Sketch of the apparatus for localization studies and exemplification of binaural cue manipulation in experiment 3. In experiments 1 and 2 sounds were played via loudspeakers while in experiment 3 spatialized stimuli were fed directly into the CI processors. The figure shows an example for binaural cue trading used in experiment 3: While the prerecorded stimulus from, e.g., , was played to the subject it was run through an all-pass filter that incorporated the ITD changes necessary to represent ITDs from . A shift in localization to the ITD direction would suggest a dominance of ITDs whereas a localized direction halfway between the ITD and ILD direction shows equal weighting of both cues. Likewise, conditions with manipulated ILDs were run.
Free-field localization results of subject BW in experiment 1. Results are presented as medians and quartiles. Panel A shows results for each single CI for pulsed WBN, whereas panels B-D show results for various stimuli with bilateral CIs (see inlets and Table I).
Free-field localization results of subject DF in experiment 1. Panel C shows additional results for LPN limited to . Results are plotted as in Fig. 2, stimuli are described in Table I.
Free-field localization results of experiment 2 of subject DF with elevated CI processors for pulsed wide-band noise (Table I). In the condition “Together” , processors were placed above the head next to each other, whereas in condition “Plate” a cardboard was placed between the two processors that were still next to each other. The elevated processors were spaced at ear distance in the condition “Distanced” (o), but there was no plate between them. Medians are connected by lines and error bars show quartiles.
Comparison of free-field localization and virtual localization of subject BW (experiment 3). The stimulus was pulsed wide-band noise (Table I).
Localization results of subject BW with virtual sound presentation without binaural cue disparities for different stimuli (Inlet and Table I).
Localization results of experiment 3 with binaural cue disparity (subject BW). A binaural recording of pulsed wide-band noise taken at was manipulated so that one binaural cue (ILDs or ITDs) stemmed from a different direction away from . Localization followed this offset only for ILDs.
Average directional shift observed in experiment 3 with subject BW for different disparities between ILD and ITD cues. Panels A–C give results for pulsed wide-band noise, high-pass noise with slow slopes and pulsed high-pass noise, respectively (cf. Table I). Data are averages over several originating directions.
ILDs computed from HRTFs measured on Med-El Combi processors. ILDs were derived from the level of HRTF-filtered narrow-band noise signals. The frequencies and the bandwidth of the noise corresponded to the log-spaced channel mapping in used by subject BW. Tickmarks on the frequency axis correspond to CI channels. Levels were normalized to be zero for frontal sound incidence to correct for amplification offsets in both CI processors. ILDs were averaged for sound from the right and the left side.
Overview of stimuli.
Summary statistics to free-field localization results of subject BW in experiment 1 for different stimulus conditions, cf. Fig. 2.
Summary statistics for localization results of subject DF in free-field experiments 1 and 2, cf. Figs. 3 and 4. Statistics as in Table II.
Summary statistics for localization results of subject BW in experiment 3, cf. Figs. 5 and 6: localization in the free field of pulsed wide-band noise and of virtual stimuli with consistent binaural cues. Virtual localization results were combined from trials without disparities from a session with ILD and one with ITD disparities. Statistics as in Table II.
Article metrics loading...
Full text loading...