Auditory Spectro-Temporal Pattern Analysis

摘要

Progress is reported here in four areas. The work on the relation between CMR andthe MLD has suggested that part of the MLD is based upon an analysis of dip information in the stimulus waveform. The work on Mechanisms of modulation detection interference (MDI) has supported the across channel nature of MDI, and has reinforced the notion that MDI depends upon the average modulation frequency, but not upon the specific modulation pattern or phase. Results suggest that the effect is widely tuned for carrier frequency, and that MDI may not be the most appropriate paradigm for investigating grouping by common modulation, where carrier specific information must be maintained. The present proposal therefore utilizes paradigms other than MDI. Studies on CMR for suprathreshold signals indicated that large CMR effects are restricted to stimuli presented near detection threshold. Furthermore, CMR information CMR information available at low SLs appears to be coarse, resulting in large DLs for amplitude discrimination, and in poor gap thresholds. Companion experiments on the MLD indicated that the magnitude of suprathreshold masking release was highly similar for the MLD and CMR. In the work on CMR and Auditory Grouping we have investigated the notion that CMR can basically be understood as an auditory grouping process. When a comodulated noise background exists, the individual narrow bands of noise comprising the stimulus may be interpreted as a single auditory source, by virtue of the comodulation among the bands. When a single pure tone signal is presented in one of the bands, a possible cue for the detection of the signal is the change in the envelope of the on signal band that does not occur in the other bands.