The Whisper plugin is a revolutionary addition to Obsidian, allowing users to effortlessly turn their speech into written notes. With the power of OpenAI's Whisper API, this plugin can recognize and transcribe audio recordings with remarkable accuracy. Users can initiate recording sessions using the ribbon button or keyboard shortcut (Alt + Q), and then pause, resume, or stop the recording as needed. The plugin also supports uploading existing audio files for transcription. Settings allow users to customize their experience, including selecting the language of the message being whispered and choosing where to save transcriptions and audio files. With Whisper, Obsidian users can now focus on capturing their thoughts without worrying about typing them out manually.
Improvements
- Reorganize settings into logical sections: API Keys, Transcription, Recording, Output, Post-processing, Advanced
- Hide section sub-fields until their parent toggle is enabled
- "Save audio file" and "Create note file" are now on by default
- Preserve scroll position when toggling settings so the panel no longer jumps
Bug Fixes
- Fix bug where "Note save path" stayed editable when "Create note file" was off