WAVparser

Read, manipulate, and write wav files in C++.

A general example for reading and writing WAV files using the WAV_t class is in the main.cpp file. The sample WAV file included and used in the main.cpp program is from https://freewavesamples.com/ensoniq-zr-76-01-dope-77-bpm.

Creating a WAV_t object

To create an empty WAV file:

WAV_t wav;

To parse an existing WAV file:

WAV_t wav("path/to/file.wav");

To write to a WAV file:

wav.write("path/to/new/file.wav");

Header

The header is stored as a private member of the WAV_t class.

struct WAV_fmt_t
{
    uint16_t audio_format;
    uint16_t num_channels;
    uint32_t sample_rate;
    uint32_t byte_rate;
    uint16_t block_align;
    uint16_t bits_per_sample;
    uint16_t extra_params_size;
    std::vector<uint8_t> extra_params;
};

The fields in a WAV_fmt_t struct generally depend on things like encoding and sample size. Most of the header details are hidden. Some can be edited:

uint16_t sample_rate() const;
void set_sample_rate(uint16_t new_rate);
int sample_size() const;
std::vector<uint8_t> &extra_params(); // extra parameters field in the header, don't know what actually goes here.

File encoding can be changed on the fly:

WAV_encoding get_encoding() const;
void set_encoding(WAV_encoding new_encoding);

block alignment and byte rate can be calculated with some helper functions:

int byte_rate = wav.calculate_byte_rate();
int block_align = wav.calculate_block_align();

I don't personally see how this information is useful outside of the file header but its available just in case. These functions also set their respective fields in the internal header struct.

Samples

Individual audio channels and samples can be retreived and assigned using the channel() function:

int sample = wav.channel(0)[0]; // first sample from channel 0
wav.channel(1)[0] = 0; // first sample from channel 1

In stereo audio left channel = 0 and right channel = 1, usually.

To find the number of bytes per sample, use sample_size():

int sample_size = wav.sample_size();

This will return the number of bytes that the currently set encoding requires per audio sample. Actual audio samples are all stored as doubles.

To clear all samples from a WAV_t object, use clear_data():

wav.clear_data();

RIFF_t

WAV files are contained within RIFF files. As such, the WAV_t class is built using the RIFF_t class. See the RIFFparser repository for more information about reading and manipulating RIFF data.