Voice Notes

A simple desktop app (Tkinter) to record voice notes, transcribe them with Whisper, and save them as markdown files. For full Windows setup details, see Prerequisites.md.

Requirements

• Windows
• Python 3.11+ (project uses a .venv at .venv/)
• FFmpeg installed and bin on PATH (the app expects C:\ffmpeg\bin by default; edit FFMPEG_BIN in voicenotes_app.py if different)
• Microphone input device

Setup

1. Create and activate the virtual environment (if not already):

python -m venv .venv
".venv\\Scripts\\activate"

2. Install dependencies:

pip install -r requirements.txt

3. Ensure FFmpeg is reachable on PATH, or adjust FFMPEG_BIN in voicenotes_app.py.

Running the app

• From Explorer: double-click run_voicenotes.bat (keeps the console open so you can see errors if any).
• From terminal:

".venv\\Scripts\\python.exe" voicenotes_app.py

Usage

1. On first launch, pick your language, notes folder, and microphone.
2. Click Record to capture audio; transcription inserts into the text box.
3. Click Save to write a timestamped markdown note into your selected folder.

Notes

• Whisper model loads on startup (base model by default) in a background thread.
• Recording uses 16 kHz mono for best Whisper compatibility.
• Temp WAV files are cleaned up after transcription.