AIC's 41st Annual Meeting

The Library of Congress (LC) holds more than 500,000 magnetic tape objects, many of which are degrading rapidly. As at many cultural heritage institutions and archives, a rapid method to identify degraded tape is needed to facilitate treatment prioritization before copying and/or digitizing. Even in ideal storage conditions, tapes degrade. Tapes produced during the 1970-1990’s often contain polyester-urethane (PEU) binders that hold magnetic particles onto polyethylene terephthalate substrates. PEU binders are known to degrade via hydrolysis, which causes squealing and/or shedding of magnetic material onto playback device heads. This condition is referred to as ‘sticky shed syndrome’ (SS). There are no known non-destructive methods for rapidly identifying degraded magnetic tapes. Several brands and models of tape are known to contain PEU binders and are known to degrade, however tapes are rarely held in their original packaging or even kept on original hubs, making classification by visual inspection impossible. Playing a tape is the currently accepted method for classifying a tape as SS or non-SS. If the tape squeals, flakes, or gums playback equipment, it is classified as SS and removed from the digitization workflow for treatment. This process can not only render the playback device unusable, but it can permanently damage the tape and lead to loss of data.

Assessment of 100 LC ¼” audio collection tapes led to the development of a method for the rapid identification of degraded tapes prior to playback. Using Fourier transform infrared (FTIR) spectroscopy combined with multivariate statistics (MVS), the method’s rate of accurately assessing tape condition was better than 98%.

This talk will focus on the application of the classification model produced from LC collection-based ¼” audiotape FTIR data to other formats such as U-matic, VHS, and ½” video. The ¼” dataset classification model was also applied to data collected from ¼” audio reference tapes of known provenance spanning multiple manufacturers and models. Establishing the applicability of the classification model to a broad and known set of tapes was completed to predict the usefulness of the model for collections outside of the LC and may allow for the identification of specific makes and models within a particular collection. In addition, direct analysis in real time mass spectrometry (DART-MS) was conducted on a set of degraded and non-degraded ¼” audio and ½” video tapes to investigate their chemical differences.

FTIR combined with statistical analysis is envisioned as a tool for rapid, on-site analysis of collection materials. However, attenuated total reflectance (ATR)-FTIR directly contacts the tape, and its use may affect the surface and possibly the sound or image fidelity. Modern ¼” audio tape was pre-recorded with standard test tones, subjected to ATR-FTIR testing, and played through an analog to digital recorder. The post-ATR-FTIR signals were compared to the pre-recorded signals at the same locations. The results of this testing will also be discussed.