Spectral flux is a signal processing feature that quantifies the rate of change of the power spectrum over time. It operates by computing the power spectral density within successive overlapping windows and measuring the L1 distance (sum of absolute differences) between consecutive spectra. Parameters: (1) window size determines spectral resolution vs. temporal resolution trade-off; (2) hop size controls temporal granularity; (3) the spectral magnitude representation (linear, dB, Mel-scale) shapes sensitivity. Persistence mechanism: encoded in audio analysis pipelines across speech recognition (phoneme boundary detection), music information retrieval (onset detection, beat tracking), and acoustic event classification — standard in librosa, Essentia, and Praat.
[formal: fluxus spectralis | substrate: behavior | horizon: hours | explicit: yes | epoch: 0.01]
Accepted ontology entry
spectral-flux
Spectral flux is a signal processing feature that quantifies the rate of change of the power spectrum over time. It operates by computing the power spectral density within successive overlapping windows and measuring the L1 distance (sum o…
Definition
Why it is in scope
A signal processing feature that measures the rate of change of the power spectrum over time. It is human-made: a numerical metric computed by DSP algorithms for audio analysis, music information retrieval, and speech processing. The technique persists through implementations in audio analysis libraries (librosa, Essentia) and research protocols. Spectral flux is used for onset detection, rhythm analysis, audio segmentation, and distinguishing voiced from unvoiced speech segments.
Names and aliases
- spectral-fluxen · CANONICAL
Relations from this entry
- cmspcwtqk04tkjlsslitlqy58DEPENDS_ON →
Spectral-flux measures the rate of change of the power spectrum over time. It operates on the power spectrum — remove it and spectral-flux has no input to operate on. Removal test passes.
- cmspcwtqk04tkjlsslitlqy58DERIVED_FROM →
DERIVED_FROM test: the power spectrum concept existed first and fed into spectral flux. Spectral flux measures the rate of change of the power spectrum over time — it is a derived metric built on the power spectrum foundation. The note pins the sense: spectral flux as a specific measure DERIVED FROM the broader power spectrum concept.
- cmsqglbgv00763e32jisx8hz3INSTANCE_OF →
Spectral-flux IS a specific kind of spectral feature: it measures the rate of change of the spectral energy over time. A competent speaker classifies it as 'a spectral feature' used in audio analysis.
- cmspqj6jh0677jlssnhgl6v8dDEPENDS_ON →
Spectral flux computes the L1 or L2 norm of the difference between consecutive magnitude spectra. Remove magnitude-spectrum and spectral flux has no input data to operate on. The removal test passes.
- cmsq65by307s4jlssrzbe9i0rDERIVED_FROM →
Spectral flux is the rate of change of spectral energy over time. Spectral energy existed first and spectral flux was derived from it. The which-came-first test: energy is a more fundamental concept that predates flux by centuries. Flux is energy in motion across frames.
- cmspltmgd05tcjlsse8vqfce1DEPENDS_ON →
Spectral flux measures frame-to-frame change in spectral magnitude. Its computation requires a time-frequency representation such as STFT; without STFT there is no magnitude spectrum per frame to compare, so flux stops operating.
- cmr9uz3vv00elhcxfruyltnd4INSTANCE_OF →
Spectral flux measures the rate of change of spectral energy between successive frames; it is a specific audio measurement used for onset detection.
- cmsrq3mlc0008sk5348kt4kl9SERVES →
Spectral flux is built and maintained for the sake of onset detection — its design purpose is to measure frame-to-frame spectral change to locate note onsets. The servant points at the master purpose, Law 8d.
Relations to this entry
- cmsrq3mlc0008sk5348kt4kl9← DEPENDS_ON
Onset detection uses spectral flux (the frame-to-frame change in spectral energy) as its primary onset function. Remove spectral-flux and the dominant onset detection method collapses — the algorithm cannot compute spectral flux without it. The removal test passes at object level.
Record identity
- Created
- Aug 12, 2026, 11:15 AM UTC
- Content hash
- 53ccf77aa08fb4d44cd64925d3bc76919ca15d1e5fd5c195d0981d3482c48cd4