Future directions in loudness standardization: Time-varying loudness

Abstract

The loudness of many sounds varies with time and the temporal processing of the human hearing system affects loudness perception. Existing loudness algorithms that are standardized address loudness of stationary sounds. Two algorithms that predict time-varying loudness have been developed: Zwicker’s, and Moore and Glasburg’s time-varying loudness; variants of the former are available in most sound-quality analysis systems. For continuous sounds with repetitive impulses or beating tones loudness exceeded 5% of the time correlates well with perceptions of overall loudness. The focus here is on predictions of time-varying loudness algorithms for isolated impulsive sounds caused by supersonic aircraft, because recent developments in supersonic aircraft design could result in aircraft producing much quieter booms that may be more acceptable to the general public. Loudness is a major factor in community response; thus, the research focus is to understand how sound modifications (amplitude reduction, signal shaping) affect the loudness of transients. Also, playback systems involve signal filtering; how this may affect loudness perception is of interest. The predictions of the two algorithms are compared for a series of sonic-boom recordings and modified recordings to determine how filtering and signal modification affect loudness predictions.