End-level dependency of loudness of increasing sounds: a time persistent effect
Résumé
Loudness of stationary sounds is very well established, but little is known when considering loudness of non stationary sounds. Several studies have considered loudness of increasing and decreasing sounds lasting several seconds. Recent results have shown that global loudness of an up-ramp measured by direct magnitude estimation is strongly dependent on the end-level (Susini et al., 2010; Teghtsoonian et al., 2005). In the present study, we examine whether this end level bias is a consequence of the experimental method used when asking participants to estimate the loudness just after the end of the up-ramp. Thus, a paired comparison experiment was performed during which participants compared the global loudness of a test and a probe tone at different inter-stimulus intervals [0.25, 1, 4, 8 s]. The test tone was either an up-ramp [65-80 dB] or a constant tone corresponding to its end-level [80 dB]. Probe tones were constant tones varying in level. Results show that the point of subjective equality (PSE) of the [80dB] constant tone decreased slightly with the ISI. A similar trend was obtained for the [65-80 dB] up-ramp, but values of the PSEs were about 1.4 dB lower than for the constant tone whatever the ISI was. Finally, using another experimental method, this study reveals that the global loudness of an upramp follows the same trend as the loudness of a tone presented at a level corresponding to its end-level. This result persists even after few seconds showing a long-term memory process.