cerkit ClearCast: AI-Powered Radio Transcription

Submit Comments or Ideas

A Brief History

Created by KR4CEI, cerkit ClearCast is a project born from the desire to seamlessly bridge amateur radio hardware with cutting-edge artificial intelligence. The goal was to capture live, raw audio from a radio receiver and transcribe it into readable text in real-time—effectively pulling a clear text signal out of the noise. Initially developed for macOS and Windows using .NET 10 and the Avalonia UI framework, the project aggressively evolved to provide a robust, universal tool for radio enthusiasts everywhere.

True Cross-Platform Compatibility

One of the standout achievements of cerkit ClearCast is its native, cross-platform architecture. Whether running a dedicated desktop in a ham shack or a lightweight portable rig in the field, this application was designed to serve you anywhere. It natively supports running as a standalone executable on:

  • Apple Silicon macOS
  • Windows
  • Linux
  • Raspberry Pi (Linux-arm64)

These single-file executables bundle everything needed to run, meaning you don’t even need to pre-install the .NET runtime on the target machine!

Current Features

  • 🎙️ Live Audio Capture: Stream incoming audio directly from any connected input device, such as a Focusrite Scarlett 18i8 or a simple USB audio interface (SignaLink, DigiRig, or HamLink).
  • 🤖 AI-Powered Transcription: The app slices audio into manageable chunks and feeds them to the Google Gemini API for fast, highly accurate speech-to-text recognition.
  • 📝 Real-Time Display: Transcribed text appears instantly on-screen as it is recognized.
  • 🆔 Call Sign Detection: Automatically detects and abbreviates amateur radio call signs and NATO phonetics.
  • 🌐 QRZ API Integration: Securely authenticates and looks up the state and country of detected call signs.
  • 📋 Tracked Call Signs List: Maintains a running list of unique call signs detected during the session, augmented with location data.
  • 📒 Auto-Logging: Saves session transcripts and alphabetical lists of detected call signs to a text file (e.g., log-[TIMESTAMP].txt), ensuring no communications are lost—even if the app is abruptly closed.
  • 🔇 Configurable Silence Detection: Features an adjustable threshold to ignore quiet periods. This fundamentally skips “dead air,” significantly reducing API usage and costs.
  • 🔊 Audio Monitor: Includes an optional playback mode for the captured audio, allowing you to easily verify your signal routing.
  • 💾 Persistent Settings: Your Google Gemini API key, QRZ credentials, selected audio device, and silence thresholds are securely persisted between app sessions.

Requirements & Setup

Satisfying Google AI Studio Requirements

Because cerkit ClearCast uses the lightning-fast Google Gemini API for speech-to-text processing, you’ll need to supply the application with an API key.

  1. Navigate to Google AI Studio.
  2. Sign in using your Google account.
  3. Select Get API key from the navigation menu.
  4. Click Create API key, generate a new key, and copy it to your clipboard.

Running the Application

  1. Download the executable completely tailored to your operating system (macOS, Windows, Linux, or Raspberry Pi) and launch it.
  2. Click File -> Settings… to open the configuration dialog.
  3. Enter your Google Gemini API key and any QRZ credentials (optional).
  4. Select your audio input device (the radio receiver source) and channel from the Audio Settings section.
  5. Click Save to apply your settings securely.
  6. On the main window, adjust the silence threshold slider to an appropriate level to ensure background static isn’t unnecessarily transcribed.
  7. Click the START button and watch the transcriptions roll in!

Audio Interfaces to Connect your Radio and your Computer

As an Amazon Associate, I earn from qualifying purchases