process-audio-capture 1.0.0

Windows DLL for capturing audio from specific processes

ProcessAudioCapture

A Windows DLL for capturing audio from specific processes (applications).

Overview

ProcessAudioCapture uses the AUDIOCLIENT_PROCESS_LOOPBACK_PARAMS API introduced in Windows 10 version 2004 (Build 19041) to capture audio output from a specific process only.

Unlike system-wide loopback capture (which records all audio), this library allows you to:

• Record audio from a single application (e.g., browser, game, media player)
• Exclude a specific application's audio from recording
• Monitor real-time audio levels from a specific process

Based on Microsoft's official ApplicationLoopback sample.

Features

• Process-specific capture - Record audio from only the specified process (INCLUDE mode)
• Process exclusion - Record all audio except from the specified process (EXCLUDE mode)
• Real-time level monitoring - Get audio level in dB via callback
• Pause/Resume support - Pause and resume recording without stopping
• WAV file output - Direct output to WAV file
• Process enumeration - List all processes currently outputting audio
• Python bindings included - Ready-to-use Python wrapper

Requirements

• OS: Windows 10 version 2004 (Build 19041) or later, or Windows 11
• Build Environment (if building from source):
  • Visual Studio 2019/2022 with C++ Desktop Development
  • Windows 10 SDK (10.0.19041.0 or later)
  • CMake 3.16 or later

Installation

Using Prebuilt Binaries

Download the latest release from the Releases page.

Extract and use:

• ProcessAudioCapture.dll - The main DLL
• python/process_audio_capture.py - Python wrapper

Building from Source

cd ProcessAudioCapture
mkdir build
cd build
cmake .. -G "Visual Studio 17 2022" -A x64
cmake --build . --config Release

Output files:

• build/bin/Release/ProcessAudioCapture.dll
• build/bin/Release/TestCapture.exe

Quick Start

Python

from process_audio_capture import ProcessAudioCapture

# Check if Windows version is supported
if not ProcessAudioCapture.is_supported():
    print("Windows 10 2004 or later required")
    exit()

# List processes outputting audio
processes = ProcessAudioCapture.enumerate_audio_processes()
for p in processes:
    print(f"PID: {p.pid}, Name: {p.name}, Window: {p.window_title}")

# Capture audio from a specific process
def on_level(level_db):
    print(f"Level: {level_db:.1f} dB")

with ProcessAudioCapture(
    pid=target_pid,
    output_path="output.wav",
    level_callback=on_level
) as capture:
    capture.start()
    time.sleep(10)  # Record for 10 seconds
    # capture.stop() is called automatically

C/C++

#include "ProcessAudioCapture.h"

// Check Windows version
if (!PacIsSupported()) {
    printf("Windows 10 2004 or later required\n");
    return;
}

// List audio processes
PacProcessInfo processes[64];
int count;
PacEnumerateAudioProcesses(processes, 64, &count);

// Start capture
PacHandle handle;
PacErrorCode err = PacStartCapture(
    targetPid,           // Target process ID
    PAC_MODE_INCLUDE,    // Capture mode
    L"output.wav",       // Output file (NULL for monitoring only)
    levelCallback,       // Level callback (can be NULL)
    nullptr,             // User data
    &handle
);

// During capture...
float levelDb;
PacGetLevel(handle, &levelDb);

// Pause/Resume
PacPauseCapture(handle);
// ... paused ...
PacResumeCapture(handle);

// Stop capture
PacStopCapture(handle);

API Reference

Functions

Function	Description
PacIsSupported()	Check if Windows version is supported
PacEnumerateAudioProcesses()	Get list of processes outputting audio
PacStartCapture()	Start capturing audio from a process
PacStopCapture()	Stop capturing
PacPauseCapture()	Pause capturing (audio data not written)
PacResumeCapture()	Resume capturing
PacIsPaused()	Check if capture is paused
PacGetLevel()	Get current audio level in dB
PacIsCapturing()	Check if currently capturing
PacGetLastErrorMessage()	Get last error message
PacGetVersion()	Get DLL version

Error Codes

Code	Value	Description
PAC_SUCCESS	0	Success
PAC_ERROR_INVALID_PARAM	-1	Invalid parameter
PAC_ERROR_NOT_SUPPORTED	-2	Windows version not supported
PAC_ERROR_PROCESS_NOT_FOUND	-3	Process not found
PAC_ERROR_AUDIO_INIT_FAILED	-4	Audio initialization failed
PAC_ERROR_FILE_CREATE_FAILED	-5	File creation failed
PAC_ERROR_ALREADY_RECORDING	-6	Already recording
PAC_ERROR_NOT_RECORDING	-7	Not recording

Capture Modes

Mode	Description
PAC_MODE_INCLUDE	Capture only the specified process's audio
PAC_MODE_EXCLUDE	Capture all audio except the specified process

Directory Structure

ProcessAudioCapture/
├── CMakeLists.txt          # Build configuration
├── LICENSE                 # MIT License
├── README.md               # This file
├── src/
│   ├── ProcessAudioCapture.h   # Public API header
│   ├── ProcessAudioCapture.cpp # DLL implementation
│   ├── LoopbackCapture.h       # Internal capture header
│   └── LoopbackCapture.cpp     # Internal capture implementation
├── python/
│   └── process_audio_capture.py  # Python wrapper
├── test/
│   └── test_capture.cpp    # C++ test application
├── prebuilt/               # Prebuilt binaries
│   └── x64/
│       └── ProcessAudioCapture.dll
└── build/                  # Build output (CMake generated)

Limitations

1. DRM-protected content: Cannot capture (Windows restriction)
2. Audio format: Fixed at 48kHz or device default, 32-bit float, stereo
3. Child processes: Audio from child processes of the target is also included/excluded

Use Cases

• Recording game audio without microphone
• Capturing browser audio (meetings, videos, music)
• Audio processing pipelines
• Accessibility tools
• Streaming/broadcasting specific application audio

License

MIT License - see LICENSE file.

Based on Microsoft's ApplicationLoopback sample.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Acknowledgments

• Microsoft for the ApplicationLoopback sample code
• Windows Audio Session API (WASAPI) documentation