Welcome to the Advanced module. These recipes involve deep dives into fluid dynamics, motion tracking, and complex math. We focus on high-performance simulations and sophisticated human-computer interaction systems.
Physics & Simulation
- (y-) GPU Fluid Simulation: A complex TOP-based feedback system for smoke and liquid dynamics based on Jos Stam’s stable fluids algorithm. Optimized for Apple Silicon.
- (y-) Dreamscape Particle Cloud: A stylized, glowing particle cloud that uses recursive feedback and blur chains to create the “bloom” effect characteristic of high-end generative art.
- (y-) Vector Field Instancing: Learn how to create millions of points following swirling vortex paths driven by an animated 3D vector noise grid on the GPU.
- (y-) Reaction-Diffusion (Gray-Scott Model): Simulate the chemical patterns found on animal skins and seashells entirely in a GLSL feedback loop. Spot, stripe, and worm-hole morphologies emerge from two parameters.
Motion Tracking & Interaction
- (y-) Fluid Brush with Hand Tracking: Your hand becomes a brush painting glowing fluid trails in real time. A MediaPipe wrist position drives a feedback loop combining diffusion, advection, and decay - the interactive wall installation technique.
- (y-) ★ Complete Hand Tracking Walkthrough: A comprehensive project covering everything from initial setup to building a watercolor brush and a generative architecture scene driven by MediaPipe landmarks.
- (y-) MediaPipe Hand Tracking Advanced: Learn multi-hand tracking, complex gesture recognition (swipes, rotations, claps), and advanced interaction systems for large-scale installations.
- (y-) Motion History & Optical Flow: Motion analysis using simple webcams. Learn how to track speed and direction without expensive infrared sensors.
- (y-) Hand Tracking Index: The central hub for all MediaPipe resources, including video links and series overview.
Math & Fractals
- (y-) 3D Sierpinski Tetrahedron: Fractal geometry controlled by hand orientation and pinch-zoom gestures. We use the Copy SOP and GPU Instancing for recursive complexity.
- (y-) ★ Hand-Tracked Chaotic Attractor: A math-heavy project using Script CHOPs to drive a Lorenz attractor system in real time, optimized for high framerates on modern hardware.
- (y-) Real-time Neural Style Transfer: A deep-learning recipe using OpenCV and pre-trained ONNX models inside the Script TOP.
(y) Return to Recipes & Projects | (y) Return to TouchDesigner | (y) Return to Home