Sound consists of small variations in air pressure that occur very rapidly one after another. These variations are caused by actions of the vocal organs. In the case of voiced sounds, for example, the vibrating vocal cords chop up the stream of air so that pulses of relatively high pressure alternate with moments of lower pressure, whereas in fricative sounds the airstream is forced through a narrow gap so that it becomes turbulent with irregularly occurring peaks of pressure[14].
The speech signal is a signal varying slowly over time in the sense that, when examined over a sufficiently short period of time (between 5 and 100 msec), its characteristics are fairly stationary; however, over long periods of time the signal characteristics change to reflect the different speech sounds being spoken. An illustration of this effect is given at the top of Figure 2.3, which shows the speech waveform corresponding to the Japanese sentence, ``Hai, onegaishimasu'' (Would you please do me a favor?) as spoken by a native male speaker. Largely three parts of the signal can be distinguished by contrasting the absence or presence of vibration, which corresponds to `` Hai'', ``onegai'', and ``shimasu''.
Although there are several ways of classifying signals in the waveform, the most conventional and straightforward is via the three-state of the speech-production source - vocal cords. The states are (1) silence, where no speech is produced; (2) unvoiced, in which the vocal cords are not vibrating, so the resulting speech waveform is aperiodic or random in nature; and (3) voiced, in which the vocal cords are tensed and therefore vibrate periodically when air flows from the lungs, so the resulting speech waveform is quasi-periodic.
Although we can learn quite a lot by a visual inspection of a speech waveform, it is impossible to analyze precisely the individual pronunciation sounds from waveforms because the waveforms of many sounds are much too similar to one another to be analyzable by eye. However, even without much experience, it is possible to see the difference among the three-state of vocal cords through the waveform.
