gms | German Medical Science

Audiological experiments or clinical tests are usually performed either by presenting sounds over loudspeakers or headphones. Especially in the context of hearing aid evaluation, it may sometimes be useful to use another presentation mode, where the signal is first presented to a dummy head over loudspeakers, and the ear canal signals are transmitted to headphones binaurally. This mode allows, for example, blind comparison of hearing aids and exposure to stimuli that would be unethical for the subject directly, such as wind over an extended time [1].

The goal of this study thus is to compare the outcome of a set of standard audiological tests for sound presentation through a loudspeaker (Genelec 8351 A), audiological headphones (Sennheiser HDA200), and binaural transmission of binaural signals captured with a KEMAR. For the binaural transmission, three different means of headphone equalization were compared, which should result in equivalences of the stimuli to loudspeaker presentation in terms of hearing threshold (RETSPL) [2], loudness perception (Gf-Gc) [2], or ear drum sound pressure level on KEMAR (flat response on KEMAR [3]). For these presentation conditions, loudness function scaling (using the ACALOS procedure) and speech intelligibility in quiet and noise (OLSA) measurements were conducted in normal-hearing subjects, in addition to transfer function analyses of all presentation modes.

Results showed equivalent outcomes for loudspeaker and binaurally transmitted sound presentations when Gf-Gc or RETSPL-based headphone equalization is use, although the sound pressure at eardrum is different. Differences with binaural transmission between RETSPL/Gf-Gc -based and the flat headphone equalization were noted for speech intelligibility in quiet, consistent with the level offset between conditions. The outcomes with headphone presentation showed some unexpected results, which could be explained based on commonly applied, non-standardized level offsets for the HDA200 and experimental conditions applied in the present study.

Figure 1 [Fig. 1]