Play
Under the hood
-
Python
→ Nova’s brain is written in Python.
→ Handles the AI API calls (OpenAI, ElevenLabs), the weather pulling (OpenWeatherMap), random chatter, file generation, and error handling. -
Liquidsoap (DSL)
→ Nova’s radio control system is programmed in Liquidsoap’s scripting language.
→ It watches songs, counts plays, triggers Nova, injects her voice clips into the live audio stream. -
Bash (Shell Scripting)
→ Used inside Liquidsoap to spawn Python scripts asynchronously.
Version 4.1
How Nova Works:
-
🎶 Listens in Real-Time:
Nova monitors every track played live on Station1. -
🧠 Tracks the Flow:
After every three songs, Nova wakes up and prepares something special for you. -
✍️ Crafts a Line of Chatter:
Nova either pulls a random quote from her “Talking Shite Bank” (handwritten, chaotic one-liners)
— or she dynamically generates brand-new banter using OpenAI’s GPT models, Inane Chatter mod, adding a bit of intelligence and unpredictability. -
🌦️ Checks the Weather:
Nova taps into real-time weather reports for Dublin, so sometimes she’ll casually mention it in her chat. -
🗣️ Speaks Like a Real DJ:
Using ElevenLabs’ neural TTS (Text-to-Speech) engine, Nova transforms her thoughts into hyper-realistic voice clips. -
🚀 Injects Her Voice Into the Stream:
Her audio clips are dynamically detected and seamlessly inserted into the live broadcast, without skipping a beat. -
💡 Always Learning, Always On:
Nova operates fully automatically, requiring no human input — she's always there, always unpredictable, and always a little bit cheeky.- 💡 Changelog
# ———————- version 4.0 ———————- - # – changelog > Bug fixes / added backend server controller
- # – changelog > Revised core code / Local Saver Mode
- # – changelog > API saver mode / recycle shite vault
# – changelog > new a.i prompt
# – changelog > added volume booster
# – changelog > added multiple voices
- 💡 Changelog
Powered by: