AlphaPixel pitches outline-based method for sharper GPU text
The company’s Slughorn implementation of the Slug technique computes glyph coverage from vector outlines, while AlphaPixel’s comparison sets out the trade-offs with texture atlases, distance fields and tessellation.
AlphaPixel has published a comparison of five approaches to GPU text rendering, making the case for Slug when text must remain sharp across changing scales and viewing angles. The company also introduces Slughorn, its C++20 implementation of the technique.
Texture atlases remain widely used because they are fast and portable, but their baked bitmaps can blur when enlarged and require storage for different sizes and glyph sets. Signed distance fields can scale more smoothly, while multi-channel distance fields are designed to retain sharp corners. Both still rely on textures baked at a chosen resolution.
Tessellation converts outlines into geometry and can suit animated vector artwork, though it requires geometry generation. Slug takes a different route: it stores glyph outlines and evaluates coverage in a fragment shader, without a baked texture atlas or per-frame tessellation, according to AlphaPixel.
The company says that approach can keep edges sharp under large scale changes and perspective transforms. Its own visual comparison uses selected examples and specified settings; it does not establish performance across all hardware or workloads.
AlphaPixel says Slughorn supports SVG, FreeType fonts and paths from Blend2D, Cairo and Skia, and targets OpenGL, Vulkan, WebGPU and DirectX. It describes the library as suitable for vector graphics as well as text.
AlphaPixel reports that Slug’s creator, Eric Lengyel, dedicated the technique’s patent to the public domain on 17 March 2026. The comparison presents the methods as serving different needs, with the choice depending on the rendering task.
