Top of page

Tracing Sound: Adding Functionality to Contactless Audio Recording Digitization

Share this post:

The following is a post by Harrison Biggs, a 2022 preservation science intern with the Preservation Research and Testing Division at the Library of Congress. He is a rising Senior at the University of Michigan, pursuing a degree in Material Science and Engineering with a minor in Museum Studies.

A digital photograph of Harrison Biggs
Harrison Biggs

This summer I had the honor of working with Peter Alyea in the Preservation Research and Testing Division, developing new functionality for the IRENE system.

IRENE, which stands for Image Reconstruct Erase Noise Etc., is a method and system for optically extracting sound from mechanical audio recordings, such as phonograph records and cylinder records. What makes IRENE special is that it works by imaging, rather than touching/playing, making the entire process contactless. This allows us to extract audio from broken, cracked, fragile and otherwise unplayable objects while reducing the possibility of damage. The system was created in the early 2000s by physicists Dr. Carl Haber and Dr. Vitaliy Fadeyev at the Lawrence Berkeley National Laboratory in California but has been in continuous development thanks to the efforts of over 75 individuals. Of particular note is Dr. Earl Cornell, a software designer at Laurence Berkeley National Lab and the current caretaker and head designer of the software that converts images to sound. Today the system is in use at the Library of Congress’s National Audiovisual Conservation Center, the Northeast Document Conservation Center, the UC Berkeley Library, and a handful of other institutions around the globe. One of the more notable projects using IRENE has been at the Smithsonian’s National Museum of American History, where they’ve used the system to digitize experimental recordings from Alexander Graham Bell’s Volta Lab.

The process behind IRENE currently takes place in two parts. The first involves a hardware setup with a either 3D imaging probe or a specialized microscope, and a set of motors and stabilizers. We start by loading an object into the system, placing it next to, but not touching the probe/lens. We then spin the object in front of the probe/lens, moving it slightly after each rotation so as to scan all of the grooves. When using the probe, this gives us a topographical map, which is then converted into a 2D, black and white image, where pixel intensity corresponds to the height on the surface of the object. When using the microscope, we receive a 2D image, which is left in color.

Wax cylinder sits on a metal bar next to a 3D probe, all resting on a metal table.
The IRENE system, scanning a cylinder record. The cylinder is placed on a stabilized mandrel, and the entire system is built on a stabilized table. The slightest vibration could ruin the scan. Photo credit: Harrison Biggs

We then send this image to the second part of IRENE: a piece of software called Weaver. In Weaver, we process the image, digitally trace the grooves, simulate the motion of a stylus through those grooves and then calculate and output an audio file. This is the part of the project I’m working on. My goal for the summer is to add a new method of manually tracking grooves, using drawing tablets. The thinking is that tracing the grooves with a drawing tablet may be more intuitive and ergonomic than using a traditional mouse and keyboard.

Drawing tablet in front of a laptop displaying a black and white close up of a record. A red line follows a thinner white line in the image.
The Weaver software in the middle of tracking a groove with a drawing tablet. Note the left side of the screen, where the plugin chain is displayed. These can be removed, adjusted, or replaced at any time. Photo credit: Harrison Biggs

Weaver is designed to be entirely modular; each aspect of its operation (loading files, modifying them, tracing grooves, and saving/exporting data) is written into different plugins which can be operated in practically any order or combination. Modifying an existing plugin to accept and process data from a tablet is relatively simple. What’s been challenging is actually getting data from the tablets and passing it to the plugin. The tablet producer provides libraries that can be used to interface with their hardware, but they are minimally documented. Most of the time I’ve put into my internship so far has been spent reverse engineering demo applications and experimenting with settings. Implementing features and making sure they work is only the first part of software design. The second step is designing and perfecting the user experience, and this is the part of the process I have the least experience in. Luckily, I’ve had Peter and Dr. Cornell to guide me through this part of the process. Not only have they offered their own opinions about the system and how it can be improved, but they’ve helped me get in touch with and solicit opinions from other regular users.

Ultimately, my summer has been fairly successful. I’ve manage to create a working plugin that can take pen, touch and button data from the tablet, and operate on that data as needed, without too many bugs or slowdowns. It’s been a challenge, and the plugin is far from perfect, but I’m proud of what I’ve accomplished. Hopefully in the future someone with more skill than me will be able to take what I’ve done and turn it into something more polished. In the meantime, my product is usable and will (hopefully) be useful to those working with IRENE and Weaver.

This internship has been an amazing opportunity to enhance my existing programming skills and see how they can be applied to collections work. It also allowed me to exercise some of my communication skills and helped me improve my documentation and note taking.

Special thanks to Peter Alyea for all his support and for giving me the opportunity to work with him this summer, and to Dr. Earl Cornell for all of his assistance though the course of my project.

Subscribe to the blog— it’s free! — learn how the largest library in the world preserves the coolest stuff in the world.

Add a Comment

This blog is governed by the general rules of respectful civil discourse. You are fully responsible for everything that you post. The content of all comments is released into the public domain unless clearly stated otherwise. The Library of Congress does not control the content posted. Nevertheless, the Library of Congress may monitor any user-generated content as it chooses and reserves the right to remove content for any reason whatever, without consent. Gratuitous links to sites are viewed as spam and may result in removed comments. We further reserve the right, in our sole discretion, to remove a user's privilege to post content on the Library site. Read our Comment and Posting Policy.

Required fields are indicated with an * asterisk.