DJ3000 Radio Project

---

Overview

The DJ3000 project was inspired when a team member was watching the Simpsons and was inspired by a futuristic robot introduced as a joke - the DJ3000. Built to replace the jobs of radio station hosts, the DJ3000 "plays CD's automatically" and even replaces the entertainment values of radio hosts with "three distinct varieties of inane chatter. 'Hey hey! What about that weather out there?' 'Woah... that was the caller from Hell' 'Well hot dog! We have a wiener!' ... 'Wow, looks like those clowns in Congress did it again!'" - The Simpsons, Bart Gets an Elephant. We have not only brought the futuristic machine to life, but have improved its humanness and made it small enough to fit in your pocket. This system uses Python, Pydub, ElevenLabs API, and a Raspberry Pi to create seamless, AI-driven radio shows that can be transmitted live over FM frequencies.

---

Features

• Playlist Management: DJ3000 shuffles a playlist of user-supplied music files.
• AI DJ Chatter: AI-generated back-and-forth chatter between DJ personalities, providing a humorous, lifelike experience ("inane chatter", news, weather).
• Song Transitions: Smooth crossfades between songs with DJ commentary.
• FM Broadcast: Generated .wav files are transmitted over FM frequencies using a Raspberry Pi.

---

How it Works

1. Playlist Generation: The user adds songs to a playlist. DJ3000 shuffles the songs and ensures no repeats occur.
2. AI Voices: The ElevenLabs API is used to generate DJ-style voice commentary, which is added between songs.
3. AI News Commentary: LLAMA 3.1B from Together.ai is used to generate summaries of news headlines and descriptions from BBC News.
4. Song Transitions: The system handles crossfading between songs and the insertion of AI-generated chatter.
5. WAV File Generation: DJ3000 exports the entire radio show as a .wav file.
6. FM Broadcast: Using fm_transmitter, the .wav file is broadcast via FM on a Raspberry Pi.

---

Demo Pre-Generated Files

• Intro Segment Demo
• Mid Segment Demo
• News Demo
• Weather Demo

Intro and Mid Segment may take some time to load as they are large files and have to be downloaded from s3 on the server side. Segments may be loud in google chrome web player

---

Installation

Requirements

• Python 3.x
• Pydub: For audio manipulation.
• ElevenLabs API: For AI voice generation.
• Together.ai: For AI Summaries
• OpenWeatherAPI: For localized weather
• Raspberry Pi: For FM broadcasting (Pi 4 reccomended).
• fm_transmitter: A tool to broadcast .wav files over FM using a Raspberry Pi.

Installation Steps

1. Clone the Repository:

   git clone https://github.com/thehomeworkclub/DJ3000/
   cd dj3000-radio

2. Install Dependencies:

   pip install -r requirements.txt

3. Set up ENV:

   • Obtain an API key from ElevenLabs.
   • Obtain an API key from Together.ai
   • Create a .env file in the project root and add your API key:

      ELEVEN_LABS_API_KEY=""
      TOGETHER_API_KEY=""
     

4. Install fm_transmitter on Raspberry Pi: Follow the installation steps in the fm_transmitter repository.

---

Usage

1. Add Songs: Place your .mp3 files into the Music folder.

2. Run the DJ3000 Script:

   python sequential.py

   This will generate a .wav file that contains the radio show.

3. Broadcast via FM:

   sudo ./fm_transmitter -f 100.1 segment_1.wav

   Replace 100.1 with your desired FM frequency.

---

File Structure

├── Music/                   # Place your .mp3 files here
├── chatter/                 # AI-generated DJ chatter files
├── sequential.py            # Main Python script
├── main.py                  # Handles AI voice generation and transitions
├── model.py                 # Handles database and segments
├── requirements.txt         # Python dependencies
├── .env                     # API keys and environment variables
└── README.md                # Project documentation

---

Example Query

Here’s an example of how the system generates a playlist and transitions between songs:

{
  getPlaylist {
    id
    title
    artist
    duration
    filePath
  }
}

---

Planned Features

• Enhanced Song Metadata: Use AI to announce the song title and artist in real-time.
• Real-time Playlist Updates: Modify the playlist while the broadcast is live.
• Improved FM Broadcasting: Enhance signal quality of the RaspberryPi FM broadcast.

---

Built With

• Python: For managing audio and running the core logic.
• Pydub: For slicing and mixing audio files.
• ElevenLabs API: For generating AI voice commentary.
• fm_transmitter: For broadcasting .wav files over FM via Raspberry Pi.

---

License

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