Version 0.0.1a5

Released December 19th, 2025

Added

• Result returned from all AudioPlayer methods, making it easy to implement further functionality.
• New is_playing property in AudioPlayer.
• Further global type checking to ensure proper data is being passed.
• New VoiceClient.move method to move channels easily.
• FileAudioSource raises FileNotFoundError if the provided filepath doesn't exist.

Changed

• Refactored file locations for further readability and organization.

Fixed

• Made VoiceGateway and VoiceServer connection methods idempotent, just in case they are re-called.
• Added guards on many methods to ensure object states are consistent and compliant.
• FFmpeg encoder/decoder is now localized to each VoiceConnection/AudioPlayer, allowing multiple voice connections with playback.

Removed

• Convenience methods under VoiceConnection like play, pause, etc. in favor of clarity with player access.

Version 0.0.1a4

Released December 19th, 2025

Added

• New AudioSources:
  1. BufferAudioSource: Allows bytes-like objects to be encoded directly.
  2. URLAudioSource: Allows the encoding of audio hosted at an endpoint on the web.
• __slots__ to all objects for memory efficiency.
• Implemented Discord's Audio/Video End-to-End Encryption protocol (DAVE) opcodes and miscelleanous logic to prepare for transition - required by March 1st, 2026.
• Support for aead_aes256_gcm_rtpsize encryption - automatically searches for best encryption mode supported.

Changed

• AudioSource implementations no longer require read or stream methods.
• Subsequently, FileAudioSource no longer requires aiofiles as a dependency; removed from project.
• Network objects (VoiceServer, VoiceGateway) moved to networking subfolder.

Fixed

• FileAudioSource using incorrect, public filepath for internal checks rather than private attribute.
• VoiceGateway internal gateway latency calculation not returning due to reference bug.

Removed

• VoiceConnection add_queue_file, remove_queue_file, and play_file removed in favor of generic add_queue, remove_queue, and play.

Version 0.0.1a3

Released December 17th, 2025

Added

• AudioPlayer.shuffle() method to shuffle the queue, as well as a corresponding VoiceConnection.shuffle() for convenience.

Changes

• Clamped dependency versions for compatibility (Thanks @vicky5124 for reporting this issue).
• Switched audio dependency from opuslib to be fully FFmpeg-dependent; minor lag improvement (Thanks @vicky5124 for recommending).
• Refactored internal client cache for slightly better lookup performance and readability.

Fixed

• VoiceReconnectEvent constructor docstring return type fixed from VoicePopulatedEvent.
• MemberJoinVoiceEvent, MemberLeaveVoiceEvent, MemberLeaveVoiceEvent, MemberDeafEvent, and MemberMuteEvent now dispatch regardless if the bot is in the channel with them.

Removed

• VoiceActiveEvent, VoiceInactiveEvent, VoicePopulatedEvent and VoiceEmptyEvent as they were redundant of logic from other events.

Version 0.0.1a2

Released December 17th, 2025

Added

• VoiceConnection.play method for completeness.
• New AudioBeginEvent and AudioEndEvent.
• New audio queue system:
  1. Player:
     • previous, next to scan forward and backward in the queue/history.
     • add_queue to add sources to the queue.
     • remove_queue, clear_queue to remove/clear sources from the queue.
     • queue property to view the queue, history property to view history.
  2. VoiceConnection (for convenience):
     • Corresponding previous, next.
     • Corresponding add_queue and additional add_queue_file.
     • Corresponding remove_queue, clear_queue and additional remove_queue_file.

Fixed

• Made all object attributes private, with public attributes made as properties.

Version 0.0.1a1

Released December 16th, 2025

Original alpha release.