Temporal weights in loudness judgments: A review
* Presenting author
During the past 10 years, a significant number of studies have shown that, in contrast to a sound level meter and widely used measures like the LAeq, human listeners do not weight all temporal portions of a sound equally when they evaluate its overall loudness. Instead, the psychoacoustic data very consistently show a primacy effect: The beginning of a sound is more important for the loudness judgment than later temporal portions. This talk provides an overview of the state of the art concerning temporal weights in loudness (data and modeling approaches). We will discuss a) how the temporal dynamics of the primacy effect can be modeled and how it depends on the sound duration and the duration of the temporal segments, b) how the temporal weights are affected by the presence of a background noise and by variations in sound level, c) how the primacy effect recovers during silent gaps in the sound, and d) whether the temporal weights are applied globally or on a frequency-channel specific basis. The data provide a challenge for current dynamic loudness models and improved modeling strategies will be suggested.