Wind noise in hearing aids with directional and omnidirectional microphones: Polar characteristics of custom-made hearing aids

Abstract

The purpose of this study was to examine the characteristics of wind noise at the output of in-the-ear, in-the-canal, and completely-in-the-canal hearing aids. The hearing aids were programed to have linear amplification with matching flat frequency responses for directional (DIR) and omnidirectional (OMNI) microphones. The microphone output was then recorded in a quiet wind tunnel when the Knowles electronic manikin for acoustic research (KEMAR) head was turned from 0 degrees to 360 degrees . The overall, 125, 500, and 2000 Hz one-third octave band flow noise levels were calculated and plotted in polar patterns. Correlation coefficients, average differences, and level differences between DIR and OMNI were also calculated. Flow noise levels were the highest when KEMAR was facing the direction of the flow and angles between 190 degrees and 250 degrees . The noise levels were the lowest when the hearing aids were facing the direction of the flow. The polar patterns of DIR and OMNI had similar shapes and DIR generally had higher levels than OMNI. DIR, however, could have lower levels than OMNI in some angles because of its capability to reduce noise in the far field. Comparisons of polar characteristics with behind-the-ear hearing aids, and clinical and engineering design applications of current results are discussed.

Wind noise at microphones within and across hearing aids at wind speeds below and above microphone saturation

“…Some previous studies of wind-noise at the hearing-aid input used wind speeds of 2.1 and/or 5 m/s (Dillon et al, 1999(Dillon et al, , 2000Beard and Nepomuceno, 2001;Thompson and Dillon, 2002), while Chung et al (2009Chung et al ( , 2010 mainly used speeds of 4.5, 9.0 and 13.5 m/s, and the current study used speeds of 3, 6, and 12 m/s. Therefore, comparisons with previous studies cannot be made at identical wind speeds.…”

“…Another difference across studies is the measurement duration, which where specified varies from 9 s (current study) to 16 s (Grenner et al, 2000) to 30 s (Chung et al, 2009(Chung et al, , 2010. For long-term-average measurements, long durations could better capture any very slow variations of turbulence generated by the head or inherent in a non-laminar, generated wind stream.…”

“…It is well established that wind noise can be greater at the output of a directional microphone than an omni-directional microphone (Beard and Nepomuceno, 2001; Thompson and Dillon, 2002;Chung et al, 2009Chung et al, , 2010. The long-term-average, wind-noise spectrum can vary with wind azimuth (horizontal plane), style of hearing aid, and/or hearing-aid shell design for a particular hearing-aid style (Dillon et al, 1999(Dillon et al, , 2000Grenner et al, 2000;Beard and Nepomuceno, 2001;Thompson and Dillon, 2002;Chung et al, 2009Chung et al, , 2010. Some studies have also shown that wind-noise levels increase with wind speed (Beard and Nepomuceno, 2001;Chung et al, 2009Chung et al, , 2010.…”

“…Hearing aids often switch from a directional microphone to the front omni-directional microphone in wind (Chung, 2004), since wind-noise levels are typically lower with an omni-directional microphone (Beard and Nepomuceno, 2001;Chung et al, 2009Chung et al, , 2010. However, it is not clear whether further reductions in wind-noise levels can be achieved by switching between the front and rear omni-directional microphones, and whether this depends on the wind speed, wind azimuth, or hearing-aid shell design.…”

“…The author is aware of two studies that used higher wind speeds, where hearing aids were fitted to a Knowles Electronics Manikin for Acoustic Research (KEMAR) (Burkhard and Sachs, 1975) and wind-noise levels were measured at the ear drum for wind speeds at multiples of 4.5 m/s up to 13.5 (Chung et al, 2010) or 22.5 m/s (Chung et al, 2009). The fitted, linear gain was used to derive wind-noise input levels for the 125, 500, and 2000 Hz one-third-octave bands and the total wideband level across all bands.…”