Research

The vibrational response of an elastic panel to incident acoustic waves is determined by the direction-of-arrival (DOA) of the waves relative to the spatial structure of the panel's bending modes. By monitoring the relative modal excitations of a panel immersed in a sound field, the DOA of the source may be inferred. In reverberant environments, early acoustic reflections and the late diffuse acoustic field may obscure the DOA of incoming sound waves. Panel microphones may be especially susceptible to the effects of reverberation due to their large surface areas and long-decaying impulse responses. An investigation into the effect of reverberation on the accuracy of DOA estimation with panel microphones was made by recording wake-word utterances in eight spaces with reverberation times (RT60s) ranging from 0.27 to 3.00 s. The responses were used to train neural networks to estimate the DOA. Within ±5°, DOA estimation reliability was measured at 95.00% in the least reverberant space, decreasing to 78.33% in the most reverberant space, suggesting an inverse relationship between RT60 and DOA accuracy. Experimental results suggest that a system for estimating DOA with panel microphones can generalize to new acoustic environments by cross-training the system with data from multiple spaces with different RT60s. Published in the Journal of the Acoustical Society of America in October 2024.

The vibrational response of a panel to an incoming pressure wave is contingent upon the relationship between the incident angle and the modal structure of the panel. Estimating the direction of arrival (DOA) of the incident wave is possible by analyzing the relative modal excitations measured by a single sensor affixed to the panel. Reverberation can distort the wave's true DOA due to reflections from numerous angles present in the recording made by the sensor. Panel microphones, known for their extended internal reverberation times, may be particularly vulnerable to these effects. In this study, we investigate the impact of reverberation on the reliability of DOA estimation performed with panel microphones. Experiments were conducted by recording wake-word utterances in eight diverse acoustic spaces, each characterized by varying reverberation times (RT60s). The vibrational responses of the panel were used to train neural networks for DOA estimation. Results indicate an inverse relationship between RT60 and the reliability of DOA estimation. Results also indicate DOA estimation utilizing panel microphones can effectively adapt to diverse acoustic environments by training the system with data from various environments. These findings have implications for applications such as smart speakers and voice- activated devices, which are commonly used in residential settings. Importantly, results consistently achieved DOA estimation accuracy exceeding 98% within 10 degrees in environments with RT60s typical of living rooms, establishing that single-sensor panel microphones are a viable alternative to traditional microphone arrays for estimating acoustic DOA. Presented at WNYISPW 2023.

Room acoustics metrics were evaluated across the University of Rochester campus to improve the acoustic characteristics of spaces and to provide students with room data for educational projects. Impulse responses were taken and distributed to students, as well as the recoreded characteristics, dimensions, and testing setup of the spaces. Pictured here is the Rush Rhees Rotunda above the campus library. RT60 = 3.00s, C80 = -2.0dB, EDT = 3.27s.

Jenna currently leads upper-level recording labs and acoustics recitations at the University of Rochester. The goal of these courses is to provide students a hands-on learning experience for analog and digital recording practices, miking techniques, mixing and editing techniques, ear training, acoustics, and more. Pictured here is the Gavett Recording Studio at the University of Rochester, equipped with an API 1608-II.

Current studies focusing on the use of flat panels as loudspeakers, microphones, smart devices, and touch interfaces are ongoing.