oximedia-audio-analysis

Advanced audio analysis and forensics for OxiMedia, providing comprehensive audio characterization for professional, forensic, and music applications.

Part of the oximedia workspace — a comprehensive pure-Rust media processing framework.

Features

• Spectral Analysis — Advanced frequency-domain analysis with multiple window functions; spectral centroid, flatness, crest factor, bandwidth, and contrast
• Voice Analysis — F0/formant analysis, gender detection, age estimation, emotion detection, speaker identification and verification
• Music Analysis — Harmonic analysis, chord progression detection, rhythmic analysis, timbral analysis, instrument identification
• Source Separation — Vocal/instrumental separation, drum and bass isolation via harmonic-percussive decomposition
• Echo and Reverb Analysis — Room acoustics measurement, RT60 estimation, early reflection pattern analysis
• Distortion Analysis — THD measurement, clipping detection, non-linear distortion characterization
• Dynamic Range Analysis — Crest factor, RMS tracking over time, loudness variation measurement
• Transient Detection — Attack/onset detection, ADSR envelope characterization, onset strength functions
• Pitch Analysis — YIN algorithm pitch tracking (patent-free), vibrato detection, F0 estimation with confidence
• Formant Analysis — F1–F4 tracking over time, vowel detection and classification, LPC-based formant extraction
• Audio Forensics — Authenticity verification, edit detection (cuts/splices), ENF analysis, compression history analysis, background noise consistency
• Noise Analysis — Noise profiling, SNR computation, noise floor estimation, noise type classification
• Psychoacoustic Analysis — Perceptual feature extraction, masking models
• Stereo Field Analysis — Width, phase correlation, mid/side analysis
• Spectral Flux — Frame-to-frame spectral change detection

Usage

Add to your Cargo.toml:

[dependencies]
oximedia-audio-analysis = "0.1.1"

use oximedia_audio_analysis::{AudioAnalyzer, AnalysisConfig};

let config = AnalysisConfig::default();
let analyzer = AudioAnalyzer::new(config);

let samples = vec![0.0_f32; 44100]; // 1 second of audio
let sample_rate = 44100.0;

let result = analyzer.analyze(&samples, sample_rate)?;

println!("Spectral centroid: {:.1} Hz", result.spectral.centroid);
println!("Spectral flatness: {:.3}", result.spectral.flatness);

API Overview (81 source files, 479 public items)

Core types:

• AudioAnalyzer — Main analysis entry point
• AnalysisConfig — Configuration for analysis parameters

Modules:

• spectral — Frequency-domain spectral analysis (FFT frame, spectral features, flux, contrast)
• voice — Voice characteristic analysis (characteristics, speaker identification)
• rhythm — Rhythmic feature extraction and tempo analysis
• music — Music-specific analysis (timbre, rhythm)
• forensics — Audio authenticity and tampering detection (compression history, noise consistency)
• pitch — Pitch tracking (YIN algorithm), vibrato detection
• pitch_tracker — Extended pitch tracking framework
• formant — Formant frequency analysis and vowel detection
• noise — Noise profiling and SNR computation
• echo — Echo and reverb detection and measurement
• onset — Transient and onset detection
• transient — Transient detection and envelope analysis
• beat — Beat tracking and rhythm analysis
• cepstral — Cepstral analysis for voice and music
• harmony — Harmonic analysis and chord detection
• psychoacoustic — Psychoacoustic feature extraction
• stereo_field — Stereo width and phase correlation
• spectral_contrast, spectral_features, spectral_flux — Advanced spectral metrics
• separate — Source separation
• tempo_analysis — Tempo estimation

License

Apache-2.0 — Copyright 2024-2026 COOLJAPAN OU (Team Kitasan)