Pure-Rust audio processing workspace: decode, encode, DSP effects, and spectral analysis.

Version: 0.1.2 | MSRV: 1.80 | License: Apache-2.0

Biquad EQ · Parametric EQ · Butterworth/Chebyshev/Elliptic/FIR filters · Compressor/Limiter/Gate/Expander/De-esser · Multiband compressor · Chorus/Flanger/Phaser/Tremolo/Vibrato · Delay · Freeverb + convolution reverb · Phase vocoder (pitch shift + time stretch) · Channel vocoder · YIN/pYIN pitch detection · Autocorrelation pitch tracker · Onset detection (spectral flux, HFC, complex domain) · Beat tracking · Spectral subtraction + Wiener filter noise reduction · EBU R128 / ITU-R BS.1770 loudness (LUFS + true peak) · ReplayGain · MFCC · Chromagram · Spectral centroid/flux/rolloff/flatness/contrast/tonnetz · STFT / iSTFT · Mel-spectrogram (via OxiFFT) · Kaiser and FlatTop window functions

• Surround layouts: Quad, 5.1, 7.1, 5.1-Side, Atmos 7.1.4
• ChannelMap / ChannelId with SMPTE/ITU-R BS.775 ordering
• Downmix (5.1 → stereo, N-ch → mono) and upmix utilities per ITU-R BS.775
• WAVE_FORMAT_EXTENSIBLE for >2 channel WAV output

• RF64/BW64 WAV for audio files exceeding 4 GB
• FLAC album art via METADATA_BLOCK_PICTURE (encode_flac_with_album_art)
• ID3v2.4 writer with UTF-8, APIC album art, USLT lyrics, extended header CRC, ReplayGain
• APEv2 tag writer for WavPack/Musepack output
• Two-pass EBU R128 loudness normalization (−14/−16/−23 LUFS targets)
• Noise-shaped (ATH-weighted) dithering for perceptually optimal bit-depth reduction

• TranscodeStream streaming transcode pipeline (decode → optional DSP → encode)
• transcode_batch parallel batch format conversion via rayon
• DspChain composable DSP effect builder
• AudioRingBuffer<T> lock-free SPSC ring buffer with wait-free overflow policy
• AudioPipeline with parallel branches, bypass/mute, and latency reporting
• AudioClock with drift_ppm, elapsed_frames, elapsed_secs

oxiaudio/                    (facade — default = ["pure"])
  oxiaudio-core              (AudioBuffer, traits, error, IPC, ring buffer, surround layouts)
  oxiaudio-decode            (SymphoniaDecoder + AIFF/AU/Opus/WavPack/Musepack/MIDI)
  oxiaudio-encode            (WAV/RF64, FLAC, AIFF, AU, ID3v2.4, APEv2; streaming + two-pass)
  oxiaudio-encode-mp3-lame   (LAME FFI adapter — opt-in, never default)
  oxiaudio-dsp               (resample, filters, dynamics, reverb, pitch, spectral, loudness)

Default features carry zero C/C++/Fortran dependencies. The mp3-encode-lame feature is the sole sanctioned FFI boundary (LGPL, opt-in).

use std::path::Path;

// Decode any supported format
let buf = oxiaudio::decode_file(Path::new("input.flac")).expect("decode failed");
println!("{} frames @ {} Hz", buf.frame_count(), buf.sample_rate);

// DSP: normalize, then add reverb
let mut out = buf.clone();
oxiaudio::dsp::normalize(&mut out, -1.0);
let with_reverb = oxiaudio::dsp::reverb(&out, 0.6, 0.4, 0.3);

// Re-encode as FLAC
oxiaudio::encode_flac(&with_reverb, Path::new("output.flac")).expect("encode failed");

All M0–M23 milestones complete; Opus CELT/SILK/Hybrid encoders conformant as of 2026-06-10 (v0.1.2).

• 1,133 tests passing, 0 clippy warnings
• 41,033+ production SLoC across 6 crates
• All major codecs, DSP algorithms, and tagging formats implemented
• Pure-Rust Opus encoder: CELT-only, SILK NB/WB, and Hybrid FB modes (RFC 6716 conformant)
• Pure Rust default features (LAME FFI is opt-in only)