OwnAudioSharp 4.0.3

OwnAudioSharp

Complete cross-platform audio library for .NET desktop applications

OwnAudioSharp is a professional-grade audio engine providing high-performance audio playback, recording, and processing for Windows, Linux, and macOS with zero external dependencies. This is the full edition: everything the library offers, including MIDI, the analysis features (chord detection, note transcription, matchering) and the Avalonia waveform display. For a lean playback/recording-only build see OwnAudioSharp.Basic; for Android and iOS see OwnAudioSharp.Mobile.

Key Features

• Native Rust Audio Engine: Built on a purpose-built native Rust core for professional-grade, low-latency audio. Device I/O, mixing, and the full effect chain run entirely in native code with a real-time-safe hot path — no PortAudio or MiniAudio dependency.
• Multi-format Support: Native pure-Rust decoder (Symphonia) with built-in support for WAV, MP3, FLAC, OGG/Vorbis, AAC/M4A, ALAC/M4A, and AIFF. For any other format, FFmpeg is used automatically as a fallback when installed — no code changes required.
• Real-time Processing: Zero-allocation design with lock-free buffers and native mixing for professional-grade performance
• Advanced Audio Features:
  • Network Synchronization: Multi-device audio sync across local network (< 5ms accuracy on LAN)
  • Master Clock: Sample-accurate timeline synchronization for multi-track playback
  • SmartMaster Effect: Intelligent audio mastering with auto-calibration
  • AI-powered vocal removal (ONNX-based neural separation)
  • Audio matchering and mastering
  • Real-time chord detection
  • Built-in effects and DSP routines

Quick Start

using OwnaudioNET;
using OwnaudioNET.Features.Vocalremover;

// Initialize the audio engine
OwnaudioNet.Initialize();
OwnaudioNet.Start();

// Create the audio mixer using the underlying engine
var mixer = new AudioMixer(OwnaudioNet.Engine.UnderlyingEngine);
mixer.Start();

// Play an audio file
var music = new FileSource("music.mp3");
mixer.AddSource(music);

// Synchronized Multi-track Playback (Master Clock)
var vocals = new FileSource("vocals.wav");
var backing = new FileSource("backing.mp3");

mixer.AddSource(vocals);
mixer.AddSource(backing);

// Attach sources to the Master Clock for sample-accurate sync
vocals.AttachToClock(mixer.MasterClock);
backing.AttachToClock(mixer.MasterClock);

// Start sources individually
vocals.Play();
backing.Play();

// Network Synchronization - Multi-Device Audio
// Server mode (control device)
await OwnaudioNet.StartNetworkSyncServerAsync(port: 9876);

// Client mode (follower devices)
await OwnaudioNet.StartNetworkSyncClientAsync(
    serverAddress: "192.168.1.100",  // Or null for auto-discovery
    allowOfflinePlayback: true);

// All clients automatically follow server commands
// Perfect for multi-room audio, DJ setups, installations

// AI Vocal Removal
var options = new SimpleSeparationOptions 
{ 
    Model = InternalModel.Best, 
    OutputDirectory = "output" 
};

using var separator = new SimpleAudioSeparationService(options);
separator.Initialize();
var result = separator.Separate("song.mp3");
// result.VocalsPath and result.InstrumentalPath contain the output files

Audio Decoding

Decoding is handled by the native Rust engine using a pure-Rust Symphonia backend — no managed decoder and no external runtime is required for the common formats:

Natively supported (built-in): WAV, MP3, FLAC, OGG/Vorbis, AAC/M4A, ALAC/M4A, AIFF

For any format the native backend cannot handle, OwnAudioSharp transparently falls back to FFmpeg when it is installed on the system. This is not part of the public API — the decoder layer selects the best backend automatically, with no code changes required.

Decoder priority: native Rust (Symphonia) → FFmpeg (optional fallback)

Optional FFmpeg installation per platform:

• Windows: Place the FFmpeg DLLs next to the executable, or anywhere on PATH.
• macOS: brew install ffmpeg
• Linux: sudo apt install ffmpeg (or equivalent)

Platform Support

• Windows: Windows 10/11 (x64, ARM64)
• Linux: Ubuntu, Debian, Fedora, etc. (x64, ARM64)
• macOS: macOS 10.13+ (x64, ARM64)

Architecture

OwnAudioSharp uses a two-layer architecture:

1. Native Rust Engine Layer: Device I/O, multi-track mixing, resampling, and the full effect chain run in a native Rust core. Audio data stays in native memory on a real-time-safe, allocation-free hot path.
2. Managed API Layer: High-level thread-safe wrappers that drive the native engine through a lock-free FFI boundary. The managed side only issues control commands, so the UI thread is never blocked and no managed code runs in the audio path.

Documentation

• GitHub: https://github.com/ModernMube/OwnAudioSharp
• Documentation: https://modernmube.github.io/OwnAudioSharp/

License

MIT License - Copyright (c) 2025 ModernMube

Development Tools

This project is developed with the following tools:

	Anthropic — Claude Code
	Microsoft — Visual Studio Code
	Microsoft — Visual Studio 2022
	JetBrains — Rider