That said, we share Unsworth’s concern about the scholarly bottlenecks and we have taken several steps to address them. First, we have taken important lessons from each of the summer institutes we have held and have improved our ability to help scholars complete their work. Looking back at our first summer institute, held in 2004, less than a year after programmers had begun work, the annotation software barely functioned, we didn’t have any annotation models for scholars to follow, and we devoted a lot of time to discussing what we wanted the end result to look like. Many of the collections accepted for that first summer institute do not yet have complete annotations and remain unpublished. By the time of our last summer institute in 2009—nine months ago—we have dramatically increased the amount of work that gets accomplished and half of the participants have already submitted their work for peer review. Second, we have been reevaluating the path we have chosen and are considering a model in which annotation is more basic and more focused on description, allowing access to move more quickly. Deeper description and cultural analysis would form another layer, manifested in work in various kinds of online journals. In this scenario, the work of more narrative writing falls into disciplinary domains and peer review apparatus outside of the EVIA Project.
Unsworth’s description of an alternative preservation and access process is useful for the sake of argument, but is simplistic in its outline and ignores key aspects of the kind of content we address. Ironically, the initial process he describes is not so different from what we do. We send out a call for depositors and select those that are appropriate. Those applications gather very general collection-level information. Unsworth does not indicate a selection process, but I think he would agree that the costs of digital preservation warrant some selectivity. Once we select collections for accession, we collect information necessary for preservation production—basic information about each recording that generates numbers, labels, and information we use for tracking the item. As for digital preservation transfers, Unsworth describes a situation of the future that does not presently exist. While depositors in our case are using largely consumer-level formats such as VHS, Hi-8, and MiniDV, we cannot rely on standard consumer transfer services for digital preservation. It remains a challenge to find vendors who understand digital preservation rather than simply copying tapes to DVD for the family media album. Our present digital preservation format for analog sources is uncompressed video which creates a 125 GB file for every hour of video. While no internationally accepted standards and best practices for video preservation exist, we must follow the models from audio preservation and the best intentions of video preservation work happening at other institutions if we want to make a sincere effort in this direction. Thus, we must have technical metadata collected about the transfer process and run quality control processes on the completed file. Our project relied on the University of Michigan for most of our transfers and commercial vendors in more difficult cases. The costs are significant in either case and so selection is necessary. In addition, as these formats age, the transfer process becomes more challenging. Even MiniDV has proven to be a vexing format from which to get good, error-free transfers.