Contribution
Memory for Room Acoustics
* Presenting author
Abstract:
The cognitive mechanisms behind auditory memory for room acoustics are still poorly understood; historically, auditory memory research has focused more on the ability to recall speech content, and to discriminate pitch or timbre. However, a characterization of how well listeners can remember the acoustics of rooms is crucial for understanding the requirements of extended reality applications, and for the design of room acoustic experiments in general. Therefore, this study aims to test the ability to detect different room acoustics at different retention intervals in the short-term memory range. For this purpose, we conducted a listening experiment using binaural auralizations of speech based on spatial room impulse response measurements of 61 rooms. In each trial, participants listened to auralizations of two rooms, separated by varying silent intervals, and were then asked whether they were identical or not. Also, either the same or a different speech signal was used for creating the two renderings, and the rooms were taken from three different, pre-determined similarity groups. The results demonstrate that memory performance declines as retention intervals increase and that significant differences in performance between using the same or different speech signals exist. Furthermore, similar rooms were harder to distinguish than dissimilar rooms.