A quick primer about Sound and Marine Mammals

Sound is likely the most important sensory medium for fully aquatic marine mammals. Sound travels approximately 5x as fast as in air, and due to water's high elasticity, can travel further distances underwater (particularly at low frequencies).

Sound is a vibrational energy. The speed of vibration corresponds to a sound's pitch, or frequency. The volume of a sound of a particular frequency is a function of the strength of that vibration, which to a receiver translates into pressure. A parcel of sound rarely contains just one frequency, but a range of frequencies, each with a corresponding amplitude. Because it is difficult to work in terms of pressure, acousticians use a log scale that relates the pressure of the sound of interest to a standard pressure. Underwater, that standard pressure is 1 µPa (micro-pascal) measured at a distance of 1 m away from the sound's source. The units of the log scale are decibels, or dB. Thus decibels, as they relate to amplitude, are meaningless unless you include to which reference they refer. A complete assessment of volume should always be of the form "X dB re 1 µPa at 1 m".

There are several ways to characterize a sound. The three most popular are the spectrum (frequency versus amplitude), a waveform (time versus amplitude), and a spectrogram (time versus frequency, with amplitude coded as a grey scale within the graph).

You will be familiar with the concept that a sound decreases in amplitude the further away from its source. This is due to wavefront spreading, which underwater can happen in two ways: spherically (up to a range equivalent to the depth of the water for that area), or cylindrically (for ranges beyond a range equivalent to the depth of the water for that area). Each type of spreading results in a loss in energy, termed TL (transmission loss), measured in dB, that is mathematically quantifiable:

TLspherical = 20 log R (when R < depth of water)

TLcylindrical = 15 log R (when R > depth of water)

Thus the received level of a sound (RL) can be calculated in terms of the original source level (SL) and TL.

RL = SL - TL

The ocean has two types of sound: specific signals of interest to the researcher, and non-specific signals. The latter category can be quantified as ambient noise, which is variable between sites depending on localized sources such as shipping traffic, and biologics. To detect a deliberate signal, it must be detectable above ambient noise. In other words, ambient noise acts to mask a signal. (Liken this to trying to hear someone's conversation on the other side of a interstate highway&emdash;if there were no cars generating noise, the conversation would be easily detectable. However, the noise of the cars&emdash;ambient noise&emdash;masks the conversation). Noise most effectively masks a signal when the frequencies of the two sounds closely match. When in the same frequency domain, a signal becomes non-detectable if its amplitude is the same or less than that of ambient noise (referred to as a Signal to Noise Ratio of unity).

The issue of detection against a noise background is further complicated by the nature of the receiving array (i.e., an organism's hearing apparatus). In situations of potentially masking noise (of a frequency that corresponds to the signal of interest), most arrays can still not detect the signal if its only a small amplitude above the noise level. The level a signal has to be above ambient noise in order to be detectable is referred to as a Critical Ratio. Critical ratios are species specific.

Finally, in order to detect a sound, the receiving array has to be sensitive to that specific frequency: most animals have an optimum frequency sensitivity, outside of which the threshold of detection increases (in other words, the sound has to be louder to be audible). Outside of a particular range of frequencies, sounds may not be audible to a species at all. Sound sensitivity for a species is characterized by an audiogram, that charts frequency by the volume (dB) of a frequency required to be heard (the threshold).

Other primers:

the following links will download .pdf files for specific acoustics topics as handled by the sound analysis program, Canary. The below text and downloads are taken from the program's website, with permission. If you need the .pdf viewer, visit here.

Digital Representation of Sound provides a brief explanation of how sound is represented digitally. An understanding of the basic principles introduced here (such as digital sampling, sampling rate, and sample size) will be helpful in using Canary or any other sound analysis software.

A Biologist's Introduction to Spectrum Analysis, provides some conceptual background for making and interpreting spectrograms and spectra with Canary. It introduces the short-time Fourier transform (STFT), the mathematical technique used by Canary for making spectrograms. A second aim of this appendix is to explain some of the limitations and tradeoffs intrinsic to spectrum analysis of time-varying signals. The discussions in this appendix assume a basic understanding of how sound is recorded and represented digitally. If you are not already acquainted with concepts such as sampling rate and amplitude resolution (sample size), read Appendix A.

Sound Amplitude Measurements provides a summary of the relationships among sound power, sound intensity, and sound pressure. These quantities are sometimes confused, in part because all three are often expressed as levels using a decibel (dB) scale. First we define each quantity and explain how they are related to each other. We then explain the use of a dB scale to express relative levels of power, intensity, and pressure. An understanding of the basic principles introduced here will be helpful in using Canary. This appendix is not intended to be a comprehensive review of any aspect of the physics of sound; further references are provided.