If a recorded audio signal is nearly periodic (as is much voiced speech), it is said to have a fundamental frequency, sometimes denoted “f0” in speech research, equal to one over the period of the signal. The signal can then be approximated as a sum of sinusoids whose frequencies are multiples of f0. Unvoiced speech (fricative consonants, for example) does not have a fundamental frequency. Fundamental frequency is a physical property of a sound or a sound recording.
Many algorithms have been proposed to estimate the fundamental frequency of the voice and/or other recorded audio signals. These are useful for all sorts of analytical tasks, and are also used in various pitch-synchronous processing tasks such as pitch shifting. In real-time machine accompaniment of live instruments or voice, the fundamental frequency is usually the most salient information used to keep track of the instrument or voice’s progress through a piece of music.