Contents:
- Background
- Introduction
- Prototype Test Setup
- Process and Methodology
- Timeline for Each Test Site
- Analysis Methodology
- Findings
- Conclusions
- Next Steps
- Community Support and Action Welcome
This is volume one of a two volume report. It describes the field test carried out in the summer of 2012 that set out to document the occurrence of interstitial errors in digital audio files, defined as lost or altered samples that represent the loss of content and integrity. These errors–often very momentary–result from a failure in the chain of digital data, i.e., in the handoff from the analog-to-digital converter (ADC) to the digital audio workstation (DAW), and in the DAW’s writing of the file to a storage medium. The field test was carried out in three (U.S.) federal audio reformatting facilities, and was designed to identify occurrences of interstitial errors in the course of actual reformatting work. The second item, titled Appendixes, provides background information.The study collected 171 samples representing output from three different devices used over a three to four week period. The goal was to (a) assess the extent to which the errors exist, and the frequency and severity with which they occur and (b) model a prototype test setup and method for detecting and reporting erros. It also includes a section Unexpected Findings that reveals other operational issues in the respective recording facilities that need attention. Given the small amount of samples tested, the results cannot be considered statistically meaningful; however the results do offer a good starting point for future investigations. It concludes that the errors do occur and thus present an issue worthy of serious consideration in audio preservation. Volume two, the appendixes, includes the procedures manual created to carry out the tests, the resulting log sheets, details on the error themselves, notes on the pilot test system and all the report deliverables from 2011.
This valuable report is a must read for IT staff and technicians developing analogue audio to digital conversion workflows for preservation purposes. The study conclusion emphasizes that such errors probably cannot be prevented and thus the need to incorporate monitoring and and file-based management tools to detect these errors in the resulting digital files. Such tools are currently under development by the report authors and will be available for archives in the future.