Temporal perceptual weights in loudness: The role of sound duration

Most current technical measures of loudness (e.g., LAeq) assume that two temporal stimulus components with identical spectrum and level have the same impact on the overall loudness of a longer sound, regardless of their temporal position within the sound (e.g., beginning versus end). In contrast to this assumption, many studies consistently showed that listeners apply strongly non-uniform temporal weights when judging the overall loudness of a longer, time-varying sound. The first 100-300 ms receive a higher weight than later portions of the stimulus (primacy effect). Some studies also showed a recency effect, i.e., a slightly higher weight assigned to the end of a signal. This talk addresses the dependency of temporal loudness weights on sound duration. While most studies used stimulus durations in the order of 1 s, some experiments presented considerably shorter or longer sounds, ranging from about 200 ms to 3 s. The results of these studies will be reviewed, and potential explanations of the effect of sound duration on temporal loudness weights will be discussed.