|
|
Hearing Aid Amplification and Related Problems Hearing aids amplify sounds. A good hearing aid amplifies sounds according to the dynamic range of hearing still available to the patient, and it reproduces sounds that are most receptive to the user, with a minimum amount of distortion and noise. It must have solutions for a number of problems related to amplification, as described below. A. Dynamic-Range Problem Conventional hearing aids with linear amplification do not provide proper solutions for the reduced dynamic range in sensorineural hearing losses. For patients with such losses, although the sensitivity to soft sounds is reduced, the loudness sensation for intense sounds remains nearly normal. As a result, the same gain that allows soft sounds to be heard will make intense sounds too loud. Therefore, the amplification scheme of the hearing aids should be able to mimic the function of outer hair cells in healthy cochlea, that is, like an automatic gain control (AGC) system that provides greater gains to softer sounds, and smaller gains to stronger sounds. When the hearing loss in a frequency region is so profound that the residual dynamic range becomes too small, even an AGC system would fail to transmit information from that region. Even worse, research has shown that amplification in such a region can actually hurt speech understanding. In this extreme case of diminished dynamic range, if the information carried in that frequency region is crucial, it might be worth transporting that information to a different frequency region where the hearing is either intact, or with a loss that can be corrected by amplification. B. Distortion Problem When a system responds in any nonlinear way, it generates distortions. In hearing aids, distortions can result from many sources, due to limited driving power, the variable gain from the wide-dynamic-range-compression (WDRC) scheme, and so on. B1. Output Saturation Like all instruments, hearing aids can deliver only so much output power. When the hearing aid is operating near the saturation output, as in cases of severe loss, the strong peaks of the output waveform will be ¡®clipped¡¯. Distortions of the signal waveform due to peak clipping can greatly deteriorate the sound quality. B2. Distortion Resulted From WDRC Because WDRC provides variable gains depending on input level (i.e., decreasing gain with increasing stimulus level), it is a nonlinear process by definition, and thus its outcome is distorted. However, this distortion is intentional, for the purpose of overcoming the reduced dynamic range. For the most part, distortions directly resulted from WDRC are perceptually tolerable. WDRC in particular distorts the original signal by smearing the spectral contrast. It adds greater gains to frequency bands where the sound levels are low, and smaller gains to frequency bands where the sound level is high. As a result, the contrasts between spectral peaks and valleys are reduced. The spectral smearing could deteriorate the quality and clarity of the original signal. This is particularly true when the sound is intense enough that the spectral valleys are above the hearing threshold. Therefore, it is important to neutralize the effect of spectral smearing for relatively intense sounds. C. Feedback Problem Feedback problems happen when the sound output of the hearing-aid receiver goes back to the microphone through a loop, called the feedback path, and gets amplified again and again. When the overall gain exceeds a certain value, the system becomes unstable and will oscillate. The oscillation will be heard as an extremely annoying whistle. It is a sever form of distortion. This distortion is also a serious problem for the function of hearing aid, as it is often the practical reason that the gain must be held below a certain limit. Therefore, to fully realize the amplifying capability of the hearing aid, one must effectively manage feedback problems. D. Noise Problem Hearing aid designers need to deal with two kinds of noise, one near hearing threshold¡ªthe ¡®low¡¯ noise, and the other above threshold¡ªthe ¡®noisy¡¯ noise. The low noise includes circuit noises (such as microphone noise) and low-level ambient noises. For people with normal hearing, most of these noises are inaudible. For hearing aid users, however, strong amplification can often bring such noises above threshold. Then, while normal-hearing people are enjoying a quiet moment, hearing aid users at the same place will still be bombarded by the constant and annoying noises, depriving them of a quiet time. To avoid this, it is desirable for hearing aids to restrict gains applied to sounds that are inaudible to normal ears. The noisy noise includes all noises that are audible to normal ears, such as noises from cars, airplanes, all other voices that interfere your conversation at a party, and so on. This kind of noise is the biggest enemy of speech communication, for people with normal or impaired hearing alike. For hearing-impaired listeners, however, such noises do more damage, because impaired ears need a higher signal-to-noise ratio to achieve the same intelligibility as the normal ears. Furthermore, the severer the hearing loss, the more is demanded on the signal-to-noise ratio. Once the noise is audible, further amplification still increases the overall output level, but it can no longer improve signal-to-noise ratio. To improve speech intelligibility, the hearing aid must be able to suppress noises while leaving the speech signal unharmed. Other than affecting speech intelligibility, noisy noises, particularly when they are amplified, also cause listening discomfort to the hearing aid users. Note that many hearing-impaired ears have nearly normal loudness sensation for intense sounds. As a result, an intense noise that sounds loud to a normal ear can sound very loud to an aided ear. Therefore, just for the purpose of reducing listening fatigue and improving listening comfort, it would be desirable for hearing aids to be able to attenuate the noise level, even when the speech is absent. |