Role
Visual Generalist
Year
2024
This project was conceived as a deep dive into high-viscosity fluid simulation and shader look development—a procedural interpretation of molten lava flowing over terrain. At its core, the work balances physical simulation discipline with artistic expression, using custom viscosity controls and temperature-based behaviors to evoke the natural movement of heavy, cooling fluid.
The FX design emphasized controlled complexity and art-directable dynamics. Rather than relying on default fluid presets, the simulation incorporated dual temperature sources with noise modulation to introduce realistic variation in velocity and surface texture, allowing hotter, faster regions to carry cooler, denser material in a visually compelling interplay. Geometry proxies and optimized collision VDBs were used to ensure performance without sacrificing the fine surface detail necessary for believable interaction with the environment.
From a lighting and rendering perspective, the piece leveraged Mantra’s physically based shading to capture emissive glow and surface displacement tied to solver attributes. Procedural color ramps and custom maps derived from simulation data were used to drive emission intensity and roughness, reinforcing the visual impression of cooling and solidifying material. This integration of simulation attributes into shader logic elevated the look development phase beyond mere rendering, allowing thermo-dynamic behavior to manifest visually.
Creative Goals
- Explore fluid dynamics beyond traditional low-viscosity sims, focusing on temperature-dependent behavior
- Integrate simulation attributes into surface shading for rich look development
- Maintain a balance of performance and detail through optimized proxy geometry and modular workflows
Outcome
The final piece functions as both a technical examination of viscosity-based simulation and a visual study of procedural material behavior. It demonstrates how custom solver configurations, data-driven shaders, and pipeline-aware strategies can converge to produce imagery that feels both physically grounded and artistically expressive—turning algorithmic motion into evocative visual storytelling.