just_audio

just_audio is a feature-rich audio player for Android, iOS, macOS, web, Linux and Windows.

Platform SupportAPI DocumentationTutorialsBackground AudioCommunity Support

Screenshot with arrows pointing to features

Quick synopsis

import 'package:just_audio/just_audio.dart';

final player = AudioPlayer();                   // Create a player
final duration = await player.setUrl(           // Load a URL
    'https://foo.com/bar.mp3');                 // Schemes: (https: | file: | asset: )
player.play();                                  // Play without waiting for completion
await player.play();                            // Play while waiting for completion
await player.pause();                           // Pause but remain ready to play
await player.seek(Duration(second: 10));        // Jump to the 10 second position
await player.setSpeed(2.0);                     // Twice as fast
await player.setVolume(0.5);                    // Half as loud
await player.stop();                            // Stop and free resources

Working with multiple players

// Set up two players with different audio files
final player1 = AudioPlayer(); await player1.setUrl(...);
final player2 = AudioPlayer(); await player2.setUrl(...);

// Play both at the same time
player1.play();
player2.play();

// Play one after the other
await player1.play();
await player2.play();

// Loop player1 until player2 finishes
await player1.setLoopMode(LoopMode.one);
player1.play();          // Don't wait
await player2.play();    // Wait for player2 to finish
await player1.pause();   // Finish player1

// Free platform decoders and buffers for each player.
await player1.stop();
await player2.stop();

Working with clips

// Play clip 2-4 seconds followed by clip 10-12 seconds
await player.setClip(start: Duration(seconds: 2), end: Duration(seconds: 4));
await player.play(); await player.pause();
await player.setClip(start: Duration(seconds: 10), end: Duration(seconds: 12));
await player.play(); await player.pause();

await player.setClip(); // Clear clip region

Working with gapless playlists

// Define the playlist
final playlist = ConcatenatingAudioSource(
  // Start loading next item just before reaching it
  useLazyPreparation: true,
  // Customise the shuffle algorithm
  shuffleOrder: DefaultShuffleOrder(),
  // Specify the playlist items
  children: [
    AudioSource.uri(Uri.parse('https://example.com/track1.mp3')),
    AudioSource.uri(Uri.parse('https://example.com/track2.mp3')),
    AudioSource.uri(Uri.parse('https://example.com/track3.mp3')),
  ],
);

// Load and play the playlist
await player.setAudioSource(playlist, initialIndex: 0, initialPosition: Duration.zero);
await player.seekToNext();                     // Skip to the next item
await player.seekToPrevious();                 // Skip to the previous item
await player.seek(Duration.zero, index: 2);    // Skip to the start of track3.mp3
await player.setLoopMode(LoopMode.all);        // Set playlist to loop (off|all|one)
await player.setShuffleModeEnabled(true);      // Shuffle playlist order (true|false)

// Update the playlist
await playlist.add(newChild1);
await playlist.insert(3, newChild2);
await playlist.removeAt(3);

Working with headers

// Setting the HTTP user agent
final player = AudioPlayer(
  userAgent: 'myradioapp/1.0 (Linux;Android 11) https://myradioapp.com',
  useProxyForRequestHeaders: true, // default
);

// Setting request headers
final duration = await player.setUrl('https://foo.com/bar.mp3',
    headers: {'header1': 'value1', 'header2': 'value2'});

Note: By default, headers are implemented via a local HTTP proxy which on Android, iOS and macOS requires non-HTTPS support to be enabled. See Platform Specific Configuration.

Alternatively, settings useProxyForRequestHeaders: false will use the platform's native headers implementation without a proxy. Although note that iOS doesn't offer an official native API for setting headers, and so this will use the undocumented AVURLAssetHTTPHeaderFieldsKey API (or in the case of the user-agent header on iOS 16 and above, the official AVURLAssetHTTPUserAgentKey API).

Working with caches

// Clear the asset cache directory
await AudioPlayer.clearAssetCache();

// Download and cache audio while playing it (experimental)
final audioSource = LockCachingAudioSource('https://foo.com/bar.mp3');
await player.setAudioSource(audioSource);
// Delete the cached file
await audioSource.clearCache();

Note: LockCachingAudioSource is implemented via a local HTTP proxy which on Android, iOS and macOS requires non-HTTPS support to be enabled. See Platform Specific Configuration.

Working with stream audio sources

// Feed your own stream of bytes into the player
class MyCustomSource extends StreamAudioSource {
  final List<int> bytes;
  MyCustomSource(this.bytes);
  
  @override
  Future<StreamAudioResponse> request([int? start, int? end]) async {
    start ??= 0;
    end ??= bytes.length;
    return StreamAudioResponse(
      sourceLength: bytes.length,
      contentLength: end - start,
      offset: start,
      stream: Stream.value(bytes.sublist(start, end)),
      contentType: 'audio/mpeg',
    );
  }
}

await player.setAudioSource(MyCustomSource());
player.play();

Note: StreamAudioSource is implemented via a local HTTP proxy which on Android, iOS and macOS requires non-HTTPS support to be enabled. See Platform Specific Configuration.

Working with errors

// Catching errors at load time
try {
  await player.setUrl("https://s3.amazonaws.com/404-file.mp3");
} on PlayerException catch (e) {
  // iOS/macOS: maps to NSError.code
  // Android: maps to ExoPlayerException.type
  // Web: maps to MediaError.code
  // Linux/Windows: maps to PlayerErrorCode.index
  print("Error code: ${e.code}");
  // iOS/macOS: maps to NSError.localizedDescription
  // Android: maps to ExoPlaybackException.getMessage()
  // Web/Linux: a generic message
  // Windows: MediaPlayerError.message
  print("Error message: ${e.message}");
} on PlayerInterruptedException catch (e) {
  // This call was interrupted since another audio source was loaded or the
  // player was stopped or disposed before this audio source could complete
  // loading.
  print("Connection aborted: ${e.message}");
} catch (e) {
  // Fallback for all other errors
  print('An error occured: $e');
}

// Catching errors during playback (e.g. lost network connection)
player.playbackEventStream.listen((event) {}, onError: (Object e, StackTrace st) {
  if (e is PlayerException) {
    print('Error code: ${e.code}');
    print('Error message: ${e.message}');
  } else {
    print('An error occurred: $e');
  }
});

Working with state streams

See The state model for details.

player.playerStateStream.listen((state) {
  if (state.playing) ... else ...
  switch (state.processingState) {
    case ProcessingState.idle: ...
    case ProcessingState.loading: ...
    case ProcessingState.buffering: ...
    case ProcessingState.ready: ...
    case ProcessingState.completed: ...
  }
});

// See also:
// - durationStream
// - positionStream
// - bufferedPositionStream
// - sequenceStateStream
// - sequenceStream
// - currentIndexStream
// - icyMetadataStream
// - playingStream
// - processingStateStream
// - loopModeStream
// - shuffleModeEnabledStream
// - volumeStream
// - speedStream
// - playbackEventStream

Credits

This project is supported by the amazing open source community of GitHub contributors and sponsors. Thank you!

Platform specific configuration

Android

To allow your application to access audio files on the Internet, add the following permission to your AndroidManifest.xml file:

    <uses-permission android:name="android.permission.INTERNET"/>

If you wish to connect to non-HTTPS URLs (typically HTTP), also add the following attribute to the application element:

    <application ... android:usesCleartextTraffic="true">

Note that just_audio's proxy (used to implement features such as headers, caching and stream audio sources) runs on a localhost HTTP server, and this also requires cleartext access to be enabled. You can either enable this via the option above which also enables access to any non-HTTPS URL, or you can instead limit cleartext access to just localhost URLs by defining a network security config. To use this approach, create the file android/app/src/main/res/xml/network_security_config.xml:

<?xml version="1.0" encoding="utf-8"?>
<network-security-config>
	<domain-config cleartextTrafficPermitted="true">
		<domain includeSubdomains="false">127.0.0.1</domain>
	</domain-config>
</network-security-config>

Then reference this network security config in your AndroidManifest.xml file by adding the following attribute to the application element:

    <application ... android:networkSecurityConfig="@xml/network_security_config">

If you need access to the player's AudioSession ID, you can listen to AudioPlayer.androidAudioSessionIdStream. Note that the AudioSession ID will change whenever you set new AudioAttributes.

If there are multiple plugins in your app that use ExoPlayer to decode media, it is possible to encounter a Duplicate class error if those plugins use different versions of ExoPlayer. In this case you may report an issue for each respective plugin to upgrade to the latest version of ExoPlayer, or you may downgrade one or more of your app's plugins until the versions match. In some cases where a plugin uses non-breaking parts of the ExoPlayer API, you can also try forcing all plugins to use the same version of ExoPlayer by editing your own app's android/app/build.gradle file and inserting the dependencies for the desired Exoplayer version:

dependencies {
    def exoplayer_version = "...specify-version-here...."
    implementation "com.google.android.exoplayer:exoplayer-core:$exoplayer_version"
    implementation "com.google.android.exoplayer:exoplayer-dash:$exoplayer_version"
    implementation "com.google.android.exoplayer:exoplayer-hls:$exoplayer_version"
    implementation "com.google.android.exoplayer:exoplayer-smoothstreaming:$exoplayer_version"
}

iOS

Using the default configuration, the App Store will detect that your app uses the AVAudioSession API which includes a microphone API, and for privacy reasons it will ask you to describe your app's usage of the microphone. If your app does indeed use the microphone, you can describe your usage by editing the Info.plist file as follows:

<key>NSMicrophoneUsageDescription</key>
<string>... explain why the app uses the microphone here ...</string>

But if your app does not use the microphone, you can pass a build option to "compile out" any microphone code so that the App Store won't ask for the above usage description. To do so, edit your ios/Podfile as follows:

post_install do |installer|
  installer.pods_project.targets.each do |target|
    flutter_additional_ios_build_settings(target)
    
    # ADD THE NEXT SECTION
    target.build_configurations.each do |config|
      config.build_settings['GCC_PREPROCESSOR_DEFINITIONS'] ||= [
        '$(inherited)',
        'AUDIO_SESSION_MICROPHONE=0'
      ]
    end
    
  end
end

If you wish to connect to non-HTTPS URLs, or if you use a feature that depends on the proxy such as headers, caching or stream audio sources, add the following to your Info.plist file:

<key>NSAppTransportSecurity</key>
<dict>
    <key>NSAllowsArbitraryLoads</key>
    <true/>
</dict>

The iOS player relies on server headers (e.g. Content-Type, Content-Length and byte range requests) to know how to decode the file and where applicable to report its duration. In the case of files, iOS relies on the file extension.

macOS

To allow your macOS application to access audio files on the Internet, add the following to your DebugProfile.entitlements and Release.entitlements files:

    <key>com.apple.security.network.client</key>
    <true/>

If you wish to connect to non-HTTPS URLs, or if you use a feature that depends on the proxy such as headers, caching or stream audio sources, add the following to your Info.plist file:

<key>NSAppTransportSecurity</key>
<dict>
    <key>NSAllowsArbitraryLoads</key>
    <true/>
</dict>

The macOS player relies on server headers (e.g. Content-Type, Content-Length and byte range requests) to know how to decode the file and where applicable to report its duration. In the case of files, macOS relies on the file extension.

Windows

Windows support is enabled by adding an additional dependency to your pubspec.yaml alongside just_audio. There are a number of alternative options:

Example:

dependencies:
  just_audio: any # substitute version number
  just_audio_windows: any # substitute version number

For issues with the Windows implementation, please open an issue on the respective implementation's GitHub issues page.

Linux

Linux support is enabled by adding an additional dependency to your pubspec.yaml alongside just_audio. There are a number of alternative options:

dependencies:
  just_audio: any # substitute version number
  just_audio_mpv: any # substitute version number

For issues with the Linux implementation, please open an issue on the respective implementation's GitHub issues page.

Troubleshooting

Most problems you encounter when playing an audio file will likely relate to the audio file format, the server headers, or the file name.

Audio file formats/encodings

Different platforms support different audio formats and encodings. For a list, see this StackOverflow answer.

Different audio formats have different seeking support. If you have control over the encoding of the audio files your app needs to play, then it is recommended to encode audio as M4A because this is capable of embedding an accurate seek table. With MP3, there are multiple different methods of encoding that permit seeking, although all such methods are approximate. When your app needs to play from arbitrary audio sources requested by the user, you are at the mercy of the source audio format.

Different audio formats may or may not embed duration metadata, and the absence of this metadata is a common reason why just_audio may sometimes return a null duration. This behaviour is platform specific, however, and Android can sometimes infer the missing duration by decoding the entire file and measuring the duration.

Server headers

A server that hosts audio content should return appropriate HTTP response headers to the client. This includes an appropriate Content-Length header and a correct Content-Type header so that the player knows which decoder to use to read the data.

Servers should also support range requests. These allow just_audio to make requests for a part of the whole file within a given byte range as opposed to always requesting the whole file. If the user seeks to a position near the end of the audio file that hasn't been downloaded yet (and if there is a seek table), just_audio will try to make a range request for the end of the file.

Range requests can also impact on just_audio's ability to determine an audio file's duration. In many cases, certain audio formats embed metadata (which includes the audio duration) at the END of the audio file, and range requests allow just_audio to jump to the end of the file to fetch the metadata first, without having to wait for the entire file to download, and then inform you up front what the audio file duration is. If just_audio returns a null duration when the audio file has duration metadata, this may suggest that the server does not support range requests.

If you host audio files on your own server, remember to correctly configure the headers described above.

File names

When playing audio from a local file, just_audio cannot use any Content-Type header to figure out which decoder to use to read the file. Instead, the file name extension will generally be used to determine the file type. For example, if a file name ends with .mp3, the MP3 decoder will be used to read it, while if the file name ends with .wav, the WAV decoder will be used. If the file name has an .mp3 extension but the actual file content is not in the MP3 format, then just_audio may potentially fail to read it. iOS enforces this fairly strictly, while Android is more likely to be forgiving by taking a peek at the data to make an educated guess as to what the data format really is. In general, however, it is recommended to use correct file name extensions for any audio file that is under your control.

Mixing and matching audio plugins

The flutter plugin ecosystem contains a wide variety of useful audio plugins. In order to allow these to work together in a single app, just_audio "just" plays audio. By focusing on a single responsibility, different audio plugins can safely work together without overlapping responsibilities causing runtime conflicts.

Other common audio capabilities are optionally provided by separate plugins:

  • just_audio_background: Use this to allow your app to play audio in the background and respond to controls on the lockscreen, media notification, headset, AndroidAuto/CarPlay or smart watch.
  • audio_service: Use this if your app has more advanced background audio requirements than can be supported by just_audio_background.
  • audio_session: Use this to configure and manage how your app interacts with other audio apps (e.g. phone call or navigator interruptions).
  • just_waveform: Use this to extract an audio file's waveform suitable for visual rendering.

Tutorials

Vote on upcoming features

Press the thumbs up icon on the GitHub issues you would like to vote on:

Please also consider pressing the thumbs up button at the top of this page (pub-web.flutter-io.cn) if you would like to bring more momentum to the project. More users leads to more bug reports and feature requests, which leads to increased stability and functionality.

Platform support

Feature Android iOS macOS Web Windows Linux
read from URL
read from file
read from asset
read from byte stream
request headers
DASH
HLS
ICY metadata
buffer status/position
play/pause/seek
set volume/speed
clip audio
playlists
looping/shuffling
compose audio
gapless playback
report player errors
handle phonecall interruptions
buffering/loading options
set pitch
skip silence
equalizer
volume boost

Experimental features

Feature Android iOS macOS Web
Simultaneous downloading+caching
Waveform visualizer (See #97)
FFT visualizer (See #97)
Background

Please consider reporting any bugs you encounter here or submitting pull requests here.

The state model

The state of the player consists of two orthogonal states: playing and processingState. The playing state typically maps to the app's play/pause button and only ever changes in response to direct method calls by the app. By contrast, processingState reflects the state of the underlying audio decoder and can change both in response to method calls by the app and also in response to events occurring asynchronously within the audio processing pipeline. The following diagram depicts the valid state transitions:

just_audio_states

This state model provides a flexible way to capture different combinations of states such as playing+buffering vs paused+buffering, and this allows state to be more accurately represented in an app's UI. It is important to understand that even when playing == true, no sound will actually be audible unless processingState == ready which indicates that the buffers are filled and ready to play. This makes intuitive sense when imagining the playing state as mapping onto an app's play/pause button:

  • When the user presses "play" to start a new track, the button will immediately reflect the "playing" state change although there will be a few moments of silence while the audio is loading (while processingState == loading) but once the buffers are finally filled (i.e. processingState == ready), audio playback will begin.
  • When buffering occurs during playback (e.g. due to a slow network connection), the app's play/pause button remains in the playing state, although temporarily no sound will be audible while processingState == buffering. Sound will be audible again as soon as the buffers are filled again and processingState == ready.
  • When playback reaches the end of the audio stream, the player remains in the playing state with the seek bar positioned at the end of the track. No sound will be audible until the app seeks to an earlier point in the stream. Some apps may choose to display a "replay" button in place of the play/pause button at this point, which calls seek(Duration.zero). When clicked, playback will automatically continue from the seek point (because it was never paused in the first place). Other apps may instead wish to listen for the processingState == completed event and programmatically pause and rewind the audio at that point.

Apps that wish to react to both orthogonal states through a single combined stream may listen to playerStateStream. This stream will emit events that contain the latest value of both playing and processingState.

Configuring the audio session

If your app uses audio, you should tell the operating system what kind of usage scenario your app has and how your app will interact with other audio apps on the device. Different audio apps often have unique requirements. For example, when a navigator app speaks driving instructions, a music player should duck its audio while a podcast player should pause its audio. Depending on which one of these three apps you are building, you will need to configure your app's audio settings and callbacks to appropriately handle these interactions.

just_audio will by default choose settings that are appropriate for a music player app which means that it will automatically duck audio when a navigator starts speaking, but should pause when a phone call or another music player starts. If you are building a podcast player or audio book reader, this behaviour would not be appropriate. While the user may be able to comprehend the navigator instructions while ducked music is playing in the background, it would be much more difficult to understand the navigator instructions while simultaneously listening to an audio book or podcast.

You can use the audio_session package to change the default audio session configuration for your app. E.g. for a podcast player, you may use:

final session = await AudioSession.instance;
await session.configure(AudioSessionConfiguration.speech());

Note: If your app uses a number of different audio plugins, e.g. for audio recording, or text to speech, or background audio, it is possible that those plugins may internally override each other's audio session settings, so it is recommended that you apply your own preferred configuration using audio_session after all other audio plugins have loaded. You may consider asking the developer of each audio plugin you use to provide an option to not overwrite these global settings and allow them be managed externally.

Libraries

just_audio