Masking functions and fixed-signal functions for low-level 1000-Hz tones

Abstract

Masking functions and fixed-signal functions were constructed using a narrow range of pedestal intensities for 10-ms, 1000-Hz gated tones. Data from three experiments agreed with previously reported data, clearly demonstrating negative masking and the pedestal effect. The data extend earlier findings by showing (1) the resilience of the pedestal effect when a background noise masker is introduced; (2) a possible indifference of the fixed-signal function to stimulus duration; (3) the ability of a set of psychometric functions to produce both masking and fixed-signal functions; (4) depending on method, the impact of unit choice on the interpretation of both the pedestal effect and negative masking data. Results are discussed in relation to current psychophysical models, and suggest that accounting for the auditory system's sensitivity to differences in low-level sounds remains a challenge.

Informational masking of negative masking

“…Namely, (1) detectability improves, (2) the psychometric function flattens, and (3) NM is obtained. All three of these "uncertainty reduction effects" are borne out in the psychoacoustic literature on NM (e.g., Hanna et al, 1986;Viemeister and Bacon, 1988;Shepherd and Hautus, 2009;Shepherd et al, 2013) and this suggests that when NM is quantified via the equation…”

“…The latter two interpretations (units selection and Laming's sensory analytical model) have been addressed previously (for units selection, see references above; for the sensory analytical model, see Laming, 1985Laming, , 1986Hanna et al, 1986;Shepherd and Hautus, 2009) and are not considered in detail here. Suffice it to say that while units selection is not wholly compelling based on the most recent data (e.g., Shepherd and Hautus, 2009;Shepherd et al, 2013) Laming's model remains a viable interpretation of NM under certain conditions. In fact, of the handful of empirical studies of auditory NM that have appeared in the literature since Laming first proposed his model (Hanna et al, 1986;Viemeister and Bacon, 1988;Shepherd and Hautus, 2009;Shepherd et al, 2013) none have presented data that are entirely inconsistent with its basic predictions (although see Moore, 1988 for criticisms of Laming's model outside of the context of NM).…”

“…Suffice it to say that while units selection is not wholly compelling based on the most recent data (e.g., Shepherd and Hautus, 2009;Shepherd et al, 2013) Laming's model remains a viable interpretation of NM under certain conditions. In fact, of the handful of empirical studies of auditory NM that have appeared in the literature since Laming first proposed his model (Hanna et al, 1986;Viemeister and Bacon, 1988;Shepherd and Hautus, 2009;Shepherd et al, 2013) none have presented data that are entirely inconsistent with its basic predictions (although see Moore, 1988 for criticisms of Laming's model outside of the context of NM). That said, in this study, we chose to focus on the CUH.…”

“…Namely, under certain conditions, the detectability of the signal can improve in the presence of a pedestal and the level of the just detectable masked signal at threshold, DA JND , can "dip" (cf. Solomon, 2009) below A 0 (Hanna et al, 1986;Viemeister and Bacon, 1988;Shepherd and Hautus, 2009;Shepherd et al, 2013). This finding has been termed negative masking (NM) (Raab et al, 1963a) due to the fact that it represents a situation in which, contrary to Weber's law, sensitivity improves under masking.…”

“…In audition, NM occurs in the context of intensity discrimination, when both the signal and the pedestal are near 0 dB SL. It has been found using both broadband noise (Raab et al, 1963a,b) and pure-tones (Hanna et al, 1986;Viemeister and Bacon, 1988;Shepherd and Hautus, 2009;Shepherd et al, 2013). In this study, we focus on pure-tone signals and tonal pedestals and so, for simplicity, all subsequent discussions of NM are in terms of such stimuli.…”