Speech understanding frequencies

Initially, hearing damage occurs at the higher frequencies. As the unprotected exposure continues, the damage and hearing loss spread to the lower frequencies. Eventually, most of the frequencies that are essential to understanding speech (500 Hz to approximately 3000 Hz) may be affected. 

Noise can also interfere with speech communication. One worker may not understand or may misunderstand what another says. A loss in the high frequency range will reduce hearing of some soft speech sounds, like s, sh, and th. The hearing loss may cause failure to hear certain words or create speech misinterpretations. 

To understand the basic concepts of modulation, we first review time- and frequency-domain representations of signals. The information present in m(t) can be completely specified by a complex function of speech amplitude vs. frequency, M f), obtained using the Fourier transform. Since M f) contains aU of the information in m t), it is possible to go back and forth between m t) and M f) using the forward and inverse Fourier transforms. 

An important effect of noise emissions in a man-machine-environment-system is the disturbance in verbal communication. When full understanding of sentences is required, the interference sound level for normal messages should be no higher than 10 dB (A) below the speech sound level. For unusual or difficult messages this difference should be increased to 20 dB (A). The maximum distance between speaker and listener which still ensures satisfactory communication can be derived from the speech interference level (SIL). The SIL is the arithmetic mean of the octave sound level in the four octaves with the mean frequencies of 0.5/1/2/4 kHz. Table 6.21 shows the maximum distance between speaker and listener for various SILs. 

High frequency hearing loss is a type of hearing loss which involves a person s inability to hear clearly speeches containing high frequency components. As an example, a person with such inability will face problem to understand... 

In compressing high frequency components, speech intelligibility must be preserved. One of the speech features that need to be preserved is the formant. Formants are the resonance frequencies of the vocal tract that shape the speech source signal [5 8]. The first 2 formants called FI and F2 are the two most important formants in understanding the speech. For a voiced speech, the ratio of FI to F2 is approximately equal even if the pitch of the speech is different. 

---