docs.rs failed to build codecpod-0.1.0
Please check the build logs for more information.
See Builds for ideas on how to fix a failed build, or Metadata for how to configure docs.rs builds.
If you believe this is docs.rs' fault, open an issue.

codecpod

English | 简体中文

An audio codec library that statically links FFmpeg, avoiding the hassle of pre-installed dependencies. Provides both a Rust API and Python bindings.

Features

Bundles a static FFmpeg; zero external shared-library dependencies at runtime.
Decoding: AAC, AC-3, ALAC, APE, FLAC, MP1/2/3, Opus, Vorbis, WavPack, various PCM/ADPCM, and more.
Encoding: WAV, AIFF, FLAC, ALAC, AAC, MP3 (LAME), Opus, Vorbis.
Optional resampling, downmixing, range trimming, normalization, and more.

Installation

pip install codecpod

Pre-built wheels are published to PyPI for Linux x86_64, macOS (Apple Silicon and Intel), and Windows x86_64.

Building from source

Building requires the following toolchain:

A Rust toolchain; the latest stable is recommended.
nasm, clang, and libclang.
make, pkg-config.

On Windows the FFmpeg sources are built with autotools, so the toolchain must be the GNU (MinGW-w64) one rather than MSVC: build inside an MSYS2 UCRT64 shell and target x86_64-pc-windows-gnullvm.

Build and install into the current environment:

uvx maturin develop --release

Or produce a wheel:

uvx maturin build --release

The first build downloads the FFmpeg and dependency source tarballs from upstream, pinned to the exact versions and SHA-256 checksums in build.rs. This takes a while, so make sure your network connection is stable.

If the CODECPOD_VENDOR_DIR environment variable is set, the build skips downloading and uses the sources there directly. See build.rs for the directory layout.

Usage

import codecpod
import numpy as np

# Read audio metadata
info = codecpod.info("input.flac")
print(info.sample_rate, info.channels, info.frames, info.codec)

# Decode audio; returns (ndarray, sample_rate) by default, with channels_first=True
waveform, sample_rate = codecpod.load("input.flac")

# Resample to 16 kHz and downmix to mono while decoding
waveform, sample_rate = codecpod.load(
    "input.flac",
    sample_rate=16000,
    mono=True,
)

# Encode to MP3 with explicit encoder parameters
codecpod.save(
    "output.mp3",
    waveform,
    sample_rate,
    codec=codecpod.Mp3(bit_rate=192000),
)

License

This project is distributed under LGPL-2.1-or-later; see LICENSE.