Engine Comparison

Covecto ships two vectorization engines, each optimized for different image types.

At a Glance

	PixelExact	Spline
Algorithm	Flood-fill → boundary trace → rectilinear SVG	Color quantization → contour trace → Bézier splines
Best for	Icons, pixel art, screenshots, UI elements	Photos, illustrations, artwork, gradients
Output style	Sharp right angles, exact pixel boundaries	Smooth curves, organic shapes
Color handling	Exact color regions	Quantized color palette
Complexity	O(n) flood fill	Iterative color quantization + spline fitting

When to Use Which

Use PixelExact when:

• The input is a small icon (≤64px)
• You need crisp, pixel-perfect boundaries
• The image has flat colors with sharp edges
• Vectorizing screenshots or UI mockups

Use Spline when:

• The input is a photograph or illustration
• You need smooth, organic curves
• The image has gradients or many colors
• File size matters more than pixel-perfect accuracy

Use Auto when:

• You're not sure which engine to pick
• Processing mixed image types in batch
• You want the best default behavior

Auto-Engine Heuristics

The auto engine analyzes the image and selects based on:

1. Image size — small images (≤64px) lean toward PixelExact
2. Color count — low color count (≤16) suggests PixelExact
3. Edge density — high edge density favors Spline
4. Aspect ratio — very tall/wide images favor Spline

The heuristic is conservative — when in doubt, it picks Spline.