TY - JOUR
T1 - Temporal integration at 6 kHz as a function of masker bandwidth
AU - Oxenham, Andrew J.
PY - 1998
Y1 - 1998
N2 - Thresholds were measured for a 6-kHz sinusoidal signal presented within a 500-ms masker. The masker was either a bandpass Gaussian noise of varying bandwidth, or a sinusoid of the same frequency as the signal. The spectrum level of the noise masker was kept constant at 20 dB SPL, and the level of the sinusoidal masker was 40 dB SPL. Thresholds for signal durations between 2 and 300 ms were measured for masker bandwidths ranging from 60 to 12 000 HZ. The masker was spectrally centered around 6 kHz. For masker bandwidths less than 600 Hz, the slope of the temporal integration function decreased with decreasing masker bandwidth. The results are not consistent with current models of temporal integration or temporal resolution. It is suggested that the results at narrow bandwidths can be understood in terms of changes in the power spectrum of the stimulus envelope or modulation spectrum. According to this view, the onset and offset ramps of the signal introduce detectable high-frequency components into the modulation spectrum, which provide a salient cue in narrowband maskers. For broadband maskers, these high- frequency components are masked by the inherent rapid fluctuations in the masker envelope. Additionally, for signal durations between 7 and 80 ms, signal thresholds decreased by up to 5 dB as the masker bandwidth increased from 1200 to 12 000 Hz. The mechanisms underlying this effect are not yet fully understood.
AB - Thresholds were measured for a 6-kHz sinusoidal signal presented within a 500-ms masker. The masker was either a bandpass Gaussian noise of varying bandwidth, or a sinusoid of the same frequency as the signal. The spectrum level of the noise masker was kept constant at 20 dB SPL, and the level of the sinusoidal masker was 40 dB SPL. Thresholds for signal durations between 2 and 300 ms were measured for masker bandwidths ranging from 60 to 12 000 HZ. The masker was spectrally centered around 6 kHz. For masker bandwidths less than 600 Hz, the slope of the temporal integration function decreased with decreasing masker bandwidth. The results are not consistent with current models of temporal integration or temporal resolution. It is suggested that the results at narrow bandwidths can be understood in terms of changes in the power spectrum of the stimulus envelope or modulation spectrum. According to this view, the onset and offset ramps of the signal introduce detectable high-frequency components into the modulation spectrum, which provide a salient cue in narrowband maskers. For broadband maskers, these high- frequency components are masked by the inherent rapid fluctuations in the masker envelope. Additionally, for signal durations between 7 and 80 ms, signal thresholds decreased by up to 5 dB as the masker bandwidth increased from 1200 to 12 000 Hz. The mechanisms underlying this effect are not yet fully understood.
UR - http://www.scopus.com/inward/record.url?scp=0031909636&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0031909636&partnerID=8YFLogxK
U2 - 10.1121/1.421229
DO - 10.1121/1.421229
M3 - Article
C2 - 9479757
AN - SCOPUS:0031909636
SN - 0001-4966
VL - 103
SP - 1033
EP - 1042
JO - Journal of the Acoustical Society of America
JF - Journal of the Acoustical Society of America
IS - 2
ER -