Sound is produced when something vibrates. These vibrations produce local variations in air pressure due to a squeezing of the air in immediate contact with the object in vibration, and this motion is then transmitted to adjacent air molecules. With increasing distance from the source, there is a loss of energy and the sound decays.

Within a well defined range, the ear is able to detect these changes from normal atmospheric pressure (the reference at all times) even when they are minute, that is above 0.00002 N/rn2 at 1000 Hz, the threshold of human hearing.

The dominant parameters of sound are frequency, wavelength and energy level (Figure A).

Frequency is the number of complete vibrations which occur in each second. Subjectively, high frequencies are high pitched sounds (treble) and low frequencies are low pitched sounds (bass).

Wavelength is the distance occupied by one complete vibration. Low pitched sounds have longer wavelengths than those of high pitch.

Energy level is a consequence of the amplitude of sound vibrations. Loud sounds produce greater amplitudes (bigger pressure changes) at normal atmospheric pressure than quiet sounds. These amplitudes are indicative of the energy levels, measured in decibels (dB); the range of interest is from 0 dB (threshold of hearing) to around 120 dB (threshold of pain) (Figure 22.3a).

The human ear is able to detect sound frequencies in the range from 20 to 20000 vibrations per second (hertz, symbol Hz). The corresponding wavelengths in air are from 17m down to 17mm (Figure 22.3b).

With older persons, it is the discrimination of high frequencies which is impaired; the low frequency response remains virtually unchanged. Frequencies above and below the human range are detected by some animals, for example dogs.

Even though people are able to detect sounds within the broad frequency range of 20 to 20000 Hz, their reception of them is not uniform. Because of the physiology of the ear, the corresponding response is not linear. The ear is more sensitive to high frequency sounds than to lower ones, so that when measuring sounds with equipment, such as a sound level meter (Figure 22.4), corrections must be built in to mimic the ear's response. The readings thus obtained can then provide a reliable guide to human subjective reaction to sounds of different mixes of frequencies. Sound readings, so modified, are measured in dBA, the letter A signifying the application of this selective frequency modification. The units dB and dBA should not be confused, or used together in calculations; consistent units are essential (Figure 22.5).

Some important approximate relationships between changes in sound pressure level and the corresponding changes in apparent loudness are as shown in Figure 22.6.