Many natural sounds including speech show common temporal level fluctuations in different frequency regions, i.e. they are comodulated. A psychoacoustical effect related to our sensitivity to such comodulation is "Co-modulation Masking Release" (CMR) [1][6]. In CMR experiments, it is found that thresholds of a sinusoidal signal masked by noise are lower if the masker is co-modulated compared to a masker with uncorrelated level fluctuations in different frequency regions. So far, CMR has been investigated using stationary spectra. However, natural sounds often show, apart from temporal comodulation, also non-stationary spectra and spectre-temporal modulations [3]. The auditory system seems to be sensitive to those modulations [4]. The present study investigates if CMR is still observed when the stimuli have these other features of spectral and spectre-temporal modulations in addition to comodulation.
展开▼
