parseAudioMetadata
Parses audio format parameters out of a container header.
function parseAudioMetadata(buffer, format?): ParsedAudioMetadata | null;
Defined in: src/utils/audioFormat.ts:509
Parses audio format parameters out of a container header.
Parameters
| Parameter | Type | Description |
|---|---|---|
buffer | ArrayBuffer | The audio data buffer to parse, starting at the first byte of the stream. |
format? | DetectedAudioFormat | The already-detected format, if known. Skips a redundant detectAudioFormat call. |
Returns
ParsedAudioMetadata | null
The parsed parameters, or null if the format is unknown or the buffer does not yet reach the parameters.
Remarks
Where detectAudioFormat answers “which container is this?”, this function answers “what audio is inside it?” — enough to fill in AudioMetadata without the caller declaring the format.
Coverage:
| Format | Sample rate | Channels | Bit depth | Encoding |
|---|---|---|---|---|
| WAV | ✓ | ✓ | ✓ | linear16 / mulaw / alaw / mp3 |
| OGG | ✓ | ✓ | — | opus (Vorbis: none) |
| MP3 | ✓ | ✓ | — | mp3 |
| FLAC | ✓ | ✓ | ✓ | — |
| Others | — | — | — | — |
Returns null when more bytes are needed. Container parameters can sit well past MIN_SNIFF_BYTES — behind a WAV LIST chunk or an ID3v2 tag, for instance — so a null result on a buffer whose format is detectable means “accumulate more data and call again”. Callers should stop accumulating at MAX_SNIFF_BYTES and fall back to the container format alone. null is also returned when no format is detected at all.
Example
import { parseAudioMetadata } from 'composite-voice';
import { readFileSync } from 'node:fs';
const wav = readFileSync('speech.wav');
parseAudioMetadata(wav.buffer);
// => { format: 'wav', mimeType: 'audio/wav', sampleRate: 16000,
// channels: 1, bitDepth: 16, encoding: 'linear16' }
See
- detectAudioFormat for container detection
- BufferInput for the input provider that applies this automatically