Global and Continuous Loudness Estimation of Time-Varying Levels

Summary Conventional psychoacoustic methods are not sufficient to produce instantaneous judgments or real-time perceptual tracking of nonstationary sounds. In order to evaluate continuously or globally the loudness of pure tones of variable duration and time-varying level, two cross-modal matching methods were used, one with continuous force feedback and another without force feedback but using a continuous analogical/categorical judgment scale. The global loudness of various predefined acoustic level profiles was estimated under two experimental conditions: one with continuous estimation during the sound sequence and the other without. The results show that the continuous judgment profiles transcribe quite well the stimulus contours, although a temporal lag on the order of 1s between stimulus contour and response profile is observed, as is an asymmetry between increasing and decreasing profiles. Global judgments are influenced by the rate or duration of level change and by the level at the end of the signal. A recency effect, similar to that observed in auditory memory research using an immediate recall task, is thus revealed for loudness estimates on nonstationary sounds lasting a few tens of seconds. Finally, global judgments are generally higher without preceding continuous evaluation for both matching methods.