Track Scalpel

Read chapters from a Blu-ray MPLS playlist file and split audio files into tracks based on those chapters

Features

• Read and write any format supported by SoundFile/libsndfile, including WAV, W64, AIFF and FLAC among others
• Can delegate splitting to SoX to support other formats
• Generate a list of sample-accurate cut points to use with other tools
• Read MPLS directly for maximum precision: MPLS timestamps are accurate to 1/45000 of a second whereas text chapter files are accurate to 1/1000s and cue sheets are accurate to 1/75s
• Configurable block/sector alignment: can be sample-accurate or round down/up/to nearest 1/n of a second, useful for aligning to audio CD sectors or frames for gapless playback

Known issues

The Python SoundFile module does not appear to provide a way to set channel layout metadata when writing files, so while this program can read and write 5.1 and 7.1 surround, it does not write channel layout. SoX does not properly preserve channel layout when working with WAV either, but does for FLAC so you can work around this problem by delegating cutting to SoX and using FLAC as both input and output format.

WAV files over 2GB are not supported. You should use a format that supports large files such as FLAC or W64.

Requirements

• Python 3.6+
• NumPy and SoundFile are required to read/write sound files
• SoX (optional)

Usage

To install, run:

pip install git+https://github.com/michaelburton/trackscalpel.git

To list the sample numbers where each chapter in playlist 00001.mpls begins given a sample rate of 48kHz, run:

trackscalpel 48000 00001.mpls

To split a file called audio.wav with playlist 00001.mpls and put the resulting files (01.wav, 02.wav, ...) into the current directory, run:

trackscalpel audio.wav 00001.mpls

To split a file called audio.wav with playlist 00001.mpls, align splits along Audio CD sector boundaries, encode the tracks in FLAC format and place them in the 'tracks' directory (tracks/01.flac, tracks/02.flac, ...), run:

trackscalpel --cd -f FLAC -o tracks audio.wav 00001.mpls

For more detailed usage information, run:

trackscalpel -h

Acknowledgements

MPLS parsing code is from PyGuymer, very lightly modified to run under Python 3 (I just replaced xrange with range).