by Dan Figurelli, Audio Preservation Engineer, Media Digitization and Preservation Initiative, Indiana University; Melissa Widzinski, Audio Preservation Engineer, Media Digitization and Preservation Initiative, Indiana University; and Jonathan Richardson, Audio Visual Specialist, Media Digitization and Preservation Initiative, Indiana University
Edited by Adam Nickel, Processing and Quality Control Specialist, Media Digitization and Preservation Initiative, Indiana University
The content of this post originally appeared as a presentation given at the 2017 Association for Recorded Sound Collections Conference.
In our time working at MDPI, we have come across a few wild things: unique media objects that presented interesting challenges. For the first post in this series, we will discuss some of the wild open-reel tapes we have found.
Over the course of preparing thousands of open-reel tapes for digitization, it is not unusual to come across a sight like this on any given day:
The tape on the left is an unfortunate sight to see, but it is a reality when artifacts have not been properly cared for or stored. The tape on the right is a similar problem that will probably take the same amount of time to fix. In both of these instances, the tape was rewound onto new spools and digitized without serious issues. Not so “wild,” but a little out of the ordinary of most of the tapes we see.
Here is a real wild thing: Scotch 206 shedding. Scotch 206 is not on the sticky shed brand list, but this shedding or stripping still happens. Not all Scotch 206 tapes display this behavior, and luckily this is the only case we have come across. The large amount of shedding, or stripping as it could be called, happened in just a few seconds of play. As far as we know there is no surefire cure for this syndrome, so the tape has been set aside until we can come up with a way to safely digitize it.
When digitizing open-reel tapes, it is not unusual to see different types of speed changes such as this:
When creating a preservation master file, we capture the speed change twice. When the recording changes from 7.5 to 3.75 ips, we continue digitizing for an additional ten seconds. We pause the transfer and adjust the playback machine to 3.75. We leave a five-second gap of silence and then capture ten seconds of content leading up to the change to 3.75. Once that change occurs, we are now at the correct speed and may continue on with the rest of the tape.
When we create the Production Master file, the goal is to make the content as listenable as possible. We edit out the incorrect speeds, so the content is seamless and the end user has a pleasant listening experience:
Many field recordings we deal with have at least a few of these speed changes. But then there are some that have way too many.
This tape changed speeds 22 times! It was a field recording from 1961, where a woman was singing old songs that her mother taught her. She would sing, then pause to remember what the next line of the song was. Each time she paused, the recording changed from 7.5 ips down to 3.75 ips. When she remembered the next lyric and started singing again, it went back to 7.5! The person doing the recording was no doubt hoping for the best possible fidelity for the songs, and probably trying to conserve tape out in the field; however, she showed very little consideration for the preservation engineer some 60 years later.
Here we have a spiritual group from the 70’s being interviewed about their career. Thankfully, they sang at the beginning of the interview, because the rest of the tape suffered from dying batteries and random speed changes. In this example, you hear the content speed up, then change speeds to 3.75. After we corrected for the change, it still didn’t sound quite right….
When we come across things like this, we digitize at a fixed speed so that proper time correction can be applied later on.
Another issue with open-reel tapes is changes in track configuration. One wild example was a tape that started as full-track for about 40 minutes at 7.5 but then changed to quarter-track at 3.75. It changed back to full-track… then to half-track mono… then to full-track again! It went back and forth like this seven more times with plenty of stops, starts, different tape stocks spliced together, and wow and flutter issues.
Here we have an example of a tape that was labeled as full-track, but didn’t sound very good…
It sounds very dull, almost like the azimuth was out of alignment. The magnetic viewer actually revealed that the content was a very narrow band near the center of the tape. We got much better results using track two of a quarter-track head.
In this next example, the tape changed speed, direction, and track configuration all at the same moment! It started as full-track and went to half-track mono going in opposite directions at different speeds!
At the change you can hear two different things happening: one is backwards music, and the other is a very slow voice that kind of sounds like Jabba the Hutt. In the effort to capture everything on the tape as accurately as possible, we start digitizing with a full-track head and then change our speed and tape head configuration, Once we are at half-track, we will digitize the left side at the appropriate speed and direction, and then the right. When we create the production master file, it reflects all of the content played back correctly.
This post will continue with Part Two.
This might be considered sacrilege, but I’d try it any way. Rather than subjecting the tape to all the starts/stops and speed changes, I would digitize the whole thing at the slowest speed possible (maybe 1 7/8)?) and then shrink/manipulate the digital signal. I believe that it would be faster and far more enjoyable to achieve results working on the computer.
Love what you guys are doing. Keep it up!
i like your blog so amazing thank you and good luck
i like your blog so amazing thank you and good luck
Thanks for this post i like your blog so amazing thank you and good luck