Prediction models of desirable levels of birdsong and water sound in a noisy environment: A laboratory experiment based on virtual reality
* Presenting author
Augmenting pleasant natural sounds in the noisy urban environment is a key strategy in soundscape design. There have been numerous studies on the positive effects of natural sounds on soundscape quality. However, little attention has been directed to predictive models that suggest appropriate levels of natural sounds at specific ambient noise levels. These models provide a blueprint for practical soundscape design. This study, aims to develop prediction models of desirable natural sound levels to enhance the soundscape quality through laboratory experiments based on virtual reality. The laboratory test consists of two sessions (I and II). In session I, participants were instructed to evaluate traffic sound scenes, ranging from 55dB to 76dB, in terms of perceived loudness of noise (PLN) and overall soundscape quality (OSQ). In session II, participants were instructed to adjust the audio levels of bird and water sounds augmented to each traffic scene to the most desirable levels. At the end of each adjustment, participants were instructed to assess PLN and OSQ for the combined sounds composed of natural sounds mixed with target traffic sounds. Based on the results, soundscape predictive models were developed using acoustic indicators to predict desirable natural sound levels corresponding to traffic noise levels.