Use of orthogonal treatment of acoustic attributes to study interaction effects in sound quality design

Abstract

Traditional methods of studying psychoacoustic attributes rely on subjective measurement of non-orthogonal sound samples. However, this approach has a narrow inference space due to the non-orthogonality of the acoustic attributes. This research aims to utilize an orthogonal treatment of acoustic attributes to study interaction effects in the sound quality of a fan-powered HVAC appliance by manipulating critical psychoacoustic attributes to create a broadened inference space. The results show that, in addition to the effects of loudness, sharpness, roughness, and prominence ratio, the roughness*prominence ratio, sharpness*prominence ratio, and the roughness*sharpness*prominence ratio interactions were statistically significant. Hence, this work demonstrates the importance of sound attribute interactions to consumers. Furthermore, comparing the empirical models without and with significant interactions demonstrated improvements in predicting the consumer response when including interaction effects. This work recommends considering psychoacoustic interactions, especially sharpness*roughness*prominence ratio, in designing products for improved consumer satisfaction.