The acoustics features and their relationship to the emotional characteristics of rain sound effects

Abstract

Rain sound effects are commonly used in games, animations, and films, but are still relatively new area in that the relationship between their features and emotional characteristics has not been studied. This paper investigates the effects of several subjective and objective features on the emotional characteristics of rain sound effects via a listening test, linear regression analysis and spectral analysis. In the listening test, listeners rated the sounds over Arousal, Valence, six features, and sixteen emotional categories. Linear regression was performed afterward to validate and further investigate the results. Meanwhile, spectral analysis using seven acoustics features was performed to provide an objective understanding for the rain sound effects. The results showed that Arousal significantly increased with increases in the properties Density, Impact, Intensity, Loudness, and Quantity, while Valence decreased. It was also found that most of the emotional categories were significantly influenced by these properties. The emotional categories Agitated, Angry, Scary, Violent, Anxious, and Nervous significantly increased with these properties, while the categories Comforting, Calm, Gentle, Relaxed, and Tender decreased. However, Mysterious was not significantly effected by any property. Generally, related emotional categories were affected by rain properties in a similar way, especially for the Angry and Calm quadrants. For Negative emotional characteristics, Arousal-related properties increased with increasing RMS Amplitude and Spectral Flux. However, they decreased with increasing Spectral Irregularity. These results give some possible general suggestions for editing and choosing rain sound effects in computer games and films.

Date of Award	2019
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology