0.6.0

Voice Receive

v0.6.0 makes voice receive useable with many utility additions. You can see a run-down below, or view the full recorder example.

Audio receive streams automatically close when their attached voice connection is destroyed.

// Join with selfDeaf set to false to receive audio
const voiceConnection = joinChannel({ ...options, selfDeaf: false });

voiceConnection.receiver.speaking.on('start', userId => console.log(`User ${userId} started speaking`));
voiceConnection.receiver.speaking.on('end', userId => console.log(`User ${userId} stopped speaking`));

// A Readable object mode stream of Opus packets
// Will only end when the voice connection is destroyed
voiceConnection.receiver.subscribe(userId);

// A Readable object mode stream of Opus packets
// Will end when the voice connection is destroyed, or the user has not said anything for 100ms
const opusStream = voiceConnection.receiver.subscribe(userId, {
  end: {
    behavior: EndBehaviorType.AfterSilence,
    duration: 100,
  },
});

// -------------------------

// You can re-use the Opus stream, e.g. to play in another channel as an echo
const resource = createAudioResource(opusStream, { inputType: StreamType.Opus });

// You can decode the Opus stream to raw audio using prism-media
const rawAudio = opusStream.pipe(new prism.opus.Decoder({ frameSize: 960, channels: 2, rate: 48000 }));

// You can save the stream to an Ogg file using prism-media@2.0.0-alpha.0
const oggWriter = new opus.OggLogicalBitstream({
  opusHead: new opus.OpusHead({
    channelCount: 2,
    sampleRate: 48000,
  }),
  pageSizeControl: {
    maxPackets: 10,
  },
});

pipeline(opusStream, oggWriter, createWriteStream('./myfile.ogg'), callback);

---

Commit log

• chore: update conventional commit URL (#167)

• test(DataStore): add missing tests (#165)

  • tests(DataStore): add missing tests

  • test(DataStore): add function spys for frame prep

• refactor(AudioPlayer): use this.once instead of events.once (#162)

• feat(Util): support AbortSignal in entersState (#171)

  • refactor(Util): use AbortSignal in entersState

  • chore(CI): drop Node v14 from CI

  • refactor(Util): remove unnecessary function

  • test(Util): add tests for abortAfter

  • test(Util): add test for entersState

• refactor(Examples): remove unused imported binding (#172)

• feat(VoiceReceive)!: improve usability (#136)

  • refactor(VoiceReceiver): begin refactor

  • feat(SSRCMap): resolve function

  • feat(VoiceConnection): close all streams in non-ready state

  • refactor(VoiceReceiver): map by user ID

  • feat(VoiceReceiver): allow specifying end type for streams

  • feat(VoiceReceiver): add SpeakingMap

  • refactor(SSRCMap): remove unused resolve method

  • test(VoiceReceiver): add test for SpeakingMap

  • test(VoiceReceiver): add tests for AudioReceiveStream

  • test(AudioReceiver): strengthen AudioReceiveStream tests

  • test(VoiceReceiver): remove inapplicable tests

  • test(VoiceConnection): fix test errors

  • test(VoiceConnection): test receiver bindings tracking

  • test(VoiceReceiver): decrypt

  • chore: remove unused code

  • fix(AudioReceiveStream): close normally

  • feat(Examples): update receiver example

  • feat(Examples): create recorder example

  • docs(VoiceReceiver): add docs for receive classes

  • refactor: suggestions from code review

  Co-authored-by: Antonio Román kyradiscord@gmail.com

  Co-authored-by: Antonio Román kyradiscord@gmail.com

• chore(Release): v0.6.0