Wind noise in hearing aids with directional and omnidirectional microphones: Polar characteristics of behind-the-ear hearing aids

Abstract

Wind noise can be a significant problem for hearing instrument users. This study examined the polar characteristics of flow noise at outputs of two behind-the-ear digital hearing aids, and a microphone mounted on the surface of a cylinder at flow velocities ranging from a gentle breeze (4.5 m/s) to a strong gale (22.5 m/s) . The hearing aids were programed in an anechoic chamber, and tested in a quiet wind tunnel for flow noise recordings. Flow noise levels were estimated by normalizing the overall gain of the hearing aids to 0 dB. The results indicated that the two hearing aids had similar flow noise characteristics: The noise level was generally the lowest when the microphone faced upstream, higher when the microphone faced downstream, and the highest for frontal and rearward incidence angles. Directional microphones often generated higher flow noise level than omnidirectional microphones but they could reduce far-field background noise, resulting in a lower ambient noise level than omnidirectional microphones. Data for the academic microphone- on-cylinder configuration suggested that both turbulence and flow impingement might have contributed to the generation of flow noise in the hearing aids. Clinical and engineering design applications are discussed.

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

“…The velocity of 4.5 m/s was the average wind speed recorded in cities in United States on a typical non-windy day ͑National Climate Data Center, 2006͒ and 13.5 m/s is the threshold for the National Weather Services to issue a wind advisory. Higher flow velocities are not included in this study because Chung et al ͑2009͒ showed that the limiters in hearing aids limited the maximum flow noise levels recorded at the hearing aid output. At 13.5 m/s, the flow noise levels, especially for the DIR mode, reached the maximum output levels of the hearing aids at almost all head angles, and the polar patterns were nearly circular in shape.…”

“…The ambient noise in the wind tunnel was measured using an ER-11 microphone shielded with an ellipsoidal windscreen ͑B&K UA 0781͒. The one-third octave band levels of the ambient noise in the wind tunnel are reported in Chung et al, 2009.…”

“…In addition, Chung et al ͑2009͒ reported that directional microphones generated higher noise levels than omnidirectional microphones at the hearing aid output for most angles at higher flow velocities ͑i.e., 9.0 and 13.5 m/s͒. At 4.5 m/s, however, directional microphones exhibited lower noise levels than omnidirectional microphones when the hearing aids were facing the direction of the flow.…”

“…The sensitivity of the directional microphone to sounds from different directions can be varied by changing the ratio of the internal and external delays to yield cardioid, dipole, hypercardioid, supercardioid, or other directivity patterns. As the most effective subtraction occurs when the amplitude of the signals received from the front and rear microphones are similar, the two omnidirectional microphones used to form the directional microphone need to have matching phases and sensitivity across frequency regions ͑Valente, 1999; Ricketts, 2001; One of the disadvantages of directional microphones is that they generate higher noise levels than omnidirectional microphones in wind due to the incoherent sound source effect and the proximity effect ͑Beard and Nepomuceno, 2001; Thompson and Dillon, 2002;Chung et al, 2009͒. When wind flows pass an obstacle ͑e.g., the head or the pinna͒, eddies are generated downstream of the obstacle. Small eddies are also formed locally over each microphone, creating large random pressure changes on the diaphragm of each microphone, thus yielding incoherent microphone signals.…”

“…Notice that the term "flow" is used here to refer to the air flow generated in wind tunnels instead of "wind," which usually refers to naturally occurring atmospheric air motion. Chung et al ͑2009͒ showed that the highest flow noise levels were measured when the flow was coming from the front or back for both directional and omnidirectional microphones. Relatively low noise levels were measured when the hearing aid was facing the direction of the flow or facing downstream.…”