Textures and Aliasing
A wall image can cover a box without modeling every brick. Textures supply surface color or material information, often through images. UV coordinates specify which image location corresponds to each surface point.
What does one pixel represent?
Many distant stripes may fit inside one screen pixel. Reading only one point can make colors jump as the camera moves, or create false patterns. This is aliasing from sampling.
- Nearest filtering: Read the closest texel; enlargement reveals sharp blocks.
- Bilinear filtering: Blend four neighboring texels by relative distance.
- Mipmaps: Store progressively reduced images and use resolutions appropriate to minification.
A small averaging example follows. Color blending generally requires converting stored sRGB values into linear space first.
linear_samples = [0.0, 1.0, 0.0, 1.0]
print(sum(linear_samples) / len(linear_samples))
The result is 0.5, representing equal coverage by black and white regions. More samples or appropriate filtering can better represent coverage near boundaries.
Perspective projection also requires appropriate UV correction. Purely screen-linear interpolation can distort a texture on an angled surface.
Check your understanding
A dense pattern on a distant floor flickers. Will increasing the source texture’s resolution alone necessarily fix it?
Show explanation
No. The issue is how details smaller than a screen pixel are reduced. Examine mipmaps, filtering, and sampling. Adding source detail differs from displaying it consistently.