System Audio Recorder

A lightweight system audio recorder for Ubuntu Linux that captures system audio (not microphone) with a simple keyboard shortcut (Ctrl+Alt+R).

Features

• Record system audio with a keyboard shortcut (Ctrl+Alt+R)
• Desktop notifications for recording status
• Optimized for speech recording (using Opus codec)
• Small file sizes without compromising speech quality

Prerequisites

• Ubuntu Linux (tested on 22.04 LTS)
• Python 3.8+
• FFmpeg
• PulseAudio
• Required packages:

  sudo apt-get install ffmpeg python3-notify2 netcat-openbsd

Quick Start

1. Clone the repository:

git clone https://github.com/YOUR_USERNAME/audio-recorder.git
cd audio-recorder

2. Run the installer:

chmod +x install.sh
./install.sh

3. Use Ctrl+Alt+R to start/stop recording
4. Recordings are saved in ~/projects/audio-recorder/recordings

Project Structure

This project intentionally deviates from the Filesystem Hierarchy Standard (FHS) to maintain a more maintainable and portable structure:

/projects/audio-recorder/
├── bin/                    # Executable scripts
│   └── audio-recorder-toggle
├── src/                    # Python source files
│   ├── recorder.py         # Core recording functionality
│   └── recorder_service.py # System service implementation
├── config/                 # Configuration files
│   └── audio-recorder.service
├── logs/                   # Log files
│   ├── debug.log
│   └── service.log
├── recordings/            # Recorded audio files
└── install.sh            # Installation script

System Integration

Instead of spreading files across the system, this project:

• Keeps all files in one project directory
• Uses symbolic links to integrate with the system:

  /usr/local/bin/audio-recorder-toggle → bin/audio-recorder-toggle
  ~/.config/systemd/user/audio-recorder.service → config/audio-recorder.service

Technical Details

• Uses FFmpeg with PulseAudio for recording

• Implements a systemd user service for background operation

• Uses Unix domain sockets for inter-process communication

• Optimized for speech recording using the Opus codec

• Default recording settings:

  • Codec: Opus
  • Bitrate: 32k
  • Channels: Mono
  • Sample Rate: 16kHz
  • Filters: Speech-optimized bandpass

Uninstallation

chmod +x uninstall.sh
./uninstall.sh