On the use of statistics of a time-varying loudness model predictions to quantify loudness of transient environmental noise.

摘要

It may be possible to build supersonic aircraft that produce low-level sonic “booms”, but even ifthey are much quieter than their more traditional counterparts, how acceptable would these be,and how does the impact relate to that of other transient environmental sounds? It is likely thatthe louder the transient, the more annoying or disturbing it is, but when the loudness variessignificantly throughout the event, how should event loudness (a single number) be determinedfrom a sound recording? The rapid rise of loudness may also evoke a startle response. Manyenvironmental transient noise events also have high levels of low frequency energy which canpropagate through building structures with little attenuation. The use of statistics of thepredictions from a time-varying loudness model, developed by Glasberg and Moore, to predictloudness and startle was investigated. This model was investigated because it incorporatesmodels of the temporal behavior of the human hearing system, clearly of importance with rapidlychanging sounds, and it is claimed that it models the response to low frequency noise better thanother loudness models. The impact of sonic boom shape changes on the estimated time-varyingloudness was investigated, as was their impact on the output of an annoyance and a startle modeldeveloped from prior research.