Localization of tones in a room by moving listeners

Abstract

It is difficult to localize the source of a tone in a room because standing waves lead to complicated interaural differences that become uninterpretable localization cues. This paper tests the conjecture that localization improves if the listener can move to explore the complicated sound field over space and time. Listener head and torso movements were free and uninstructed. Experiments at low and high frequencies with eight human listeners in a relatively dry room indicated some modest improvement when listeners were allowed to move, especially at high frequencies. The experiments sought to understand listener dynamic localization strategies in detail. Head position and orientation were tracked electronically, and ear-canal signals were recorded throughout the 9 s of each moving localization trial. The availability of complete physical information enabled the testing of two model strategies: (1) relative null strategy, using instantaneous zeros of the listener-related source angle; and (2) inferred source strategy, using a continuum of apparent source locations implied by the listener's instantaneous forward direction and listener-related source angle. The predicted sources were given weights determined by the listener motion. Both models were statistically successful in coping with a great variety of listener motions and temporally evolving cues.

Speech Intelligibility in Reverberation is Reduced During Self-Rotation

“…Such scanning is evident in A-only for the frontal target ( Figure 3 ) in which trajectories are more complex than trajectories in the AV condition and self-rotations at the end of the target sentence show higher normalized variance in A-only than AV, possibly because the target was more difficult to localize. Macaulay and Hartmann (2021) analyzed sound localization strategies of sitting people in a task with unrestricted head movements and showed that people integrate and weight localization cues to guide their behavior. In the present study, the task was to understand speech from various target locations and respond with self-rotation as if somebody was talking to the participant, which inherently includes both sound localization and speech perception.…”

“…In the present study, the task was to understand speech from various target locations and respond with self-rotation as if somebody was talking to the participant, which inherently includes both sound localization and speech perception. It is possible that people used a similar strategy for guiding behavior in the current experiment as in the experiments by Macaulay and Hartmann (2021) . However, this search and localize behavior is likely to be dominant in the A-only condition, while the appearance of the visual cue might have triggered another strategy.…”