Audio visualization engine · Linux/Wayland/Hyprland
Kitsune
Audio visualization engine for Linux (Wayland/Hyprland) featuring modular architecture, multi-monitor support and automatic CPU/GPU fallback.
Base pipeline
PulseAudio/PipeWire -> CAVA -> Rust Renderer -> RGBA FIFO -> mpvpaper/layer-shell
Status and classification
Overview
Kitsune is a Linux desktop audio visualization engine focused on Wayland/Hyprland and designed for daily production use.
Its decoupled architecture captures audio from PulseAudio/PipeWire via CAVA, processes it in a Rust-based renderer (CPU or GPU), and outputs RGBA frames through FIFO to mpvpaper or a custom layer-shell binary.
The system is built with fault tolerance and true multi-monitor support, allowing isolated instances per display with independent configuration (FIFOs, PIDs, logs and runtime).
Kitsune serves as the graphical core of the Kitotsu ecosystem and foundation for future integration with Kitowall and advanced rendering pipelines.
Components and capabilities
- Stable CPU backend baseline
- Optional GPU backend with automatic fallback
- Single and group modes (multi-layer alpha composition)
- Independent DSP and visual profiles
- PostFX (blur + glow)
- Spectrum-reactive particles
- Dynamic wallpaper-based accent color
- Full CLI lifecycle: start/stop/status/doctor/config
- Hot-reload test runtime
Role and responsibilities
- Full modular architecture design
- Rust renderer implementation (CPU/GPU)
- Multi-monitor isolated instance system
- Full CLI lifecycle and diagnostics implementation
- DSP/visual profile system design
- Alpha layer composition engine
- Wayland and layer-shell integration
- Automatic backend fallback system
Tech stack
Links
Public demo: Not available
Gallery
Application screenshot:
