WaveExploder

Splits multi-channel WAV files into individual per-channel WAV files. Built for post-processing multi-track live recordings from a Behringer Wing mixer.

Scripts

Script	Purpose
wing_sync.py	Queries the Wing mixer via OSC and writes channel names to config.ini
main.py	Reads config.ini and splits multi-channel WAV files into per-channel files

Typical workflow: run wing_sync.py before a session to pull channel names from the mixer, then run main.py after recording to split the files.

Requirements

• Python 3.10 or newer
• Dependencies:

pip install -r requirements.txt

Windows Firewall — one-time setup for wing_sync.py

Windows blocks inbound UDP by default. Run this once in an elevated (Admin) PowerShell before using wing_sync.py:

netsh advfirewall firewall add rule name="WingSync UDP 62000" dir=in action=allow protocol=UDP localport=62000

Usage

wing_sync.py — sync channel names from the mixer

# Auto-discover the Wing on the local network, update config.ini
python wing_sync.py

# Skip discovery, use a known IP
python wing_sync.py --ip 192.168.1.103

# Preview what would be written without changing config.ini
python wing_sync.py --dry-run

# Show raw OSC traffic for debugging
python wing_sync.py --verbose

After the first successful run, the Wing's IP address is saved to config.ini under [Wing], so subsequent runs skip network discovery automatically.

main.py — split multi-channel WAV files

# Run with config.ini defaults
python main.py

# Override input/output directories
python main.py --input "C:\path\to\recordings" --output "C:\path\to\output"

# Override recursion behaviour
python main.py --recurse
python main.py --no-recurse

# Use a different config file
python main.py --config other.ini

# Show per-channel RMS values
python main.py --verbose

Configuration (config.ini)

Copy config.ini.example to config.ini and edit it. The file has three sections:

[Setup]

Key	Description
initial_input_directory	Root directory to scan for multi-channel WAV files
initial_output_directory	Directory where per-channel WAV files are written
input_file_prefix	Filename prefix used to filter input files (e.g. 0000)
max_num_channels	Upper bound of channels to scan (default: 32)
recurse_sub_directories	If True, process all subdirectories recursively
use_dir_name_as_output_file_prefix	If True, prepend the source directory name to output filenames
use_input_filename_as_output_file_prefix	If True, prepend the input WAV filename (without extension) to output filenames; takes priority over use_dir_name_as_output_file_prefix
explode_unnamed_channels	If True, extract channels with no name in [Channel.Names]
explode_silent_channels	If True, extract channels whose RMS is below the silence threshold
silent_channel_threshold	RMS value below which a channel is considered silent (default: 0.007)

[Wing]

Key	Description
ip	IP address of the Wing mixer — written automatically by wing_sync.py

[Channel.Names]

Maps channel numbers to human-readable names. Written automatically by wing_sync.py, or set manually:

[Channel.Names]
Ch09 = Vocal_Lead
Ch17 = Drum_Kick
Ch18 = Drum_Snare

Channels not listed here are considered "unnamed". Whether unnamed channels are extracted is controlled by explode_unnamed_channels in [Setup].

Examples

Two typical workflows are supported depending on how recordings are organised on disk.

Method 1 — one subdirectory per song (Wing SD card layout)

The Wing creates a separate folder for each recording session. Each folder contains a single multi-channel WAV file.

Recordings/
├── Song01/
│   └── 0000_Song01.wav   (32-channel)
├── Song02/
│   └── 0000_Song02.wav   (32-channel)
└── Song03/
    └── 0000_Song03.wav   (32-channel)

config.ini:

[Setup]
initial_input_directory            = C:\Recordings
initial_output_directory           = C:\Recordings\Exploded
recurse_sub_directories            = True
use_dir_name_as_output_file_prefix = True
use_input_filename_as_output_file_prefix = False
explode_unnamed_channels           = False
explode_silent_channels            = False

[Channel.Names]
Ch09 = Vocal_Lead
Ch17 = Drum_Kick
Ch18 = Drum_Snare

Running python main.py produces:

Exploded/
├── Song01/
│   ├── Song01_Ch09_Vocal_Lead.wav
│   ├── Song01_Ch17_Drum_Kick.wav
│   └── Song01_Ch18_Drum_Snare.wav
├── Song02/
│   ├── Song02_Ch09_Vocal_Lead.wav
│   ├── Song02_Ch17_Drum_Kick.wav
│   └── Song02_Ch18_Drum_Snare.wav
└── Song03/
    ├── Song03_Ch09_Vocal_Lead.wav
    ├── Song03_Ch17_Drum_Kick.wav
    └── Song03_Ch18_Drum_Snare.wav

Unnamed channels (e.g. Ch01–Ch08, Ch10–Ch16, Ch19–Ch32) and silent channels are skipped because explode_unnamed_channels and explode_silent_channels are both False.

Method 2 — flat directory, WAV files named by song

All multi-channel WAV files are in a single directory and are already named by song.

Recordings/
├── Song01.wav   (32-channel)
├── Song02.wav   (32-channel)
└── Song03.wav   (32-channel)

config.ini:

[Setup]
initial_input_directory            = C:\Recordings
initial_output_directory           = C:\Recordings\Exploded
recurse_sub_directories            = False
input_file_prefix                  =
use_dir_name_as_output_file_prefix = False
use_input_filename_as_output_file_prefix = True
explode_unnamed_channels           = False
explode_silent_channels            = False

[Channel.Names]
Ch09 = Vocal_Lead
Ch17 = Drum_Kick
Ch18 = Drum_Snare

Running python main.py produces:

Exploded/
├── Song01_Ch09_Vocal_Lead.wav
├── Song01_Ch17_Drum_Kick.wav
├── Song01_Ch18_Drum_Snare.wav
├── Song02_Ch09_Vocal_Lead.wav
├── Song02_Ch17_Drum_Kick.wav
├── Song02_Ch18_Drum_Snare.wav
├── Song03_Ch09_Vocal_Lead.wav
├── Song03_Ch17_Drum_Kick.wav
└── Song03_Ch18_Drum_Snare.wav

Output filename format

[Prefix_]ChNN_ChannelName.wav

The prefix comes from exactly one source (use one or neither, not both):

Setting	Prefix source	Example
use_dir_name_as_output_file_prefix = True	Source directory name	Song01_Ch09_Vocal_Lead.wav
use_input_filename_as_output_file_prefix = True	Input WAV filename stem	Song01_Ch09_Vocal_Lead.wav
Both False	No prefix	Ch09_Vocal_Lead.wav

Running tests

python -m pytest tests/ -v

To Do

• Behringer X32 Support

  I think it would be pretty easy to add Behringer X32 functionality as well as Behringer Wing, but I no longer have an X32 to test on.