A perceptual frequency scale in hertz, designed to model human auditory perception, where equal-distance steps are perceived as equal pitch intervals. The mel scale maps linear frequency f (in Hz) to perceived pitch m via the formula m = 2595 · log₁₀(1 + f/700). It was developed by Stevens, Volkmann, and Newman (1937) as part of their psychophysical experiments on pitch discrimination. The scale compresses high frequencies (where human pitch discrimination is coarser) relative to low frequencies (where discrimination is finer). It persists through its use in the design of mel-frequency cepstral coefficients (MFCCs), mel-filter banks, and mel-scaled spectrograms — the standard representation for speech recognition, music information retrieval, and audio classification systems. [formal: mel-frequency | substrate: mind | horizon: generations | explicit: yes | epoch: 0.11]
Accepted ontology entry
mel-frequency
A perceptual frequency scale in hertz, designed to model human auditory perception, where equal-distance steps are perceived as equal pitch intervals. The mel scale maps linear frequency f (in Hz) to perceived pitch m via the formula m = 2…
Definition
Why it is in scope
A perceptual frequency scale in hertz, designed to model human auditory perception, where equal-distance steps are perceived as equal pitch intervals. Human-made as a mathematical mapping (linear-to-mel transformation) and used persistently in speech processing and audio analysis for mel-frequency cepstral coefficients, spectrograms, and loudness modeling.
Names and aliases
- mel-frequencyen · CANONICAL
Relations from this entry
- cmsoqq3ec02lqjlssdy5o9rzdDERIVED_FROM →
The mel scale (1937, Stevens-Volkmann-Newman) was developed from the concept of pitch perception to create a scale where equal steps are perceived as equal pitch intervals. Pitch as a perceptual concept existed before the mel frequency scale was created to quantify it.
- cmspywurr06ztjlssi76045rfSERVES →
The mel scale is designed and maintained for the sake of speech recognition and audio analysis. It underpins MFCCs, mel-filterbanks, and mel-spectrograms — the standard representations in speech processing systems.
- cmsqz0txz004nswyn3j0xokafINSTANCE_OF →
Mel-frequency is a specific perceptual frequency scale where equal pitches from human listeners appear equally spaced. It is a concrete instance of the general concept of frequency scale.
Relations to this entry
- cmsqwdhy2004jax3hdwf3yowy← DERIVED_FROM
The mel scale (1937, Stevens-Volkmann-Newman) existed first as a perceptual frequency mapping. The mel-spectrogram builds on this by applying mel-scale filterbanks to a time-frequency decomposition. Which existed first: the mel frequency scale predates the mel-spectrogram concept by decades.
Record identity
- Created
- Aug 13, 2026, 2:12 AM UTC
- Content hash
- 3935016b336816420fe86b1bf8c577a91dc4f4674589e71af66305f22cf29fd3