Effect of sound image spatial splitting caused by interaural differences on speech intelligibility

This study aimed to clarify the effect of sound image splitting perception by Interaural Time Difference (ITD) and Interaural Level Difference (ILD) on speech intelligibility. Thus, different ITDs or ILDs were applied to target (speech) and masker (pink noise) sounds, and mora intelligibility was measured. The sound stimuli were prepared such that either the sum of target and masker ITD values equaled 0, 0.1, 0.2, 0.4, 0.6, 0.8 ms, or the sum of target and masker ILD values equaled 0, 5, 10, 15, 20, 25, 30 dB. The results of the ITDs-added experiment showed that the mean mora intelligibility improved from around 0.2 ms for ITDs, consistent with the detection limit of spatial splitting of sound images. In the ILDs-added experiments, the mean mora intelligibility improved even when ILDs were smaller than the detection limit of spatial splitting of sound images, owing to reduced unilateral masking. Therefore, the ILDs-added experiment was conducted under conditions where unilateral masking was not reduced by adding ILDs. As a result, the mean mora intelligibility improved at ILDs around 10 dB and worsened above 15 dB. These two values are respectively near the detection limit and saturation point of spatial splitting of sound images.