|
Blender V5.3
|
#include <gpu_shader_colorspace.bsl.hh>
Public Member Functions | |
| float4 | rec709_srgb_to_output_space (float4 srgb_color) const |
Static Public Member Functions | |
| static float4 | scene_linear_to_rec709_srgb (float3x3 scene_linear_to_rec709, float4 scene_linear_color) |
Public Attributes | |
| const bool | srgbTarget |
Definition at line 11 of file gpu_shader_colorspace.bsl.hh.
Input is Rec.709 sRGB. Output is Rec.709 linear if hardware will add a Linear to sRGB conversion, NOOP otherwise.
As per GPU API design, all frame-buffers with SRGBA_8_8_8_8 attachments will always enable SRGB rendering. In this mode the shader output is expected to be in a linear color space. This allows to do the blending stage with linear values (more correct) and then store the result in 8bpc keeping accurate colors.
To ensure consistent result (blending excluded) between a frame-buffer using SRGBA_8_8_8_8 and one using RGBA_8_8_8_8, we need to do the sRGB > linear conversion to counteract the hardware encoding during frame-buffer output.
For reference: https://wikis.khronos.org/opengl/framebuffer#Colorspace
IMPORTANT: srgbTarget denote that the output is expected to be in linear space. https://wikis.khronos.org/opengl/framebuffer#Colorspace
Definition at line 29 of file gpu_shader_colorspace.bsl.hh.
References max, mix, pow, srgbTarget, and step.
|
inlinestatic |
Definition at line 50 of file gpu_shader_colorspace.bsl.hh.
References max, mix, pow, and step.
Referenced by builtin::image::image_frag(), and builtin::image::image_overlay_merge_frag().
| const bool builtin::ColorSpace::srgbTarget |
Definition at line 12 of file gpu_shader_colorspace.bsl.hh.
Referenced by rec709_srgb_to_output_space().