SystemSculpt AI

Recorder Module - Settings and Configuration

The Recorder Module offers various customizable settings to tailor the recording and transcription experience to your needs.

Key Settings

  • Recordings Path: Set the storage location for audio recordings.
  • Transcriptions Path: Set the storage location for transcription files.
  • Auto-transcription: Toggle automatic transcription of recordings.
  • Whisper Provider: Choose between OpenAI and Groq for transcription.
  • Whisper Model: Select the Whisper model for OpenAI transcription.
  • Microphone Selection: Choose the preferred microphone for recordings.
  • Save Audio Clips: Toggle whether to save audio files after transcription.
  • Paste into Active Note: Automatically paste transcriptions into the active note.
  • Copy to Clipboard: Automatically copy transcriptions to the clipboard.
  • Show Recorder Button on Status Bar: Toggle visibility of the recorder button.
  • Transcription Language: Select the language of the audio for transcription (English, Spanish, French, German)

Configuration Steps

  1. Go to Settings > SystemSculpt AI > Recorder.
  2. Adjust the settings according to your preferences.
  3. Ensure your OpenAI or Groq API key is set in the Brain module settings.

Whisper Provider Configuration

  1. Select your preferred Whisper provider (OpenAI or Groq) in the Recorder settings.
  2. If using OpenAI, select the desired Whisper model.
  3. Ensure the appropriate API key is set in the Brain module settings.

Storage Configuration

  1. Set the "Recordings path" to your preferred location for audio files.
  2. Set the "Transcriptions path" to your preferred location for transcription files.
  3. Toggle "Save audio clips" based on whether you want to keep audio files after transcription.

Integration Settings

  1. Toggle "Paste into active note" to automatically insert transcriptions into your current note.
  2. Toggle "Copy to clipboard" to automatically copy transcriptions for easy pasting elsewhere.
  3. Toggle "Copy to clipboard" to automatically copy transcriptions for easy pasting elsewhere.

Status Bar Configuration

  1. Toggle "Show recorder button on status bar" to control the visibility of the quick access button.

Best Practices

  1. Regularly review and update your settings to optimize your workflow.
  2. Experiment with different Whisper models to find the best balance of accuracy and speed.
  3. Organize your recordings and transcriptions in dedicated folders for better file management.

Transcription Settings

Custom Whisper Prompt

The custom Whisper prompt allows you to improve transcription accuracy:

  • Enable/disable custom transcription prompt in settings
  • Maximum 244 tokens per prompt (with real-time token counter)
  • Helps with:
    • Correcting specific words or acronyms
    • Preserving context for split audio files
    • Ensuring proper punctuation and filler words
    • Specifying preferred writing styles for different languages

Post-Processing Prompt

The post-processing prompt uses GPT models to enhance transcription quality:

  • Enable/disable post-processing in settings
  • Uses GPT models for advanced improvements:
    • Correcting misspellings
    • Improving overall accuracy
    • Enhancing readability
    • Handling larger context windows for better understanding

Best Practices

  1. Monitor token usage for Whisper prompts using the built-in counter
  2. Keep prompts focused and specific to your needs
  3. Use post-processing for more complex formatting requirements
  4. Test different prompt combinations to find the optimal setup for your use case
  1. Include link to recording: When enabled, automatically adds a link to the original audio file at the top of each transcription
    • Uses Obsidian's internal linking format
    • Helps maintain connection between transcriptions and source recordings
    • Enabled by default