Recently the MHL completed its first progress report for the National Endowment for the Humanities on work done during the first year of our serials digitization project. We’re going to be highlighting some of our finds here during the next couple of weeks.
Lets continue our highlights by looking at some of the things we’ve learned from our work:
Gaps in runs: Gaps present major obstacles to the work of researchers. We did some informal completion and exchange of claimed titles during our selection on the basis of holdings problems encountered, but no actual loaning of volumes has been attempted. Completing runs by sharing responsibility for digitization has proved more practical.
Quality expectations and assurance: At the outset of the Countway scanning, approximately 20% of the volumes returned were assessed. This number turned out to be needlessly high, as very few items merited rework orders. One approach might be to examine 20% or higher for the first shipments, until a baseline level of acceptable quality can be recognized. After that point, moving to 5% would likely be acceptable, with more thorough checks coming periodically to assure that the overall quality of scanning does not slip. The College had a similar experience and has moved to a 5% quality assurance level.
Pre-shipment agreement with the digitization service provider regarding acceptable types of deviations in image quality are key. With Internet Archive (IA,) there are wide variations in hue between rectos and versos of single books, as well as variations in contrast, brightness, and focus from one page to the next (though these deviations are usually uniform within a book scanned on the same scanning station). For many projects, these variations would be deemed unacceptable, and not in line with the benchmark for faithful reproduction of monographs and serials. For the MHL, however, the benefit in cost reduction was deemed sufficient to merit the less-than facsimile-grade (though more than adequate) variations in quality mentioned above.
When setting up, it is important to consider the type of image being scanned. Originals can generally be described as one of three types: Line Art, Greyscale (i.e. Black & White) or Colour. Making the right choice at this stage can make a huge difference to the quality of the image, the size of the file and whether the output device will be able to handle the file.