Spectral contrast is a signal processing feature that quantifies the tonal contrast of a frequency spectrum. It computes the difference between peak and valley energy values within user-defined frequency bands of a power spectrum, producing a vector of contrast ratios per band. This feature is widely used in audio classification, music genre recognition, and speech analysis to capture the spectral shape's peakedness. The technique was formalized by Jansen et al. (2005) as part of the MTG-Jamendo feature extraction suite and persists as a standard computation in audio analysis libraries. [formal: contrastus spectralis | substrate: behavior | horizon: a moment | explicit: yes | epoch: 0.75]
Accepted ontology entry
spectral-contrast
Spectral contrast is a signal processing feature that quantifies the tonal contrast of a frequency spectrum. It computes the difference between peak and valley energy values within user-defined frequency bands of a power spectrum, producin…
Definition
Why it is in scope
A human-made signal processing feature that quantifies the tonal contrast of a spectrum by measuring the energy difference between spectral peaks and valleys within defined frequency bands, computed via band-wise peak-to-valley analysis of a power spectrum.
Names and aliases
- spectral-contrasten · CANONICAL
Relations from this entry
- cms68l1jh00cw5ygsxx4vl25kSERVES →
Spectral contrast measures the difference between peaks and valleys in the power spectrum, producing a feature vector that captures the tonal characteristics of a sound. It was specifically designed and is maintained for the purpose of timbre analysis and musical instrument classification. The designed purpose is to serve timbre representation — spectral contrast → timbre follows the servant→master direction of SERVES.
- cmspcwtqk04tkjlsslitlqy58DEPENDS_ON →
Spectral contrast quantifies the difference between peaks and valleys in a spectral representation — this computation fundamentally requires the power spectral density as input. Remove the power spectrum and spectral contrast has no data to operate on. This is a present-tense operational dependency: spectral contrast cannot function without the power spectrum as its computational substrate.
- cmspqj6jh0677jlssnhgl6v8dDEPENDS_ON →
Spectral contrast computes energy differences between peaks and valleys across frequency bands in the magnitude spectrum. Remove magnitude-spectrum and spectral contrast has no input data to operate on.
- cmsop0xz702e7jlssmzq3gk17DEPENDS_ON →
spectral-contrast measures the energy distribution across frequency bands from the spectrum. Remove spectrum and the contrast computation has no data. Genuine operational dependency.
- cmsqglbgv00763e32jisx8hz3INSTANCE_OF →
Spectral-contrast is a specific kind of spectral-feature: it quantifies the energy distribution across different bands of the spectrum relative to the mean. A competent speaker would call spectral contrast 'a spectral feature.'
- cmspltmgd05tcjlsse8vqfce1DEPENDS_ON →
Spectral contrast measures the difference between peaks and valleys in the spectral histogram, computed from the magnitude spectrum of the STFT. Remove STFT and there is no spectrum to histogram — contrast is inoperable.
Relations to this entry
No accepted relations in this direction.
Record identity
- Created
- Aug 12, 2026, 12:01 PM UTC
- Content hash
- 6b3b239f05f548ee87473bb56af540a34ccd5f43619bd1a77c9731720a9aa4c6