Estimated variability of real-ear insertion response (REIR) due to loudspeaker type and placement

The real-ear insertion response (REIR) of a hearing aid is estimated as the difference between the aided response and the unaided response in the ear canal. Changes in the position of the loudspeaker relative to the head, between the two measurements, may reduce the accuracy of the estimate. The spatial variability of the sound field at distances close to the loudspeaker is less for a 'flat-panel' loudspeaker than for a conventional cone loudspeaker; the panel might thus lead to a reduced influence of loudspeaker position. To assess this, we measured the real-ear unaided response (REUR) as each of three loudspeakers (two cone type and one panel) was moved in a three-dimensional space centred at either 0degrees or 45degrees azimuth, at a distance of 50 cm from a KEMAR manikin. Contrary to our expectation, the variability of the REUR was larger for the panel than for the cone loudspeakers. The REUR varied less with position for the 0degrees than for the 45degrees azimuth. The variability of the REUR decreased with increasing distance of the loudspeaker from KEMAR. We tentatively suggest that loudspeaker-to-client distances of 40-50 cm should be used and that a 0degrees azimuth is preferable.