20 lines
1.3 KiB
Markdown
20 lines
1.3 KiB
Markdown
# Chapter 6: Textures
|
|
|
|
As I read this book, I summarize what I think is wrong. If you think my comments are wrong then please let me know. We can dicuss more and update your opinion.
|
|
|
|
## sRGB color space
|
|
|
|
* Actually we can't do math on non-linear space because `1 + 1 != 2` in non-linear space. Let's explain a reason in detail.
|
|
|
|
|
|
|
|
Data in the reality are in linear space and those are encoded non-linear space because the limitation of memory.
|
|
It made sRGB color format that compromised between how cathode ray tube monitors worked and what colors the human eye sees.
|
|
Let's think about that you are calculating `0.5 + 0.5` and the result must be `1.0`. However `0.5` is in non-linear space
|
|
so it is `0.729` in linear space. So `0.5 + 0.5` in non-linear space equals to `0.729 + 0.729` so the result is `1.458` not `1.0`.
|
|
Thus when you do the math, you must check values are in linear space.
|
|
|
|
* If texture's format is sRGB then OpenGL, Vulkan and DirectX convert from non-linear space to linear space automatically
|
|
when you fetch a texel with a sampler. However Metal don't do that. This book give a code how to convert from non-linear space to linear space.
|
|
But the code is not compiled when you copy. Try this code `baseColor = pow(baseColor, 1.0 / 2.2);`. You can do the gamma correction by yourself.
|