Wing Mixer Voice Control A voice-controlled interface for Behringer Wing digital mixer, allowing hands-free control through natural language commands. The system uses Rasa for intent recognition and Google's Gemini AI to process and refine voice commands. 🎚️ Features

Natural Language Processing: Control your mixer with conversational commands Multiple Command Types:

Channel selection, muting, and unmuting Fader level adjustments Gain control Phantom power toggling Solo and gate controls EQ and compressor activation Clear all solos

Intelligent Command Interpretation: Understands variations in command phrasing Audio Knowledge Base: Ask general audio engineering questions through the integrated AI assistant

🔧 System Architecture The system consists of the following components:

Rasa NLU: For intent classification and entity extraction Command Interpreter: Processes and translates natural language into structured mixer commands OSC Communication Layer: Sends commands to the Wing console via OSC protocol Google Gemini AI: Refines ambiguous voice commands and provides audio knowledge responses

📋 Requirements

Python 3.8+ Rasa (for NLU) google-generativeai spaCy with English language model (en_core_web_sm) pythonosc word2number Behringer Wing digital mixer with network connectivity

💻 Setup

Clone this repository: Copygit clone https://github.com/yourusername/wing-mixer-voice-control.git cd wing-mixer-voice-control

Install the required dependencies: Copypip install -r requirements.txt python -m spacy download en_core_web_sm

Set up your environment variables: Copyexport GOOGLE_API_KEY="your_gemini_api_key"

Configure your Wing mixer IP and port in the code: pythonCopyWING_IP = '192.168.1.183' # Change to your Wing console IP TCP_PORT = 2222 # TCP port used by the Wing console UDP_PORT = 14135 # Port where you want to receive meter data OSC_PORT = 2223 # Port for sending OSC messages

Train the Rasa model: Copyrasa train

🎯 Usage

Start the Rasa server: Copyrasa run

In a separate terminal, start the actions server: Copyrasa run actions

Connect a microphone and start the voice input client (not included in this repo): Copypython voice_client.py

Speak commands such as:

"Mute channel 5" "Raise the fader on channel 1 by 3 dB" "Set the gain on channel 7 to 20 dB" "Enable phantom power on channel 3" "Solo channel 2" "Clear all solos" "Turn on the gate for channel 8" "Enable EQ on channel 4" "Turn on compressor for channel 6"

Ask audio-related questions:

"What's the best way to mic a drum kit?" "How should I set up compression for vocals?" "Explain the difference between shelving and peaking EQs"

🧠 Command Interpreter Logic The system processes commands through multiple stages:

Voice Command Validation: Using Gemini AI to clean up speech recognition results Command Type Detection: Identifying the action (mute, fader adjustment, etc.) Channel Extraction: Finding channel numbers in the command Parameter Extraction: Determining values for settings Command Execution: Translating to OSC messages for the mixer

📚 Classes and Functions Key components:

ActionControlMixer: Main Rasa action for interpreting and executing mixer commands ActionUseGemini: Handles general audio knowledge questions using Gemini AI CommandInterpreter: Parses natural language into structured commands send_osc_message(): Sends OSC messages to the Wing console OSC message utility functions for each mixer operation

🤝 Contributing Contributions are welcome! Please feel free to submit a Pull Request.

Fork the repository Create your feature branch (git checkout -b feature/AmazingFeature) Commit your changes (git commit -m 'Add some AmazingFeature') Push to the branch (git push origin feature/AmazingFeature) Open a Pull Request

📄 License This project is licensed under the MIT License - see the LICENSE file for details. 🙏 Acknowledgements

Behringer for the WING OSC implementation Google for Gemini AI The Rasa community