Designing Effective Speech Interfaces

Chapter 3

The Nature of Sound

“I'm sorry, you'll have to upgrade to first class if you want to use our ISDN line.”

“Look with thine ears.”

William Shakespeare's King Lear (Act 4, Scene 6)

Speech technology revolves around the physics of sound. Speech is a particular type of sound and therefore inherits the universal characteristics of sound in a scientific sense. Sound is defined as “a vibration in an elastic medium at a frequency and intensity that is capable of being heard by the human ear” (Isaacs et al., 1996). The “elastic medium” we are most concerned with is air. A vibration, such as that caused by the movement of human vocal chords that produces speech, travels through the air, and reaches our ears by physically striking our eardrums. Our ears then send an electronic signal to the brain. That message is perceived as sound. Sound may also be received by a microphone and then electronically processed in any of several ways related to the development of speech software, such as recording or broadcasting. This chapter covers the nature of sound, including its measurement and properties.