The phon is a unit of loudness level for pure tones. Its purpose is to compensate for the effect of frequency on the perceived loudness of tones. The phon was proposed in DIN 45631 and ISO 532 B by S. S. Stevens.
By definition, the number of phon of a sound is the dB SPL of a sound at a frequency of 1 kHz that sounds just as loud.[clarification needed][example needed] This implies that 0 phon is the limit of perception, and inaudible sounds have negative phon levels.
The equal-loudness contours are a way of mapping the dB SPL of a pure tone to the perceived loudness level (LN) in phons. These are now defined in the international standard ISO 226:2003, and the research on which this document is based concluded that earlier Fletcher-Munson curves and Robinson-Dadson curves were in error.
The phon model can be extended with a time-varying transient model which accounts for "turn-on" (initial transient) and long-term, listener fatigue effects. This time-varying behavior is the result of psychological and physiological audio processing. The equal-loudness contours on which the phon is based apply only to the perception of pure steady tones: tests using octave or third-octave bands of noise reveal a different set of curves, owing to the way in which the critical bands of our hearing integrate power over varying bandwidths and our brain sums the various critical bands