Crate resonant_analysis 

resonant-analysis — high-level audio analysis features.

Provides onset detection, beat tracking, pitch estimation, MFCCs, chroma features, and spectral descriptors. All algorithms build on the primitives in [resonant-fft] and [resonant-core].

Quick start

use resonant_analysis::spectral;

let magnitudes = [0.1, 0.5, 1.0, 0.3];
let frequencies = [0.0, 500.0, 1000.0, 1500.0];

let centroid = spectral::spectral_centroid(&magnitudes, &frequencies).unwrap();
let flatness = spectral::spectral_flatness(&magnitudes).unwrap();

Re-exports

pub use error::AnalysisError;

pub use key::KeyDetector;

pub use key::KeyEstimate;

pub use key::Mode;

pub use key::PitchClass;

pub use loudness::LoudnessAnalyser;

pub use lufs::ChannelConfig;

pub use lufs::LufsAnalyser;

Modules

chroma

Chroma features — 12-bin pitch class profiles. Chroma features — 12-bin pitch class profiles.

error

Error types for audio analysis operations.

key

Key detection via Krumhansl-Schmuckler profiles. Key detection via Krumhansl-Schmuckler key profiles.

loudness

Loudness analysis: RMS, peak, and crest factor. Simple loudness analysis: RMS, peak, and crest factor.

lufs

LUFS measurement per ITU-R BS.1770-4. ITU-R BS.1770-4 integrated loudness (LUFS / LKFS) measurement.

mfcc

Mel-frequency cepstral coefficients (MFCCs). Mel-frequency cepstral coefficients (MFCCs).

onset

Onset detection via spectral flux and adaptive thresholding. Onset detection via spectral flux and adaptive thresholding.

pitch

Pitch estimation using the YIN algorithm. Pitch estimation using the YIN algorithm.

spectral

Spectral feature extraction: centroid, spread, flatness, rolloff. Spectral feature extraction: centroid, spread, flatness, rolloff.

tempo

Tempo estimation via autocorrelation of onset strength envelope. Tempo estimation via autocorrelation of the onset strength envelope.

time_stretch

Time-stretching and pitch-shifting via STFT phase vocoder. Time-stretching and pitch-shifting via STFT phase vocoder.