Development and application of a semantic differential for perception-based optimization of wind turbine and other broadband sounds

Abstract

Growing numbers of the population are exposed to wind turbine (WT) and other broadband sound sources from the environment. Acoustically optimizing emissions of environmental noise sources to minimize negative effects on the population therefore becomes increasingly important. The objective of this study was to develop and apply a semantic differential (SD) for multidimensional perception-based assessment of WT and other broadband sounds, to be potentially applied for acoustic optimization. An SD was developed specifically for broadband sounds and applied in a laboratory listening experiment, using outdoor WT and other broadband sounds covering a wide spectral range, at 40 dBA playback level. Fifty-two participants rated the sounds using the SD. Relevant perceptual dimensions were extracted, and a prediction model for noise annoyance was established on this basis. The experiment revealed that participants could well describe and discriminate the sound characteristics using the affective-evaluative everyday language of the SD. Four perceptual dimensions (or factors) were identified: Evaluation, Timbre, Activity and Randomness, with the latter three describing spectral shape, periodic amplitude modulation and random amplitude modulation, respectively. The factors were strongly linked with annoyance and well suited to establish an annoyance prediction model. The results can be applied in a perception-influenced design to identify how to optimize (perceived) acoustical characteristics and thus minimize annoyance effects on the population. Also, they might potentially be used in field surveys for a multidimensional assessment of broadband sounds to study the long-term annoyance potential of specific perceptual characteristics. This might eventually help to refine exposure–response relationships using predictors beyond basic noise metrics.