, must be effectively communicated to those constructing the database. Whether it is a vendor or in-house personnel, those constructing the imaging system need feedback to make the system both dependable and user friendly. Likewise, collection managers need to know about how an image database is constructed and what specifications and procedures must be followed to ensure a successful implementation. The preservation and maintenance of an image database requires careful planning. Universal Resource Names http://lcweb2.loc.gov/ammem/award/docs/handle-server.html need to be assigned to collections for use on the internet. Derivative image files from digitized collections can be used for preservation treatment planning, condition reports, collection control, rights and permissions and internal public relations and publications. Because digital collections can be integrated into a system-wide database, it should be designed to integrate with the network, not the collection or division at hand. Copyright 2002 Anthony Troncale |
