Function compressNormals

Compresses a flat array of 3D normal vectors using oct-encoding.

This function compresses an array of 3D normal vectors into a smaller array of oct-encoded normals. The compression minimizes storage requirements while preserving the direction of the normals. The last element in each triplet is redundant due to a workaround for WebGL issues where 2-element normals would not behave correctly in the shader.

The function performs various oct-encoding operations using combinations of ceil and floor to minimize rounding errors during compression.

const normals = new Float32Array([1, 0, 0, 0, 1, 0, 0, 0, 1]);
const compressed = compressNormals(normals);
console.log(compressed); // Compressed oct-encoded normals
Copy