The loudness model used in ISO532-3: Development, evaluation and prospects

Abstract

ISO 532-2 is based on the loudness model for stationary sounds described by Moore and Glasberg [JASA (2007)]. This model is similar to that in ANSI S3.4-2007, except that the model in ISO 532-2 incorporates binaural inhibition: a strong input to one ear in a given frequency region reduces the effective level of a weaker input to the other ear in nearby frequency regions. ISO 532-3 has been proposed as an extension to ISO 532-2 to deal with time-varying sounds. It generates predictions of short-term loudness, the loudness of a short-segment of sound such as a word in a sentence or a single note in a piece of music, and of long-term loudness, the loudness of a longer segment of sound, lasting 1–5 s, such as a whole sentence or a musical phrase. The model gives reasonably accurate predictions of the overall loudness of technical sounds (e.g., factory noises), of speech whose dynamic range has been compressed or expanded, and of sounds whose time pattern and spectra differ at the two ears. However, the model needs to be extended to generate predictions of the overall loudness impression of a sound environment over a period of several minutes to an hour.