6. SOUND PRODUCTION AND DETECTION
155
velopment of the sound spectrograph. This instrument is now well known
in many areas of acoustic research, including virtually all phases of bioacoustics. In brief, the sound spectrograph takes a sample of the sound
and produces a graph of frequency against time. Other displays such as
frequency against relative intensity are also possible. Many examples of
its use in bioacoustics can be found in Lanyon and Tavolga (1960) and
Busnel (1963). Spectrographic displays of various fish sounds are shown
in practically all current publications in this field.
As applied to underwater bioacoustics, the use of the sound spectrograph has been carefully evaluated by Watkins (1967). The capabilities
and limitations of this instrument are often not fully understood by biologists. The display of apparent harmonics, for example, can result from
repetition of pulses, and other complexities can be introduced if the sound
consists of short, repeated bursts of pulses, with each pulse consisting of
a brief tone or complex of tones. Often the so-called fundamental frequency is actually a pulse repetition rate. In addition, the verbal descriptions of bioacoustic phenomena have had to be standardized, and
attempts at this standardization resulted in glossaries compiled by
Broughton (1963) and Bondesen and Davis (1966), although only a
minority of the terms they listed apply to the sounds of fishes.
A serious source of difficulty in making original recordings of animal
sounds is the effect of the reverberations, reflections, absorptions, and
other acoustic phenomena in the environment. Field recordings are always plagued by such problems, including the presence of background
noise. Such problems are particularly troublesome in underwater recordings. The spectral and other characteristics of sea noise and man-made
noise have been summarized by Wenz ( 1964), but the acoustic properties
of the ocean bottom, surface, suspended particles, and air bubbles are
still under intensive study ( Albers, 1965; Richardson, 1957). The problems of recording sounds of captive fishes are even more complex, and
the acoustic field generated in an aquarium tank, for example, virtually
defies analysis (Parvulescu, 1964, 1967). Any spectral analysis of an underwater sound, therefore, must be cautiously interpreted, especially if
the acoustic conditions are not controlled and not specified (Schneider,
1967; Tavolga, 1965).
These difficulties are especially evident in attempts to identify unknown sound sources. Sounds of marine animals are often recorded without any visual confirmation of the source of the sound, and it is tempting
to make comparisons between such field recordings and recordings of
known species in captivity. This problem was discussed in detail by
Tavolga ( 1965).
Tables of data on pitch and other characteristics of fish sounds are
155
velopment of the sound spectrograph. This instrument is now well known
in many areas of acoustic research, including virtually all phases of bioacoustics. In brief, the sound spectrograph takes a sample of the sound
and produces a graph of frequency against time. Other displays such as
frequency against relative intensity are also possible. Many examples of
its use in bioacoustics can be found in Lanyon and Tavolga (1960) and
Busnel (1963). Spectrographic displays of various fish sounds are shown
in practically all current publications in this field.
As applied to underwater bioacoustics, the use of the sound spectrograph has been carefully evaluated by Watkins (1967). The capabilities
and limitations of this instrument are often not fully understood by biologists. The display of apparent harmonics, for example, can result from
repetition of pulses, and other complexities can be introduced if the sound
consists of short, repeated bursts of pulses, with each pulse consisting of
a brief tone or complex of tones. Often the so-called fundamental frequency is actually a pulse repetition rate. In addition, the verbal descriptions of bioacoustic phenomena have had to be standardized, and
attempts at this standardization resulted in glossaries compiled by
Broughton (1963) and Bondesen and Davis (1966), although only a
minority of the terms they listed apply to the sounds of fishes.
A serious source of difficulty in making original recordings of animal
sounds is the effect of the reverberations, reflections, absorptions, and
other acoustic phenomena in the environment. Field recordings are always plagued by such problems, including the presence of background
noise. Such problems are particularly troublesome in underwater recordings. The spectral and other characteristics of sea noise and man-made
noise have been summarized by Wenz ( 1964), but the acoustic properties
of the ocean bottom, surface, suspended particles, and air bubbles are
still under intensive study ( Albers, 1965; Richardson, 1957). The problems of recording sounds of captive fishes are even more complex, and
the acoustic field generated in an aquarium tank, for example, virtually
defies analysis (Parvulescu, 1964, 1967). Any spectral analysis of an underwater sound, therefore, must be cautiously interpreted, especially if
the acoustic conditions are not controlled and not specified (Schneider,
1967; Tavolga, 1965).
These difficulties are especially evident in attempts to identify unknown sound sources. Sounds of marine animals are often recorded without any visual confirmation of the source of the sound, and it is tempting
to make comparisons between such field recordings and recordings of
known species in captivity. This problem was discussed in detail by
Tavolga ( 1965).
Tables of data on pitch and other characteristics of fish sounds are