Setting High Image Quality Guidelines

Unforeseen Costs of Imaging for Short Term Needs

Unfortunately, many digitization programs are built around short-term needs. Items are often scanned on a flatbed or photographed with general-purpose cameras (e.g. Canon/Nikon) and planetary devices without any verification of the resulting image quality beyond a cursory visual comparison. Worse yet, the quality claims of a hardware manufacturer are often taken at face value, and are not assessed by the institution. The institution justifies using these lower-quality solutions because they meet the requirements of a particular immediate use; they are “good enough.” This philosophy has two main sources of unforeseen cost: the myriad costs of inevitable re-imaging and the cost of maintaining the Preservation Digital Object (PDO).

Cost of Re-Imaging: Cost, Time, Condition Degradation

If, at any point after creation, the image quality of a PDO is found to be insufficient for a new initiative, the object must be re-imaged. This  requires that most of the Digitization Chain must be repeated as well: object retrieval, object prep, digitization, QC, object return. An even greater concern is the inherent risk of degradation of the object’s condition from  additional handling. The additional internal resources required of the institution and the longer period of time during which the object is unavailable to other interested parties (patrons, researchers, conservation staff, etc.) must also be considered. Moreover, it’s likely both versions (the non-preservation first-pass and the PDO from the preservation grade re-imaging) will be maintained, increasing the cost of maintaining the digital collection. When examining the costs of re-imaging, it generally makes sense to “do it once, and do it right.” It’s safer for the object, it’s more cost and time efficient, and it’s a simpler approach.

“There is so much to image at most libraries and museums that the chance to re-image an object is small, and the opportunity cost to do so is high. Therefore it is only logical to ensure that all digitization is done appropriately.”

– Ian Bogus, MacDonald Curator of Preservation, University of Pennsylvania Libraries

Cost of Maintaining a Preservation Digital Object

A PDO is meant to be kept indefinitely. However, the cost of maintaining and migrating the PDO Substrate [see: 2.1.2 – The PDO Substrate] is high, especially when you sum this cost up over decades or longer. Maintaining a non-preservation digital asset with limited applicability (e.g., only good enough for patron web access, but not adequate for research needs) costs the same as a PDO created for broad scope-of-use both now and in the future. When considering all of these factors, it can be said that the only type of digital asset worth creating (and indefinitely maintaining) is a PDO that achieves the image quality standards that were formulated specifically for preservation-grade cultural heritage digitization.

“There is no advantage in taking the short-sighted view of a “once and done” digital imaging project – it is a waste of time and assets. It is imperative that all stakeholders, especially those who are further removed from the day-to-day work of the project, but perhaps have more authority over allocation of funds, are in agreement and fully understand that a long-term commitment is necessary for the all-important and never-ending responsibility of file storage, organization and access. In comparison, the actual digital capture of the object is the “easy” part of the project!”

– Barbara Katus, Manager of Imaging Services, Pennsylvania Academy of the Fine Arts

The Minimal Cost of Imaging in the Full Digitization Chain

Imaging collections is a Herculean task in terms of labor, costs, and mindshare. Many departments are involved in the Digitization Chain. This includes the executive arm of the institute, preservation conservation teams, cataloguing personal, imaging specialists, and information technologies departments. The scope is so wide and the chain so long that the overwhelming majority of the time and resources spent in the Digitization Chain are not related to the actual imaging. The preponderance of the time and costs are spent on administrative planning, internal communication and project organization, object retrieval, metadata entry, quality control, object return, and file management. For example, a single technician could image more than 10,000 photographic prints in one shift using a DT RG3040 Reprographic System provided that was the only step in the digitization process. In reality, the imaging technician is only one part of the total Digitization Chain, and will spend more time retrieving, organizing, and returning the boxes/containers of prints than actively imaging.

It’s also important to consider that beyond the initial Digitization Chain, there is a commitment to maintain the image file in perpetuity. Even with a decreasing cost of storage, there is still a recurring cost associated with storage and providing organization of, and access to, digital archives. Taking into account the entire Digitization Chain and the life span of the resulting assets, the difference in cost to create and maintain a true Preservation Digital Object versus a mediocre image is negligible.

Imagine opening a restaurant in the middle of Manhattan: designing an elegant menu with a well-researched flavor palate, filling the dining area with the finest decor, and spending the going rate for top-notch staff. Now, imagine installing Easy Bake Ovens in the kitchens of this restaurant. The core product, the meals served, will now be mediocre, despite the rest of the efforts to create a high-end restaurant. While an industrial-grade oven may be significantly more expensive than an Easy Bake Oven, it represents only a fraction of a cost of running the restaurant, and is worth the investment

“Thinking about the value of the [Keith Albee] collection and the value of the staff provided by the [NEH] grant… we’re investing a lot of time money and effort and we didn’t want the imaging hardware to hold us back. We invested in Digital Transitions solutions in order to get a great end product.”

– Bethany Davis, Digital Processing Coordinator Librarian, University of Iowa Libraries

Keeping in mind the cost of the entire Digitization Chain, it is imperative to execute the imaging itself at the best possible quality, which requires the best imaging hardware and software. Otherwise, the rest of the institutional resources involved in the Digitization Chain will have been squandered.

“The percentage of overall cost of digitization at NYPL represented by capture and processing probably falls somewhere between 20 and 30%. This is an extremely rough estimate based on an estimated average salary and factoring in an estimate of average annual equipment cost.”

– Eric Shows, Assistant Manager, Digital Imaging Unit, New York Public Library

“High quality images do not have to cost much more if the equipment is already available. Much of the real costs are human, so the time required to retrieve an item, prepare it, shoot it, process the images and reshelve it can be up to 80-90% of all costs incurred. Even when investing in high-end hardware and software these costs make up only 10-20% of all costs incurred as long as the equipment is consistently used.”

– Ian Bogus, MacDonald Curator of Preservation, University of Pennsylvania Libraries

No Need to Compromise in the First Place

In years past, slow scanning systems and multi-shot systems meant that high-quality images took significantly more time to produce. Increasing resolution and quality with these legacy systems required a direct compromise on capture speed. As a result, a slow workflow was, for many years, the only viable way to accomplish high-resolution, high color-fidelity, sharp PDOs. However, in the last several years, the advent of high-resolution single-shot capture systems has revolutionized digital capture; now it takes the same time to digitize with preservation-grade image quality standards as it does to produce lower quality images that are only suitable for access and patrons.

For instance, a DT BC100 Book Capture System can digitize bound material up to an A2 page spread at 600ppi with FADGI 4-Star sharpness; its capture rate of approximately 24 pages per minute is the same whether capturing an A2@600ppi@FADGI-4 or an A2@150ppi@FADGI-2. There is no longer a need to compromise between high-quality and high-productivity.

“The  SRLF Preservation Imaging team is committed to providing quality digital images of collections held at the UCLA Library. We work with a range of materials including rare manuscripts, photographs, maps, artifacts, and other sensitive materials. The ability to achieve high resolution capture and an efficient throughput with the Phase 645DF and Digital Transitions’ RCam has allowed us to meet the mission of the University while consistently providing access to collections and continuing their preservation as digital images.”

– Colleen A Carlton, Director, Southern Regional Library Facility, UCLA