2.1.1 Human hearing system
Human’s hearing is an obligatory and sophisticated system which has high sensitivity, sharp frequency tuning, and wide dynamic range. A normal ear is able to distinguish and process acoustic signals varying in large magnitude and frequency range (from twenty to twenty thousand hertz). The ear can detect fine variations in pitch, loudness, and intonation.
The physical processing of acoustic information occurs in three groups of structures, commonly known as the outer, middle, and inner ears as described in Figure 2-1. Each of them has specific function and plays an important role in hearing the sound.
Figure 2-1 Human’s hearing system
The outer ear
The outer ear has three main components: the pinna, the auditory canal, and the eardrum.
The outer ear serves to collect the sound, assist in sound localization and function as a protective mechanism for the middle ear. The resonance of the canal favors the high pitches that are important to understand many consonants in the spoken word.
The middle ear
The middle ear is an air-filled space located within the temporal bone of the skull. It consists of the eardrum and the ossicles (malleus、incus、stapes), linking the membrane to the oval window of the cochlea. Sound pushes the eardrum, thus vibrating the eardrum at the same frequency of the sound wave. The middle ear function as a bridge between the air-borne pressure wave and the fluid-borne traveling of the cochlea.
The inner ear
The inner ear consists of cochlea, semicircular canals, and the auditory nerve. The inner ear is important for hearing and balance. The cochlea is the sensory end-organ of hearing which consists of fluid-filled membranous channels within a spiral canal that encircles a bony central core. Here the sound waves, transformed into mechanical energy by the middle ear, set the cochlea into motion in a manner consistent with their intensity and frequency. There are thousands of cells along the cochlea. The cells convert the mechanical motions into electrical signals and sent to the brain via auditory nerve.
2.1.2 Hearing Loss
There are two types of hearing loss: conductive hearing loss and sensorineural hearing loss. Conductive hearing loss happens when there is a problem conducting sound waves from the outer ear through eardrum and middle ear to the inner ear. The causes of the conductive hearing loss include earwax blocking the canal, middle ear inflections, or perforation of the eardrum. The conductive hearing loss can recover after some treatments. The earwax can be
removed, the eardrum can be reconstructed and the diseases in the middle ear usually can be cured.
Sensorineural hearing loss is the most common form of hearing loss. The sensorineural hearing loss results from damage to the inner ear. Hearing loss involves a multifaceted loss of hearing ability. The acoustic distortions faced by people with sensorineural hearing loss can summarize into four categories. [1]
Decreased audibility
Hearing-impaired people do not hear some sounds at all. People with a severe or pro-found hearing loss may not hear any speech sounds, unless they are shouted at close range.
People with a mild or moderate hearing loss are more likely to hear some sounds and not others. In particular, the softer phonemes, which are usually consonants, may not be heard.
We recognized the sound by noting which frequencies contain the most energy.
Hearing-impaired people have trouble understanding speech because essential parts of some phonemes are not audible. The hearing loss causes some frequencies components to be inaudible. Figure 2-2 is the audiogram of typical hearing-impaired person. The hearing-impaired person can’t hear clearly the sounds of the frequencies which are above 1000 Hz. For approximately 90% of hearing-impaired adults and for 75% of hearing-impaired children, the degree of impairment worsens from 500Hz to 4000Hz. Furthermore, the sound energy is dominated by low-frequency component.
Figure 2-2 Audiogram of typical hearing loss
In order to overcome these difficulties, a hearing aid needs to compensate the frequency-dependent audibility loss by amplifying the signal with various gains on each frequency. A filter bank is needed to decompose the signal into different frequency bands.
Then the hearing aid has the capability to provide different amount of gain in different frequency regions.
Decreased dynamic range
Dynamic range is a term used frequently in numerous fields to describe the ratio between the smallest and largest possible values of a changeable quantity. The human ear has a dynamic range of about 130 dB between the threshold of just hearing and threshold of uncomfortable loudness level. In the above argument, soft sounds can be made audible by amplifying the sounds. Unfortunately, it is not appropriate to amplify everything for the same amount of gain. Because the dynamic range of a hearing-impaired ear is less than the dynamic range of the normal ear.
Figure 2-3 Decreased dynamic range
Figure 2-3 shows the problem at one of the frequencies which has decreased dynamic range. For normal person, the dynamic range of sounds can fit between the threshold of hearing and threshold of uncomfortable loudness level. The normal people can hear sounds comfortably in a large dynamic range. For hearing-impaired person, the dynamic range reduces. The weak sounds below the threshold of hearing and the intense sound exceed threshold of uncomfortable loudness level. They can’t hear weak to moderate sounds. If we only amplify the sound to make the weak sounds audible, the moderate to intense sound may exceed the uncomfortable loudness level and it is unacceptable for hearing-impaired people.
For overcoming this difficult, hearing aids need to compress the input sounds. The hearing aid must provide more gains for soft sounds than intense sounds. The function reducing dynamic range of sounds is called dynamic range compression.
SNR loss
The sensorineural hearing loss can cause a hearing-impaired person to understand much
less than a normal-hearing person in the same environment, even the hearing-impaired person is wearing hearing aid. In the other way, the hearing-impaired person needs a better signal-to-noise ratio (SNR) than does a normal-hearing person. Loss of clarity results in a loss of ability to understand speech, especially in noise. The noise reduction of the hearing aid can improve the signal-to-noise ratio and thus, improve speech intelligibility in the noisy listening environment.
Frequency resolution loss
Another difficulty faced by people with sensorineural hearing loss is separating sounds of different frequencies. Different frequencies are represented most strongly at different place within the cochlea. When the cochlea gets damaged, it decreases the ability of the sensitivity to frequencies. If these frequencies are close enough, the cochlea will have a single broad region of activity rather than separate regions. The normal-hearing cochlea would separate the two broad regions. The impaired cochlea just recognizes a single broad region. For compensating the hearing loss, the speech enhancement function of the hearing aid improves some perceptual aspects of speech for the human listener.