To synthesize binaural room transfer functions (BRTFs) with minimal measurement burden, we proposed a method of adding reverberation characteristics obtained at a certain point in the room to the listener's HRTFs. However, the reverberation characteristics physically differ depending on the listener’s position in the room. To investigate whether the difference of those characteristics could be subjectively distinguished, we conducted two experiments. First, a listening experiment was conducted using stimuli of pinknoise, instruments, and speech that were convoluted with the impulse responses measured at four locations in a room from a certain source position. In the experiment, one location was used as the reference, and participants were asked to score the auditory differences using the MUSHRA method. Second, a similar experiment was conducted using synthesized impulse responses that differed only in the reverberation components. Namely, the reverberation components were extracted from the impulse responses of each of the four measurement locations and connected to the direct and first reflected sound of the reference. From these experiments, it was found that the difference in measurement position was difficult to be distinguished when only reverberation components were replaced. Identifying differences in measurement position was most difficult with speech signals.
Kotsugai et al. (Wed,) studied this question.