Codex.Voice.Input.AudioInput (Codex SDK v0.7.2)

A single, complete audio input buffer.

This struct holds static audio data along with its format parameters. The default format is PCM16 at 24kHz mono, which is the standard format for OpenAI's voice APIs.

Summary

Types

t()

Functions

new(data, opts \\ [])

Create a new audio input from binary data.

to_audio_file(input)

Convert audio to WAV file format.

to_base64(audio_input)

Encode audio data to base64.

Types

t()

@type t() :: %Codex.Voice.Input.AudioInput{
  channels: pos_integer(),
  data: binary(),
  frame_rate: pos_integer(),
  sample_width: pos_integer()
}

Functions

new(data, opts \\ [])

@spec new(
  binary(),
  keyword()
) :: t()

Create a new audio input from binary data.

Options

:frame_rate - Sample rate in Hz (default: 24000)
:sample_width - Bytes per sample (default: 2)
:channels - Number of audio channels (default: 1)

Examples

iex> data = <<0, 0, 255, 127>>
iex> input = Codex.Voice.Input.AudioInput.new(data)
iex> input.frame_rate
24000

to_audio_file(input)

@spec to_audio_file(t()) :: {String.t(), binary(), String.t()}

Convert audio to WAV file format.

Returns a tuple of {filename, wav_data, mime_type} suitable for multipart uploads or file operations.

Examples

iex> data = <<0, 0, 255, 127>>
iex> input = Codex.Voice.Input.AudioInput.new(data)
iex> {filename, _wav_data, mime} = Codex.Voice.Input.AudioInput.to_audio_file(input)
iex> {filename, mime}
{"audio.wav", "audio/wav"}

to_base64(audio_input)

@spec to_base64(t()) :: String.t()

Encode audio data to base64.

This is useful when sending audio data to APIs that expect base64-encoded audio.

Examples

iex> data = <<0, 0, 255, 127>>
iex> input = Codex.Voice.Input.AudioInput.new(data)
iex> Codex.Voice.Input.AudioInput.to_base64(input)
"AAD/fw=="