c - highp float在GLSL中是如何呈现的？

Question

引用opengles_shading_language第 4.5 节:

It is strongly advised that the vertex language provide a floating point range and precision matching that of an IEEE single precision floating point number, or better. It is required that the vertex language provide floating point variables whose range is at least (-262, 262) and whose precision is at least one part in 65536. This is stated in more detail in the following tables

如果假设它是一个 IEEE single precision floating point number ，相对精度如何只有 2^-16 和 23 位小数？

Answer 1

你引用的文字说:

• 要求范围至少为 (-262, 262)，精度至少为 2^-16
• 强烈建议
使用完整的 IEEE 单精度 float

所以如果顶点语言提供 IEEE 单精度 float ，它确实会比 2^-16 具有更好的分辨率。但是，这不是必需的。