Orithet is a cutting-edge generative video art tool that seamlessly blends four conceptual systems to create mesmerizing, algorithmic visual experiences:
- PulseForge: Beat-driven, audio-reactive timing & pulses
- EchoMosaic: Semantic dream-like chaining via similarity graph + emotional flow
- GlitchGarden: Procedural ecosystem where clips act like living creatures that collide, fuse, mutate
- FractalFusion: Recursive picture-in-picture layers and infinite-depth echoes
Transform any collection of media files into hypnotic generative music videos or visual albums that feel like living digital organisms.
- Analyzes strongest audio file for BPM, beat grid, and energy bands
- All cuts and transitions lock to beat boundaries or sub-beats
- Energy bands: low/mid/high frequency analysis for dynamic visual responses
- Videos: PySceneDetect + extra random "memory cuts" every 3β12 seconds
- Images/GIFs: Variable-length clips (2β15 seconds) based on visual energy
- All clips become "creatures" with metadata (avg color palette, motion energy, dominant hue)
- Build a similarity graph (color + motion + audio mood)
- Lightweight 2D "ecosystem simulation" (30Γ30 grid) where creatures move, collide, and evolve
- Same-color creatures fuse (cross-fade + palette average)
- High-motion creatures "infect" others with glitch effects
- Random mutations (pixel sort, datamosh, RGB shift, mirror) are inherited
- At every major beat or collision event, spawn recursive PiP layers (up to depth 3)
- Each smaller layer shows a mutated version of the entire current composition
- Overlays pulse and scale with audio energy
- Beat-reactive zoom/pulse, color bleed between adjacent clips
- Echo trails, chromatic aberration, glitch probability
- Dream dissolves, vignette that reacts to bass
- Python 3.7+
- FFmpeg (required for video processing)
- Install Python 3.7+ from python.org
- Install FFmpeg:
- Download from ffmpeg.org
- Add FFmpeg to your system PATH
- Open Command Prompt and run:
git clone https://github.com/orithet/orithet.git
cd orithet
pip install -r requirements.txt- Install Python 3.7+ using Homebrew:
brew install python- Install FFmpeg:
brew install ffmpeg- Open Terminal and run:
git clone https://github.com/orithet/orithet.git
cd orithet
pip install -r requirements.txt- Install Python 3.7+ and FFmpeg:
sudo apt update
sudo apt install python3 python3-pip ffmpeg- Open Terminal and run:
git clone https://github.com/orithet/orithet.git
cd orithet
pip install -r requirements.txtpip install gradiopython orithet/cli.py --input_folder /path/to/media/folder --duration 120 --chaos 0.5 --seed 42 --resolution 1920 1080 --style dreampython orithet/gradio_ui.pyThen visit http://localhost:7860 in your browser.
NexusWeave accepts a diverse mix of media files:
- Videos: .mp4, .mov, .avi, .mkv
- Images: .jpg, .jpeg, .png, .gif
- Audio: .mp3, .wav, .flac
Recommended: 5-200 files in a single folder for best results.
| Parameter | Default | Description |
|---|---|---|
--input_folder |
(required) | Path to folder containing media files |
--duration |
120 | Target video duration in seconds |
--chaos |
0.5 | Randomness factor (0.0-1.0) |
--seed |
Random | Random seed for reproducibility |
--resolution |
1920 1080 | Output resolution (width height) |
--style |
dream | Preset style ("dream", "glitch", "psychedelic", "ambient") |
- Dream: Subtle effects, moderate chaos, 2-level recursion
- Glitch: Heavy glitch effects, high chaos, 3-level recursion
- Psychedelic: Intense visual transformations, high motion sensitivity
- Ambient: Minimal effects, low chaos, 1-level recursion
- Load Media: Scan input folder for supported file types
- Scene Detection: Use PySceneDetect to identify scene changes in videos
- Metadata Extraction: Calculate color palettes, motion energy, and dominant hues
- Audio Analysis: Extract BPM and energy bands from audio files
- Creature Creation: Each clip becomes a "creature" with unique properties
- Similarity Graph: Build relationships between clips based on visual/audio characteristics
- Ecosystem Dynamics: Creatures move, interact, and evolve over time:
- Same-color creatures fuse together
- High-motion creatures spread glitch effects
- Random mutations occur with configurable probability
- Behavioral Rules: Creatures follow physics-inspired movement and interaction rules
- Ordering: Sort creatures by age and position to create logical flow
- Timing: Adjust clip durations based on motion energy and chaos factor
- Beat Synchronization: Align transitions with audio beats
- Recursive Layers: Add PiP overlays at beat events
- Composition: Layer all elements with appropriate timing
- Effects Application: Apply beat-reactive visual effects
- Output: Generate final high-quality MP4 video (1920Γ1080, 30 fps)
python orithet/cli.py --input_folder ./media_samples --duration 180python orithet/cli.py --input_folder ./media_samples --duration 120 --chaos 0.8 --style psychedelicpython orithet/cli.py --input_folder ./media_samples --seed 12345 --style glitchpython orithet/cli.py --input_folder ./media_samples --resolution 3840 2160 --duration 240python orithet/cli.py --input_folder "C:\Users\YourName\Documents\media" --duration 120 --chaos 0.5python orithet/cli.py --input_folder ~/Documents/media --duration 120 --chaos 0.5orithet/
βββ __init__.py # Package initialization
βββ core.py # Core processing logic implementing all systems
βββ cli.py # Command-line interface
βββ gradio_ui.py # Web interface using Gradio
βββ requirements.txt # Python dependencies
βββ setup.py # Package installation configuration
βββ README.md # This file
To add new visual effects, modify the effects pipeline in core.py:
- Add effect function to the creature processing logic
- Include in the effect application stage
- Configure via style presets or command-line parameters
Each system is modular:
- PulseForge: Modify audio analysis in
process_audio()method - EchoMosaic: Adjust similarity calculation in
calculate_similarity()method - GlitchGarden: Extend creature interaction logic in
handle_creature_interaction()method - FractalFusion: Modify recursive overlay generation in
add_recursive_overlays()method
- Memory Usage: Processes media files in chunks to manage memory efficiently
- CPU Usage: Optimized for parallel processing where possible
- Recursion Depth: Limited to 3 levels to prevent render explosion
- Frame Rate: Default 30fps for smooth playback
- Ensure FFmpeg is added to your system PATH
- Use double quotes for paths with spaces
- Some antivirus software may block FFmpeg execution
- Use Homebrew to install dependencies for optimal performance
- May need to grant accessibility permissions for screen recording features
- Most distributions require explicit FFmpeg installation
- Consider using system package managers for optimized builds
- GPU acceleration support varies by distribution
Contributions are welcome! Please:
- Fork the repository
- Create a feature branch
- Commit your changes
- Push to the branch
- Open a Pull Request
MIT License - see LICENSE file for details.
Orithet is designed to work seamlessly across all major operating systems:
- Windows 10/11 - Full functionality with proper path handling
- macOS 10.15+ - Native support with Homebrew optimization
- Linux (Ubuntu 18.04+) - Package manager optimized installations
- Windows: Requires FFmpeg in PATH; uses Windows-style paths
- macOS: Recommended Homebrew installation for dependencies
- Linux: Standard package management for system dependencies
The tool maintains consistent behavior across platforms while adapting to each OS's specific requirements.
NexusWeave draws inspiration from:
- Audio-reactive visualizations
- Generative art principles
- Procedural content generation
- Digital ecosystems and evolution theory
Created with β€οΈ for artists, musicians, and creators exploring the intersection of technology and creativity.