The effect of cross-channel synchrony on the perception of temporal regularity
Schematic representation of the stimuli: In this example, the target band (bottom trace) contains the regular target click train (solid lines) and the irregular masker click train (dotted lines). The three different flanker conditions are illustrated separately above the target band; : flanker-like target; : flanker-like target, but with a 5 ms delay; : flanker-like masker. Only one flanker was presented with the target band in the experiment.
Mean (top left panel) and individual (other panels) regularity thresholds as a function of flanker frequency; the parameter is the flanker mode (see the legend in the top left panel). The horizontal solid lines show the threshold for the no-flanker condition. The vertical dotted lines mark the center frequency of the target band.
Mean threshold, averaged over flanker frequencies below (squares) or above (circles) the target band, for each of the three flanker modes. As in Fig. 2, the horizontal line shows the threshold for the no-flanker condition.
Summary AIs for the target (solid line) and flanker bands (dashed line). In this example, the target band contained the regular target click train, and the flanker click train was centered at 0.4 kHz and was regular and synchronous with the regular target click train .
Results of the first set of model simulations, in which the cross-channel interactions were implemented after the time-interval analysis stage. In all cases, the abscissa represents the flanker frequency and the ordinate shows the squared difference between the stimuli in which the target band contained the regular target or the irregular distracter click train at a signal-to-masker level of 0 dB (see the text). The upper panel shows the results when is calculated from the summary AI of the target band alone, without any consideration of the flanker image (no interaction). The middle panel depicts the results of the simulation in which was calculated from a weighted average of the target and flanker images; the weighting factors corresponded to the number of channels encompassed by the target and flanker band, respectively. The lower panel shows the results of the similarity weighting model, in which was calculated from the target image after weighting that image with an image representing the similarity between the target and flanker images.
Results of the second set of model simulations, in which cross-channel interactions were implemented at the time-interval analysis stage by weighting the transfer of time-interval information from a given NAP channel to the respective time-interval histogram with a factor that represented the amount of synchronous strobing activity in the other channels. The upper panel shows the results of the model in which the modified time-interval analysis was applied to the original NAP simulation, which had also been the front end of the previous set of simulations. The lower panel depicts the results of the model in which the NAP channels were shifted in time to compensate for the traveling-wave delay before the time-interval analysis was applied.
Article metrics loading...
Full text loading...