cs2-voice-tools

A modular suite of CLI tools for extracting, transcribing, and analyzing player voice data from CS2 demo files. The tool can extract voice data to various audio formats and filter by specific players.

Forked from: DandrewsDev/CS2VoiceData

---

Features

• Modular CLI tools for each stage of CS2 voice data processing:
  • Extraction (cs2voice extract): Extracts per-player voice data from CS2 demos with support for:
    • Multiple output formats (WAV, MP3, OGG, FLAC, AAC, M4A)
    • Player filtering by SteamID64
    • Safe filename handling for cross-platform compatibility
    • Structured error handling with specific error types
  • Transcription (cs2voice transcribe - planned)
  • Analysis (cs2voice analyze - planned)
  • Unified pipeline (cs2voice pipeline - planned)

---

Installation

• Requires Go 1.23+ and dependencies listed in go.mod.
• Requires ffmpeg installed and available in PATH when using formats other than WAV.

Usage

Global Flags

All commands support the following global flags:

• -v, --verbose: Enable verbose logging (shows additional debug information)
• -o, --output-dir: Directory to save output files (default: current directory)
• -f, --force: Force overwrite existing files (default: skip existing files)

Extract Command Flags

The extract command supports these additional flags:

• -p, --players: Filter to specific players by SteamID64 (comma-separated list)
• -t, --format: Output audio format (wav, mp3, ogg, flac, aac, m4a - default: wav)

Note: Using formats other than WAV requires ffmpeg to be installed on your system

Examples:

# Run extraction with verbose logging
cs2voice extract --verbose my-demo.dem

# Save extracted files to a specific directory
cs2voice extract --output-dir /path/to/output my-demo.dem

# Short forms work too
cs2voice extract -v -o /path/to/output my-demo.dem

# Directory will be created if it doesn't exist
cs2voice extract -o ./new-directory my-demo.dem

# Force overwrite existing files
cs2voice extract --force my-demo.dem

# Extract voice for specific players only
cs2voice extract --players 76561198123456789,76561198987654321 my-demo.dem

# Extract voice in MP3 format
cs2voice extract --format mp3 my-demo.dem

# Extract voice in FLAC format (lossless compression)
cs2voice extract --format flac my-demo.dem

# Combine multiple flags
cs2voice extract -v -o ./output -f -p 76561198123456789 -t mp3 my-demo.dem

---

Troubleshooting

Common issues and solutions:

• No voice data found in demo: Some demos may not contain voice data. Try another demo file.
• ffmpeg not found: Install ffmpeg when using formats other than WAV.
• Invalid SteamID64 format: Ensure player IDs are in the correct format (17-digit numbers starting with 7656).
• Output directory is not writable: Check permissions on the output directory.
• Demo file ended unexpectedly: The demo file might be corrupt or incomplete.

For more detailed error information, run with the --verbose flag.

---

Acknowledgements

• Forked and extended from DandrewsDev/CS2VoiceData.
• Thanks to @DandrewsDev, @rumblefrog, @markus-wa, and all contributors to the original project and libraries.
• Special thanks to demoinfocs-golang and Reversing Steam Voice Codec blog post.

---