How Much Do We Gain by Gain Control in Hearing Aids?

Abstract

This paper considers three rationales for the use of automatic gain control (AGC) in hearing aids, and particularly multi-channel AGC. One rationale is to restore loudness relations among the acoustical elements of speech to what they would be for a normal ear. At present, there seems to be little evidence to support the idea that such loudness relations are critical for speech intelligibility. A second rationale is to reduce the effects of narrowband interfering sounds. This may be applicable in some restricted situations, but not in many others. The third rationale, the one considered here to be most important, is to ensure that all of the important elements in speech are comfortably audible. This can be achieved using two forms of AGC: slow-acting AGC operating on the whole speech signal, to compensate for variations in overall sound level from one situation to another; and fast-acting (syllabic) compression acting independently in a number of bands, to compensate for differences in level of acoustical elements within speech. To reduce spectral and temporal distortions introduced by the syllabic compression, the number of bands should be small (probably two or three) and the amount of compression should be the minimum required to ensure that all of the important elements in speech are audible. It is concluded that gain control can be useful for increasing user comfort and for ensuring the audibility of speech sounds. It is not likely to restore impaired hearing to normal for most patients, but may usefully be employed in conjunction with other forms of signal processing.

Key words:

• automatic gain control
• compression
• loudness recruitment