Loudness and noise rating: How may standardization bridge the gap between science and engineering?

Abstract

Current noise regulations and standards, in Europe and to some extent also elsewhere, are primarily based on A-weighted sound levels. Most perceptual aspects are—if at all—taken into account by additional correction levels, for example, punishing tonality or impulsive content. On the contrary, instrumental loudness-prediction methods have been available for more than half a century and were first standardized 40 years ago. Against this background, the question arises why loudness—regardless of being calculated according to Zwicker or Moore and Glasberg—has not yet been accepted more widely in the field of noise control and related standardization. This contribution attempts to raise some possibly influential aspects, and to discuss them in the light of next-generation loudness standards. Advantages of stationary and time-varying loudness predictions over corrected weighted levels will be discussed and contrasted with potentially contradicting arguments. The discussion is primarily intended to provide arguments for the selection of future loudness-prediction standards. However, also a broader perspective is taken, looking at interrelations with other standards, algorithmically and regarding limiting values.