---

name: visualization-algorithms description: Evaluate and implement visualization algorithms for ocean/foam rendering. Use when discussing rendering techniques, contour algorithms, performance optimization for graphics, or evaluating algorithm tradeoffs. Auto-apply when editing files in src/render/.

Visualization Algorithms Skill

This project renders ocean waves and foam using Canvas 2D with custom algorithms. Apply this skill when evaluating or implementing visualization approaches.

Current Algorithm: Marching Squares

Location: src/render/marchingSquares.js

How It Works

1. Build intensity grid from foam row data
2. Apply box blur to smooth the data
3. Extract contours at threshold levels
4. Render as lines or filled shapes

CopyFoam Rows → Intensity Grid → Blur → Marching Squares → Contours → Canvas

Key Functions

Dispersion Options (A/B/C)

Three approaches for foam dispersion over time:

Option A: Expanding Segment Bounds

• Segment bounds expand outward over time
• Core intensity fades faster than edges
• Creates halo effect around original foam

Option B: Age-Based Blur

• More blur passes as foam ages
• Simpler but affects all foam uniformly
• Good for performance

Option C: Per-Row Dispersion Radius

• Each row tracks individual expansion
• Spreads in X and Y directions
• Most physically accurate

Algorithm Evaluation Criteria

When comparing visualization algorithms, consider:

1. Visual Quality

• Smoothness of contours
• Natural appearance of foam dispersion
• Correct representation of physics (energy dissipation)

2. Performance

• Target: 16ms frame budget (60fps)
• Grid resolution tradeoffs (80x60 default)
• Blur pass count vs. quality

3. Configurability

• Threshold levels for contour density
• Color/opacity mapping
• Animation parameters

4. Debugging

• Can render raw line segments for inspection
• Multi-threshold shows algorithm behavior
• Stories in Ladle for visual comparison

Performance Optimization Patterns

Grid Resolution

Copy// Lower resolution = faster, less detail
const gridW = 80;  // Default
const gridH = 60;

// For mobile/low-end: 40x30
// For high quality: 120x90

Blur Optimization

Copy// Box blur is O(n) per pixel with separable passes
// Multiple passes approximate Gaussian blur
const blurPasses = 2;  // Default - good balance

Typed Arrays

Copy// Use Float32Array for grid data
const grid = new Float32Array(gridW * gridH);

Stories for Visual Testing

Located in src/stories/FoamRendering.stories.jsx:

When Evaluating New Algorithms

1. Create a story for visual comparison
2. Benchmark performance with perf tests
3. Test edge cases: empty data, single row, many rows
4. Compare to physics expectations from 00-principles.md
5. Document tradeoffs in a plan document

Alternative Algorithms to Consider

Metaballs / Implicit Surfaces

• Smoother organic shapes
• Higher computational cost
• Better for sparse data

SDF (Signed Distance Fields)

• Resolution-independent
• GPU-friendly
• More complex implementation

Particle Systems

• Individual foam bubbles
• Very expensive at scale
• Most realistic appearance

Noise-Based

• Procedural detail
• Good for texture, not structure
• Can augment other methods

Reference Files

• plans/visuals/60-foam-effects.md - Foam design
• plans/visuals/132-foam-rendering-layers.md - Layered approach
• plans/visuals/foam-dispersion-v2-tests.md - Dispersion testing
• plans/00-principles.md - Physics basis for visual behavior