Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/Vanilagy/mediabunny/llms.txt

Use this file to discover all available pages before exploring further.

​
What is @mediabunny/ac3?

@mediabunny/ac3 is an extension package that adds AC-3 (Dolby Digital) and E-AC-3 (Dolby Digital Plus) support to Mediabunny. Browsers have no support for these codecs in their WebCodecs implementations. This package provides both a decoder and encoder, implemented using Mediabunny’s custom coder API. Under the hood, it uses a fast, size-optimized WASM build of FFmpeg’s AC-3 and E-AC-3 coders.

​
AC-3 and E-AC-3 support

AC-3 (Dolby Digital) is a widely-used audio codec in:
  • DVD and Blu-ray discs
  • Digital television broadcasts
  • Streaming services
  • Cinema sound systems
E-AC-3 (Dolby Digital Plus) is an enhanced version offering:
  • Better compression efficiency
  • Support for more channels (up to 15.1)
  • Higher quality at lower bitrates
  • Common in streaming platforms like Netflix and Amazon Prime
This extension enables you to work with both formats directly in the browser.

​
Installation

This library peer-depends on Mediabunny. Install both packages: 
npm install mediabunny @mediabunny/ac3
Alternatively, you can include them directly using script tags. Download the distribution files from the releases page.

​
Usage

​
Registering the decoder

Register the AC-3 and E-AC-3 decoder before starting any decoding tasks: 
import { registerAc3Decoder } from '@mediabunny/ac3';

registerAc3Decoder();
Mediabunny will now automatically use the registered decoder when it encounters AC-3 or E-AC-3 audio.

​
Registering the encoder

To enable encoding to AC-3 or E-AC-3 formats: 
import { registerAc3Encoder } from '@mediabunny/ac3';

registerAc3Encoder();

​
Registering both

You can register both the decoder and encoder in a single setup: 
import { registerAc3Decoder, registerAc3Encoder } from '@mediabunny/ac3';

registerAc3Decoder();
registerAc3Encoder();

​
Decoding AC-3 audio

Once registered, Mediabunny automatically uses the AC-3 decoder when processing AC-3 or E-AC-3 audio: 
import { Input, ALL_FORMATS, BlobSource } from 'mediabunny';
import { registerAc3Decoder } from '@mediabunny/ac3';

registerAc3Decoder();

const input = new Input({
    source: new BlobSource(file), // File containing AC-3 or E-AC-3 audio
    formats: ALL_FORMATS,
});

// Mediabunny will automatically decode AC-3/E-AC-3 audio tracks
The decoder (defined in decoder.ts:18):
  • Supports both 'ac3' and 'eac3' codecs
  • Decodes compressed packets to PCM audio samples
  • Handles timestamps and sample rate conversions automatically
  • Runs in a worker thread for optimal performance

​
Encoding to AC-3

To encode audio to AC-3 or E-AC-3 format: 
import {
    Input,
    Output,
    BlobSource,
    BufferTarget,
    ALL_FORMATS,
    Conversion,
} from 'mediabunny';
import { registerAc3Encoder } from '@mediabunny/ac3';

registerAc3Encoder();

const input = new Input({
    source: new BlobSource(inputFile),
    formats: ALL_FORMATS,
});

const output = new Output({
    format: {
        kind: 'ac3', // or 'eac3' for E-AC-3
        audioCodec: 'ac3', // or 'eac3'
        audioChannels: 2,
        audioSampleRate: 48000,
        audioBitrate: 192000, // 192 kbps
    },
    target: new BufferTarget(),
});

const conversion = await Conversion.init({ input, output });
await conversion.execute();

output.target.buffer; // => ArrayBuffer containing the AC-3 file

​
Encoder configuration

The AC-3 encoder (defined in encoder.ts:20) supports: AC-3:
  • Channels: 1-8 channels
  • Sample rates: 48000 Hz, 44100 Hz, 32000 Hz
  • Bitrate: Required (e.g., 192000 for 192 kbps)
E-AC-3:
  • Channels: 1-8 channels
  • Sample rates: Standard rates (48000, 44100, 32000) plus reduced rates (24000, 22050, 16000)
  • Bitrate: Required
Both AC-3 and E-AC-3 encoding require you to specify a bitrate in the encoder configuration. The encoder will not work without it.

​
Implementation details

The package uses a shared worker architecture:
  1. Worker client (worker-client.ts): Manages communication with the worker
  2. Decoder class (decoder.ts:18): Handles decoding workflow
  3. Encoder class (encoder.ts:20): Handles encoding workflow
  4. Worker thread (codec.worker.ts): Loads FFmpeg WASM and processes audio
The encoder:
  • Accumulates incoming audio samples until a full frame is ready
  • Converts samples to f32 interleaved format (required by FFmpeg)
  • Encodes complete frames and outputs timestamped packets
  • Pads the final frame with silence if needed during flush
The decoder:
  • Receives compressed AC-3/E-AC-3 packets
  • Decodes them to PCM audio samples
  • Preserves timestamps and sample rate information
The WASM build is highly optimized for size and performance, using FFmpeg with only AC-3 and E-AC-3 codecs enabled. This keeps the bundle size minimal while providing full codec support.
For more ways of using Mediabunny, refer to the guide.

Build docs developers (and LLMs) love

Get started for freeTalk to us