Temporal loudness weights in background noise: Data and models

Abstract

Previous studies consistently showed that human listeners primarily consider the beginning of a time-varying sound when judging its overall loudness, and place less weight on subsequent temporal portions. However, all experiments studying this primacy effect in temporal loudness weights presented the target sound in quiet. Here, we compared temporal weights when the target sound was either presented in quiet or in a continuous background noise, and for a variation in the level of the target sound across a range of 60 dB. The target sound was a time-varying narrowband noise, the background noise was a continuous bandpass-filtered noise. In all conditions, we observed the expected primacy effect, well described by an exponential decay function using parameters based on previous studies. The patterns of temporal weights were very similar in conditions with and without background noise, and largely independent of the variation in target level. Simulations using a dynamical model for the partial loudness of time-varying sounds in background noise showed that the model does not predict the observed temporal loudness weights. Strongly depending on the loudness statistic extracted from the loudness model, the predicted loudness weights were either uniform, or showed an inverse U-shaped pattern with higher weights predicted for temporal portions in the middle of the sound compared to temporal portions near onset or offset.Previous studies consistently showed that human listeners primarily consider the beginning of a time-varying sound when judging its overall loudness, and place less weight on subsequent temporal portions. However, all experiments studying this primacy effect in temporal loudness weights presented the target sound in quiet. Here, we compared temporal weights when the target sound was either presented in quiet or in a continuous background noise, and for a variation in the level of the target sound across a range of 60 dB. The target sound was a time-varying narrowband noise, the background noise was a continuous bandpass-filtered noise. In all conditions, we observed the expected primacy effect, well described by an exponential decay function using parameters based on previous studies. The patterns of temporal weights were very similar in conditions with and without background noise, and largely independent of the variation in target level. Simulations using a dynamical model for the partial loudness of ti...