Toward Technology-Driven Metadata Encodings for Historical Recordings

Pre-digital recording technology presents numerous challenges and idiosyncrasies for cataloguing and use. Discs and cylinders underwent considerable change in the recording and playback speed over the first twenty-five years of the 20th century before final industry standardization of 78.2 rpm in 1925. From 1910 to 1925, most, but by no means all, utilized a speed of 78-80 rpm. Before 1910, however, speeds could vary widely between 72-86 rpm (or more). Even with the advent of electronically driven rotational motors in the 1920s, changing over to them from mechanical motors was not instantaneous across the recording industry, shrouding their use in research with broad ambiguity. The continual degradation of recording media – wax, Ebonite, Durinoid, shellac, magnetic tape, vinyl – with each playback instance further obscures their usability. What results for modern users is that unless the precise recording and playback speeds are known, significant uncertainty exists for how one accesses the disc’s audio contents. Even using pre-digital playback equipment, what a user ends up accessing is a hypothetical or simulacrimal sound document rather than an assuredly exact reflection of the sonic events at the time of recording. Digitization of these recordings does not automatically mitigate such challenges, but instead adds additional layers of complication for precisely defining attending metadata. Metadata accounting for these kinds of idiosyncratic properties of individual items suffer from a lacuna in current systems, even as they should readily appear with the catalogue items and research data, enabling FAIR/CARE adherence.