A voice transcriptor for SDR

Using GPU as AI transcriptor (CUDA if available), defaults to CPU usage instead if CUDA is not available

Prequisites: Python 3.8+

Make sure you have the following installed:

BASH: pip install torch numpy sounddevice scipy openai-whisper

Ubuntu/Debian: sudo apt update && sudo apt install ffmpeg

macOS: brew install ffmpeg

Hardware & Audio Notes Make sure your microphone input device is set correctly (defaults to device=1, change as needed).

Works best with 48kHz input, automatically resampled to 16kHz for Whisper compatibility.

You should see:

“🎧 Listening for voice...”

“🎤 Voice started...” when speech is detected

Transcribed results printed and saved to a daily log file

Knows issues: Transcriber does not always capture voice and comes up blank in the logs Repeated words recorded

💡 Customization Ideas Swap Whisper model size: "tiny", "base", "small", "medium", "large"

👨‍💻 Acknowledgements Huge thanks to ChatGPT (a.k.a. the whispering code goblin) for helping architect, debug, and document this project like a caffeinated co-pilot. 😄

This project is licensed under the MIT License.