Recently, I led a test project to digitize a select portion of the John R. Hickman Collection, housed at American University Archives and Special Collections. Donated in 1994, the Hickman Collection consists of more than 10,000 broadcast quality audio recordings of vintage radio news and entertainment programs, from the 1920s through the 1970s. It contains reel-to-reel tapes, metal and vinyl discs, electronic transcripts from studio masters, wax cylinders, and books. The Hickman Collection includes a wide variety of entertainment programs, including comedies, detective shows, dramas, mysteries, quiz shows, soap operas, talk shows, and westerns. The collection also boasts historic radio news broadcasts and commentary from such notable broadcasters as Edward R. Murrow, H.V. Kaltenborn, Robert (Bob) Trout, Mary Margaret McBride, Robert Baukhage, and Fulton Lewis Jr.
I first encountered the collection during the spring of 2005 when, during my part-time work at the American University Archives, I began taking inventory of the reel-to-reel tapes and building databases with metadata about the news programs and some of the classic series, such as Cavalcade of America, Gunsmoke, and The Lone Ranger. At the start of 2006, however, none of the collection had been digitized or even recently played, offering a significant opportunity for a history graduate student like me to work with an untapped collection. For my “Digital History” project, therefore, I sought to combine my archival and historical interests, experience with the Hickman Collection, and personal enjoyment of vintage radio programming to kick-start the effort to digitize reel-to-reel tapes in the collection.
American University is part of the Washington Research Library Consortium (WRLC) and WRLC already boasts more than a dozen digital collections, each of which has its own website, allowing users access to the digital material. Nearly all of these collections, however, are visual collections, including digitized letters and other documents, photographs, and cartoons. There is currently one collection featuring audio interviews, but none with radio programming. My specific objective therefore was to create an online WRLC repository for the Hickman Collection, digitizing ten hours of tape for this test project.
While managing this two-month effort, I glimpsed the future role of the historian in the digital realm. Unlike a basic term paper in which a student independently conducts research and writes her analysis and conclusions, digitizing the Hickman Collection provided a very tangible sense of how historians must work fluidly with other disciplines and professions in order to create relevant and meaningful scholarship using technology. Prior to this experience, I had some technical background, such as working knowledge of computers and familiarity with the Windows environment, and basic skills, like scanning images. The scope of the Hickman Collection project, however, gave me an entirely new understanding of technology and the inter-connections of the historian with IT specialists, audio experts, catalogue librarians, law professors, and archivists.
Phase One: Audio Digitization
After securing permission to proceed from the American University Library Archives and WRLC, the next and, perhaps most important, step was to find a way to play the reels and transfer the data. This involved approaching American University’s Audio Technology Department to secure access to equipment to digitize the reels and find student assistance to help with the effort. Thankfully, the Audio Tech Department was only too happy to assist, as its current students rarely (if ever) have an opportunity to use the analog equipment required for the digital conversion process. One of the Audio Tech professors explained to me that letting his students work with analog technology was akin to allowing an auto enthusiast to peer under the hood of a Model T!
With a designated audio technology student as my constant mentor and guide, we jumped into the project. For the audio digitization effort, we used a Digidesign Digi002 analog to digital converter and a NAGRA reel-to-reel tape player (also used on movie sets in the past) as capture devices plus Pro Tools version 6.9 audio software as the capture software. We captured the audio as 24-bit, 96 kHz uncompressed WAV files, which were subsequently bounced to MP3s at 128 kps using Toast. While we planned to stream the audio online using MP3s, we are keeping the WAV master files on the server for archival posterity. Overall, despite a few hurdles, the audio digitization process was relatively smooth. Since we could not play the reels in the Hickman Collection without this newfound equipment, I found that relying on the information listed on existing labels on the reels was one of the biggest challenges. While this information generally turned out to be valid, a reel labeled “John Kennedy Inauguration Address” that I wanted to digitize inexplicably turned out to be a short clip of bluegrass music instead. In addition, I learned that several of the reels had previously been stored incorrectly, we needed to affix leader tape to each reel in order to not cut off any of the audio while feeding it into the NAGRA machine, and it takes approximately two minutes to digitize every minute of audio.
During this phase of the process, I needed to select which of the hundreds of reel-to-reel tapes in the collection to digitize. After discussion with the head archivist at American University, I chose eleven news programs and nine entertainment programs to provide a balanced reflection of the collection. They range in length from approximately fifteen minutes to more than forty-five minutes, totaling approximately ten hours overall. Mindful of the needs of future researchers who might use the collection, I sought news clips of historic importance with notable commentators from multiple time periods. Since the majority of the most noteworthy news clips in the collection are from World War II, however, my digitization project reflects this constraint including news coverage literally an hour after Pearl Harbor was bombed, early reports from D-Day, commentary from President Roosevelt’s funeral, President Truman’s speech on V-E Day and live reports from the Pacific on V-J Day. We also digitized Marian Anderson’s concert at the Lincoln Memorial in 1939 and episodes from two popular radio shows — Cavalcade of America, which recounted stories of U.S. history, and Biographies in Sound, which related the life stories of famous Americans through clips of voices from their lives – as the Hickman Collection houses a significant number of episodes of each.
Phase Two: Metadata
Although I had been collecting and recording metadata about the reels in the Hickman Collection for nearly one year, I had not planned to digitize the collection and build a web site with WRLC when I started working in the archives. Therefore, once I had to think about presenting metadata to researchers online, my approach changed. For instance, I needed to consult Dublin Core elements to plan the appropriate metadata fields, add subject fields to allow users to search and browse by the names of important events or people referenced in a specific broadcast, and include technical information about each audio clip, both the original analog format and the new digital one.
These additional steps created new challenges. The head archivist recommended only using Library of Congress Authorized Subject Headings for the subject fields, but I learned that this pool of headings is not as comprehensive as I assumed it would be. For example, several of the news programs are from the Mutual Broadcast Network (which is no longer in operation), the NBC Blue Network (today’s NBC) or the NBC Red Network (today’s ABC), but the Library of Congress has no headings for any of these stations. Therefore, I included the network name in the description field, which users may search by key word. My boss and I also disagreed with WRLC about how much of the metadata to show to actual users of the site. For instance, WRLC consulted with a catalogue librarian who insisted we have a field for language although every item in the collection is recorded in English. From the historian’s perspective, I believe that is extraneous information, cluttering the screen and making it more difficult for a researcher to find the information he or she needs on the web page. However, WRLC agreed not to include metadata on the web site that lists all of the technical specifications for the capture process, illustrating the art of compromise in this type of project.
Phase Three: Web Component
Once we finished the audio digitization and built a new Access database to list the metadata, I needed to upload the WAV, MP3 and Access files to WRLC’s server. As I was previously unfamiliar with FTPing and needed to use a Mac in the Audio Tech Department for digitization and a PC in the archives to create the metadata file, I spent about one week educating myself and evaluating different FTP program options for Macs versus PCs. While the library’s IT staff ultimately installed an FTP program on my PC, we downloaded a free trial of Fetch on the Mac. Prior to FTPing the metadata, I converted it from Access to XML, requiring me to repeatedly correct errors in the metadata that occurred during the conversion process.
As WRLC has other digital collections, there already is a customized web publishing system set up that I was able to tap into, learn about, and ultimately work with thanks to the help of WRLC’s very patient IT personnel. Two IT contacts at WRLC moved the metadata and audio files to the Digital Object Catalog or DOC, a DSpace repository to manage and preserve the files, and then imported them to Greenstone in XHTML. However, after this initial configuration, I’ve been able to make all changes, imports and updates in DOC templates, which then are automatically uploaded onto the site. I also drafted all text for the site and created the banner image in Photoshop.
Although WRLC’s digital collections boast a similar look and feel, I insisted on several unique navigation components for the John R. Hickman Digital Collection. This includes dividing the clips into two series, news and entertainment, with entertainment further broken down by program. In addition, the entire site is searchable by keyword and browseable by title, subject, date, person, and series.
Historian as Coordinator
Although I am a graduate student and a part-time employee in the American University Archives, my boss has a great deal of confidence in me and provided permission to lead the Hickman Collection digitization test project. Hence I coordinated and actively participated in all phases of the effort, secured support from other departments at American and WRLC, and provided regular status updates to the head archivist. If there was a problem that needed addressing, I handled the situation, from the most mundane, such as securing a spare key to the audio studio where the digitization took place, to the complicated, including obtaining permission for Mac server space and resolving issues with FTP programs and XML conversion. In addition, during the project, I consulted with a law professor at the Washington College of Law who was already familiar with the Hickman Collection about potential copyright issues. His input was especially helpful in determining whether to allow only the WRLC community or the general public to access the audio clips at this time. Metaphorically, I felt like the conductor of a potentially disparate symphony, always ensuring that I effectively wove together various sections of my little orchestra.
Summing It All Up
Overall, I believe the John R. Hickman Digital Collection test project meets my expectations for the semester. Although I began planning for this project in late February/early March, several steps required more time than I originally anticipated, such as FTPing uncompressed WAV files which actually took longer to upload than the length of the actual audio clips. However, I am overjoyed that we were able to digitize so many reels and create a clean, navigable site that will be of use to historians and radio enthusiasts even as a test initiative. This type of archival online repository spans both academic and public history. Classic radio programming is an important audio resource that may be used by social or cultural historians, but simultaneously appeals to members of the general public with an interest in old-time radio.
Unfortunately, this intensive effort only digitized a minute fraction of the entire John R. Hickman Collection. Under ideal conditions, I would arrange for the entire collection to be digitized, providing researchers and enthusiasts with access to hundreds of vintage radio programs. Perhaps this test project will serve to garner grant funding to a allow for a larger-scale digitization effort.
However, I would even love to be able to just listen to all of the reels in the collection myself, determine if the labels on them are accurate, and then evaluate if the episodes in the collection are particularly unique in the world of radio. For instance, less than sixty of the 1928-29 syndication recordings of the now controversial show Amos ‘n’ Andy are known to survive, and even fewer episodes have surfaced from the show’s 1929-43 network run. According to unconfirmed labels on the reels, the Hickman Collection boasts three reels from Amos ‘n’ Andy’s 1928-29 syndicated season and one, possibly, two reels, from the 1929-1943 period. Similarly, Bing Crosby starred in more than 300 Kraft Music Shows during World War II, but complete recordings of most of the hour-long broadcasts did not survive the War. In 1971 NBC donated its entire collection of radio recordings to the Library of Congress including just six Kraft shows aired prior to 1944. A few NBC affiliates recorded some of the wartime Kraft broadcasts, and their recordings serve as the source of the few early shows that currently circulate. However, the Hickman Collection boasts one original reel believed to include a 1943 episode.
Currently, the John R. Hickman Digital Collection resides on a test server with plans to migrate it to a production server and make it accessible to the WRLC community shortly. As a next step, I plan to engage in copyright research at the Library of Congress this summer to determine if the site may be opened to the general public. In addition, the head archivist and I are going to use Dreamweaver to add pages to the site to simply list metadata about the other news programs we have in the collection, and then perhaps create similar lists of the other Biographies in Sound and Cavalcade of America programs. Based on the results of the copyright research, more select digitization may commence in the fall semester as well.