Seeing Voices: Experience with Imaging Methods for Recorded Sound Preservation

The IRENE project began as a collaboration between Lawrence Berkeley National Laboratory and the Library of Congress to research and build systems to digitally image recorded sound collections for high throughput digitization, for extracting sound from broken and degrading carriers and as an analysis tool to assess the condition of media. After more than a decade of research, multiple systems have been built and contributed to the preservation of recordings long thought beyond rescue. The IRENE System has proven itself to be a robust methodology that is benefiting from modern imaging and data intensive processing technologies. Digitizing audio media in their physical state before they are converted to a time based signal allows for analysis, selection and restoration of the groove features to include in the processed audio. This capability was unimaginable without the ability to process large data sets with modern workstation class computers. Both 2D and 3D imaging capabilities provide a balance of speed and data density that has been leveraged to image commercial, unique and experimental media. Software development to capture sound from badly broken and degrading media has broaden dramatically in contrast to the capabilities of the early system. This panel will act as an historical summary of the the IRENE System, gauge its position in the preservation of recorded sound collections, provide an understanding of the underlying technologies and advance future development efforts.
 
Dr. Carl Haber, Senior Scientist, Lawrence Berkeley National Laboratory, “The IRENE System: The First 10+ Years of Preserving Audio through Imaging”
Mr. Peter Alyea, Digital Conversion Specialist, Preservation Reformatting Division, Library of Congress, “Unusual Solutions to Unusual Problems: How Imaging Saved Lost Sound”