PACIFIC NORTHWEST LABOR ARCHIVE
Final Report
Grant (Contract) # G-2069

Project Description:
There is a growing interest in our region's rich history, its environment and its trade along the Pacific Rim. The citizenry needs access to primary resources such as are collected in the University of Washington Libraries. The Libraries currently offers to the general public Web-accessible digital reproductions of some primary sources. Manuscripts, photographs and maps of the Pacific Northwest history and heritage are now available. But newspapers-where the micro-filming quality is uneven-present a particular problem. Likewise, oversize print materials such as posters, newspapers and maps are not easily scanned with affordable equipment. Outsourcing costs for high quality scanning and printing of graphics and maps are prohibitive, especially for public institutions. And rare materials, which are non-circulating and difficult to access now, are at risk when they must be removed from the building in order to scan.

In order to provide the public with easy access to accurate digital reproductions and to low-cost, but high quality, print copies of primary source materials, the University of Washington will test feasibility of in-house digital photography and printing for oversize materials. This project will also test recommended benchmarks for digital file size transmission and display across various access speeds and platforms. Legibility and accuracy of content display for remote users must be balanced with network bandwidth and common platform capacity to insure that public libraries and schools have access to the wealth of materials held in the State's libraries.

Our project primarily tested the feasibility of using in-house digital photography to digitize oversize materials, in this case, bound newspapers. The camera used was the Olympus C2000z. Although in the end we were able to create a viable result, we would not recommend use of low- or moderate-end equipment for this purpose. While this camera is capable of creating useful images from large items, such as posters, where there is generally less detail, it was incapable of taking photographs of large sections of the newsprint that were still readable. As a result, it was necessary to take photographs of sections of columns, no more than 10 inches in length (preferably 5-6 inches long), with the camera no more than 20 inches over the newspaper, and splicing these photographs together with Photoshop. To add further complication, even with photographing such small segments of the whole, the camera lens still added a distorted 'fish-eye' effect. Although this was able to be somewhat compensated for in Photoshop, and did in the end create a usable product, it was extremely time-consuming. If the means are available, we would recommend either a higher-end camera, or an oversized-materials scanner. We do, however, still feel that digital photography can be done in-house, providing the equipment and the materials selected suit each other well.

Another task of our project was to test the feasibility of meeting existing benchmarks for digital file size transmission and display. To test this we did 'field trips' by sampling our target audience (public libraries and a variety of schools). We accessed our local collections remotely from the public libraries and schools and recorded very specific data on the types of lines and transmission speed they had. We were pleased (and surprised) to learn that, across the board, our target audience has exceptionally good access to the internet, and as a result file size is not a major concern when tailoring a project to this audience. We did, however, try to keep our jpeg file sizes near the classic benchmark numbers (200K), as this did help speed the transmission of data to personal computers. Since file size was not the overwhelming concern we expected it to be, we were able to retain the color in our images, which we feel created a better final product.

The University Libraries already has a number of digital collections, with a number of skilled selectors, so the process of selecting materials to offer as much detail as possible went very smoothly. Experience pays off. Similarly, our staff has already scanned and cataloged these collections, so it was not difficult to create an efficient workflow for the processing of the materials selected. Nonetheless, the addition of new equipment and new types of material to be processed did add to our knowledge base. On a basic level, we learned that the lights needed for digital photography break easily and are extremely hot. Therefore, the equipment really needs to be in a space with good ventilation. Further, we would recommend sunscreen and sunglasses for staff, and making sure that long hair is safe from touching the lights. Be very aware of shadows, be they from equipment wires, or arms. For optimum use, the camera set-up and the computer used for processing need to be in close proximity. The use of a camera mount, to keep the camera steady, is recommended. Unfortunately, the newspapers we were using had been bound. If possible, we strongly recommend avoiding such large bound objects, as they were extremely awkward to handle, and if using the camera mount, there were areas of text that were unable to be photographed in this fashion. These 'blind spots' required hand-held camera work, which is a more cumbersome process. If bound newspaper volumes must be used, we recommend the use of overhead scanners, or possibly even outsourcing the digitization task.

Until this project, we had never had the opportunity to scan hand- and typewritten manuscripts, which we feared might be a difficult task. Luckily, we were wrong, and found out that with minimal manipulation in Photoshop these images turned out quite well. This knowledge will prove useful to future projects.