Handcrafting a WAV file

library(ctypesio)

WAV files

WAV files are a standard format for storing audio data.

This vignette will use the WAV file structure to write bespoke data as a WAV file, and then read it back in.

WAV file structure

The 44-byte header structure for a simple wave file is shown below.

WAV file structure
Positions Sample.Value Description
1 - 4 ‘RIFF’ Marks the file as a riff file. Characters are each 1 byte long.
5 - 8 File size (integer) Size of the overall file - 8 bytes, in bytes (32-bit integer). Typically, you’d fill this in after creation.
9 -12 “WAVE” File Type Header. For our purposes, it always equals “WAVE”.
13-16 ‘fmt’ Format chunk marker. Includes trailing space
17-20 16 Length of format data as listed above
21-22 1 Type of format (1 is PCM) - 2 byte integer
23-24 2 Number of Channels - 2 byte integer
25-28 44100 Sample Rate - 32 byte integer. Common values are 44100 (CD), 48000 (DAT). Sample Rate = Number of Samples per second, or Hertz.
29-32 176400 (Sample Rate * BitsPerSample * Channels) / 8.
33-34 4 (BitsPerSample * Channels) / 8. (1) 8 bit mono (2) - 8 bit stereo/16 bit mono (4) - 16 bit stereo
35-36 16 Bits per sample
37-40 ‘data’ data chunk header. Marks the beginning of the data section.
41-44 File size (data) Size of the data section.

Audio waveform

Create 2 seconds of random noise data to be played at 44.1 kHz.

Data values should be in range for a 16-bit integer i.e. [-32768, 32767]

audio <- as.integer(runif(44100 * 2, -32768, 32767))

Write the WAV file

Data is stored little-endian (which is the default endian mode for ctypesio)

wav_file <- tempfile(fileext = ".wav")
con <- file(wav_file, "wb")

write_utf8_raw(con, "RIFF")
write_uint32(con, length(audio) * 2 + 44 - 8) # total file length - 8
write_utf8_raw(con, "WAVE")
write_utf8_raw(con, "fmt ")
write_uint32(con, 16)                 # size of header so far
write_uint16(con, 1)                  # sample format = 2-byte integer
write_uint16(con, 1)                  # single channel
write_uint32(con, 44100)              # sample rate
write_uint32(con, 44100 * 16 * 1 / 8) # rate * bits * channels / 8
write_uint16(con, 2)                  # 16-bit mono audio
write_uint16(con, 16)                 # bits per sample
write_utf8_raw(con, 'data')
write_uint32(con, length(audio) * 2)  # sampled data size
write_int16(con, audio)

close(con)

Play the audio

This audio isn’t very interesting, but it can be played using the audio package

library(audio)
wav <- audio::load.wave(wav_file)
audio::play(wav)

Parsing a WAV file

This is just the reverse of writing the file.

con <- file(wav_file, "rb")

read_utf8_raw(con, 4)
#> [1] "RIFF"
read_uint32(con)      # total file length - 8
#> [1] 176436
read_utf8_raw(con, 4)
#> [1] "WAVE"
read_utf8_raw(con, 4)
#> [1] "fmt "
read_uint32(con)      # size of header above
#> [1] 16
read_uint16(con)      # sample format = 2-byte integer
#> [1] 1
read_uint16(con)      # single channel
#> [1] 1
read_uint32(con)      # sample rate
#> [1] 44100
read_uint32(con)      # rate * bits * channels / 8
#> [1] 88200
read_uint16(con)      # 16-bit mono audio
#> [1] 2
read_uint16(con)      # bits per sample
#> [1] 16
read_utf8_raw(con, 4)
#> [1] "data"
(data_size <- read_uint32(con))         # sampled data size
#> [1] 176400
data <- read_uint16(con, data_size / 2) # read in 16-bit samples

close(con)

ctypesio is pipe-friendly

All the write_*() functions in ctypesio return the connection object, so that the functions can piped together.

wav_file <- tempfile(fileext = ".wav")
con <- file(wav_file, "wb")

write_utf8_raw(con, "RIFF") |>
  write_uint32(length(audio) * 2 + 44 - 8) |> # total file length - 8
  write_utf8_raw("WAVE")           |>
  write_utf8_raw("fmt ")           |> 
  write_uint32(16)                 |> # size of header so far
  write_uint16(1)                  |> # sample format = 2-byte integer
  write_uint16(1)                  |> # single channel
  write_uint32(44100)              |> # sample rate
  write_uint32(44100 * 16 * 1 / 8) |> # rate * bits * channels / 8
  write_uint16(2)                  |> # 16-bit mono audio
  write_uint16(16)                 |> # bits per sample
  write_utf8_raw('data')           |> 
  write_uint32(length(audio) * 2)  |> # sampled data size
  write_int16(audio)

close(con)