Sound Stimulus

Our entire scientific knowledge of how hearing works derives from the stimuli used in auditory experiments. Hearing scientists have devoted [Page 901]much effort and insight into designing the sounds for their experiments so that each is just right for its purpose. This has become much easier in recent years, as the near-universal use of computers means that almost any stimulus that can be thought of can be used, provided only that a computer program can be written to make it. The result is a true revolution in experimental method: The days when specialized electrical hardware was needed for auditory experiments—and which only a few laboratories had—are long past.

The three classic stimuli of auditory science discussed in this entry are pure tones, noises, and clicks. Their popularity comes from the ease with which they can be generated, be it electrically or computationally, but also from their simple properties: They are the building blocks underlying all other stimuli. Indeed, it is rare to encounter an experiment that does not use at least one of them.

The pure tone is the simplest. Its only physical parameters are frequency, measured in Hertz; power, measured in decibels; and duration, measured in seconds. The first two parameters are closely related to the perceptions of pitch and loudness: A pure tone has fixed, unchanging pitch and loudness (indeed, they are often taken as references when measuring the perceptions of more-complex stimuli). Pitch is particularly important: A pure tone has only a single frequency component and evokes only a single pitch, but combinations of pure tones may have one or more pitches. The search for the rules underlying which particular pitch results from a pure tone or a set of pure tones has driven much of the work on theories of pitch perception. A “white” noise, in contrast, contains every frequency: It is often made by adding together every possible pure tone, each with a random level and phase (the name “white” is by analogy with white light, which contains all visible frequencies). Its spectrum is broad and evokes no pitch. Countless experiments have used white noises as masking sounds that measure the detection threshold of some other sound in the noise. Countless others have shaped the spectrum of the noise in some way to give maskers with special characteristics, such as by giving proportionally less weighting to higher frequencies than lower (e.g., “pink” noise) or by filtering it to include only certain frequencies (e.g., “notched” noise).

When used in experiments, tones and noises can last from a few milliseconds (ms) to a few minutes. In contrast, the shortest stimulus that can be made is a click, usually with a duration less than 0.1 ms. But a single click is rarely used in experiments, as it is just too short. Instead, experiments usually use sets of them, from two-click pairs to multiclick trains lasting a second or more.

The definite pitch and rich timbre produced by musical instruments are due to the large number of pure tones they generate. The frequencies of these “harmonics” are often multiples of the lowest harmonic (i.e., if the fundamental frequency measured in hertz [Hz] is at f Hz, then the rest are at 2f, 3f, 4f, 5f, and so on), but the intensity of each one will vary across instruments. This pattern helps to determine the timbre of the sound and distinguishes one instrument from another (e.g., oboes give very little power at the fundamental frequency f, while clarinets give very little power at the even-numbered harmonics 2f, 4f, etc.).

...