by Sam Springman, Unit Manager, Media Digitization and Preservation Initiative, Indiana University
Until May 2016, MDPI sorted open reel audio tapes into groups for digitization based almost entirely on a straightforward physical inspection: looking at the tapes themselves, reading the technical information written or printed on their boxes, and so forth. That approach worked well as long as our preservation work was limited to relatively consistent and well-documented materials, but once we began digging deeper into Indiana University’s holdings, we found that the information we needed to sort tapes by technical characteristics was either missing or—worse—shown incorrectly. No matter how intently we stared at a tape, held it up to the light, measured it with calipers, or compared it to previously digitized open reel from similar collections, we were not improving our odds of a tape being successfully digitized. It was clear, as described in Part One, that a different solution would have to be found.
It was decided that the only way for us to get this critical information would be to actually play the tapes on a machine and listen to the recording. We were given two tools in order to accomplish this task. The first was a Studer Model A80 ¼” open reel audio playback machine with a half track tape head.
The second was an Arnold B-1022 Magnetic Viewer. For the uninitiated, a magnetic viewer is essentially a liquid and iron filled small flat object, metallic on the underside, and clear on top so that the iron can be seen. The viewer is placed over the recorded side of the open reel, and the iron reacts to the magnetic tape and aligns with the signal so that it can actually be seen visually!
A select group of experienced members of the SMARTeam were trained on the use of the playback machine and given the responsibility of diagnosing and sorting open reel with unknown or contradictory technical specifications. Below is a brief rundown of the things required to diagnose each tape, and some of the more… eccentric ways recordists chopped up their track configurations while recording.
Each tape was wound through the machine and played. The signal was listened to through a headphone mixer as it has no internal or external speakers. As a tape first begins playing, the operator is only looking and listening for a strong signal. If the audio signal is faint, the magnetic viewer will be unable to read the signal. Once a strong signal is found, the operator then begins to listen for certain things that will help them to identify the tape. The first thing the diagnosis operator is listening for is the speed of the recording. As you can see:
The machine has only two speed settings, 7.5 inches per second, and 15 ips. What happens if the tape was recorded at a different speed? Well, a 30 ips tape played at 15 ips will sound too slow, and a 3.75 ips tape played at 7.5 ips will sound too fast. Unfortunately, it can be hard to tell a 3.75 ips tape from a 1.875 ips tape in this way, but we had occasional access to another machine with a 3.75 ips setting for resolving the difference between those speeds.
We’ve discovered tapes with extremely unique ways of arranging the configuration or other outliers. For example:
- A quarter track mono tape that is recorded fully, but every side is played in the same direction. Each side has discrete content.
- A quarter track tape that contains two forward mono recordings, and one forward stereo recording.
- A 30 second tape with a recording of a political jingle almost too short to thread in the playback machine.
- A tape with multiple problem stocks spliced together. In this case, a sticky shed polyester tape (sticks to itself and the tape heads, must be baked at a low temperature to alleviate) spliced to an acetate tape degrading from vinegar syndrome (a process in which acetic acid is released as the plastic tape base degrades, and in which the acetic acid acts as a catalyst, further degrading and breaking down the plastic). That one went to the IU Media Digitization Studios!
- Audiocassette tape (⅛” wide) wrapped around empty 1/4” reels.
And these are just to name a few!
The main problem we faced with this workflow was time. For a more difficult tape it might take as long as 15 minutes to diagnose completely, and even if the SMARTeam was lucky enough to hit a run of uniform tapes, there was still only one machine, and one possible operator at a time trying to keep up with as many as 650 tapes per week. Eventually we were able to hire experienced Memnon staff to augment our own workforce in order to maintain this pace, though it was later determined that to save time and effort it would be easier to incorporate the cost of diagnosis as part of the digitization package. The changes made to the workflow brought on by the shift to Memnon diagnosis will be the subject of Part Three.