News & Events

Towards an understanding of earplug fidelity: distortion mechanisms and measurement techniques

David Anderson, Assistant Professor, Department of Electrical Engineering, University of Minnesota - Duluth

601 Computer Studies Building

Abstract: Earplugs are an essential piece of protection against hearing damage, whether you’re a concertgoer at a music festival or a soldier exposed to high-intensity bomb shockwaves. While we have an ever-growing array of earplugs and other hearing protection devices (HPDs) available to us, the simplicity of the Noise Reduction Rating (NRR) metric used to compare devices obscures complex distortions that introduce directional confusion and a lack of speech intelligibility, even at low insertion loss levels. I will present results from an ongoing effort to create a suite of new laboratory ratings for HPDs that will be correlated with human subject performance in tests involving sound source localization, speech intelligibility, and level matching. In addition, preliminary results will be shown from the development of an in-ear microphone designed to sit behind earplugs in human subjects and record HRTF measurements, illuminating distortions to pinna cues essential to sound localization, when a variety of HPDs are in use. High-fidelity earplugs for musicians and concertgoers also have the extra requirement that musical dynamics and timbre be preserved in the sound reaching the eardrum. Statistical metrics informed by evaluation of musical compression algorithms can be used to test for musical earplug fidelity, providing a roadmap for future design and evaluation of HPDs meant specifically for musicians.

Bio: David A. Anderson is an Assistant Professor in the Department of Electrical Engineering at the University of Minnesota Duluth. He received his PhD from the University of Rochester in 2017 and was previously an Assistant Teaching Professor at the University of Pittsburgh and a Senior Scientist at Applied Research Associates, Inc. His research focuses on measurement of acoustic perceptual distortions, low-power embedded machine learning algorithms and hardware, and development of electroacoustic educational tools. He teaches courses in embedded systems and computer architecture.