WAVsplit

Split WAV files based on stored cue points

Build

make

This will create an executable file called wavsplit in the build/apps/ directory.

wavsplit takes a single argument:

wavsplit file.wav

observe.wav is a sample WAV file with cue points. Running the shell command wavsplit observe.wav will split the WAV data along the cue points into individual files in the observe directory. Note that the WAVsplitter class does not create directories. The observe directory will need to be created before observe.wav can be split into it.

Individual wav files are made based on cue points within the file. If no cue chunks are found, nothing happens.

Internals

The cue chunk (cue_chunk_t) stores the file's individual cue points. These points are read into a struct (cue_point_t):

struct cue_chunk_t
{
    uint32_t cue_points;
    std::vector<cue_point_t> data;
};

struct cue_point_t
{
    uint32_t identifier;
    uint32_t position;
    uint32_t data_chunk_id;
    uint32_t chunk_start;
    uint32_t block_start;
    uint32_t sample_start;
};

The labl chunk stores text identifiers for each cue point which are read and assigned based on cue point identifiers.

Access to the WAV file and the underlying RIFF data is coordinated with WAVparser using the WAV_t class.