Wave file format Analysis

As one of the audio waveform file formats used in multimedia, the wave file is standard in riff (resource Interchange File Format) format. The four first bytes of each wave file are "riff ". A wave file consists of a file header and a data body. The file header is divided into two parts: Riff/WAV file ID segment and audio data format description segment. For the content and format of each part of the wave file, see the following.

There are two types of common audio files, which correspond to single-channel (11.025 KHz sampling rate, 8-bit sampling value) and dual-channel (44.1 KHz sampling rate and 16-bit sampling value ). Sampling Rate refers to the number of times the audio signal is sampled per unit time during the "modulo → count" conversion process. Sample value refers to each sampling period
The integral value of the internal voice analog signal.

For single-channel audio files, the sample data is short int 00h-ffh. For two-channel stereo audio files, the sample data is a 16-bit integer (INT ), the 8-bit high and the 8-bit low represent both the left and right channels.

The data block of the wave file contains samples in Pulse-Coded Modulation (PCM) format. A wave file is composed of samples. In a single channel wave file, channel 0 indicates the left channel, and channel 1 indicates the right channel. In a multi-channel wave file, samples are generated alternately.

In addition to the data organization described in the previous section of the wave file header, the data block is the original sound sampling data. Although the wave file can be compressed, it generally uses a non-compressed format. 44.1 KHz sampling rate, 16-bit resolution, dual-channel, so wave can save sound files with high sound quality requirements, CD also uses this format, sound experts or music enthusiasts should be very familiar with it. However, the size of such files is also very large. Taking 44.1 kHz 16-bit dual-channel data as an example, the volume of sound data in one minute is 4100*2 byte * 2channel * 60 S/1024/1024 = 10.09 m. Therefore, it is not suitable for online transmission.

The following is a detailed analysis of the wave file format.

Endian	Field name	Size
Big	Chunkid	4	File Header ID, which is generally four letters: "riff"
Little	Chunksize	4	The size of the entire data file, excluding the above ID and size itself
Big	Format	4	Generally, it is four letters, namely, "wave ".
Big	Subchunk1id	4	Format description block. This field is generally "FMT"
Little	Subchunk1size	4	The size of the data block, excluding the ID and size fields.
Little	Audioformat	2	Audio format description
Little	Numchannels	2	Channels
Little	Samplerate	4	Sampling Rate
Little	Byterate	4	Bit rate, the number of bytes required per second
Little	Blockalign	2	Data Block alignment Unit
Little	Bitspersample	2	Resolution of modulus conversion during sampling
Big	Subchunk2id	4	Real sound data block. This field is generally "data"
Little	Subchunk2size	4	The size of the data block, excluding the ID and size fields.
Little	Data	N	Audio sampling data

The following is a detailed description of each field:

Chunkid	4 bytes	The "riff" in the ASCII code ". (0x52494646)
Chunksize	4 bytes	36 + subchunk2size, or 4 + (8 + subchunk1size) + (8 + subchunk2size ), This is the size of the entire data block (excluding the chunkid and chunksize)
Format	4 bytes	The ASCII code represents the "wave ". (0x57415645)
Subchunk1id		New data block (format description block) The ASCII code indicates "FMT", which is a space at the end. (0x666d7420)
Subchunk1size	4 bytes	The size of the block data (for PCM, the value is 16 ).
Audioformat	2 bytes	PCM = 1 (for example, linear sampling). If it is another value, it may be compressed.
Numchannels	2 bytes	1 => Single Channel | 2 => dual channel
Samplerate	4 bytes	Sampling Rate, such as 8000,44100 equivalent
Byterate	4 bytes	Equal to: samplerate * numchannels * bitspersample/8
Blockalign	2 bytes	Equal to: numchannels * bitspersample/8
Bitspersample	2 bytes	Sample resolution, that is, each sample is represented by a few bits, usually 8 bits or 16 bits.
Subchunk2id	4 bytes	New data blocks, real sound data The ASCII code indicates "data", which is a space at the end. (0x64617461)
Subchunk2size	4 bytes	The data size, that is, the subsequent sample data size.
Data	N Bytes	Real sound data

For data blocks, the layout is as follows (each column represents 8 bits) based on the number of channels and the sampling rate ):

1. 8 bit single channel:

Sample 1	Sample 2
Data 1	Data 2

2. 8-bit dual-channel

Sample 1		Sample 2
Channel 1 data 1	Channel 2 data 1	Channel 1 Data 2	Channel 2 Data 2

1. 16 Bit Single Channel:

Sample 1		Sample 2
Data 1 low byte	Data 1 high byte	Data 1 low byte	Data 1 high byte

2. 16-bit dual-channel

Sample 1
Channel 1 data 1 low byte	Channel 1 data 1 high byte	Audio channel 2 data 1 low byte	Audio channel 2 data 1 high byte
Sample 2
Audio channel 1 Data 2 low bytes	Audio channel 1 Data 2 high bytes	Audio channel 2 Data 2 low bytes	Audio channel 2 Data 2 high byte

 

The following is an example of a sound file:

52 49 46 46 24 08 00 00 57 41 56 45 66 6d 74 20 10 00 00 00 01 00 02 00 22 56 00 00 88 58 01 00 04 00 10 00 64 61 74 61 00 08 00 00 00 00 00 00 24 17 1e f3 3c 13 3c 14 16 f9 18 f9 34 e7 23 a6 3c f2 24 f2 11 ce 1a 0d 

Shows the corresponding analysis:

This article references:

• Http://www.diybl.com/course/3_program/game/200798/70450.html
• Https://ccrma.stanford.edu/courses/422/projects/WaveFormat/