Maybe sharper egui soon #216
Conversation
|
investigation progress: #215 (comment) |
| } | ||
|
|
||
| @fragment | ||
| fn fs_main_gamma_framebuffer(in: VertexOutput) -> @location(0) vec4<f32> { |
There was a problem hiding this comment.
do we really need a separate entry point? it's largely doing the same thing, we might as well just have an if/else at the end based on a runtime flag
There was a problem hiding this comment.
Runtime flag can do it. Just replicated egui-wgpu for the quick draft.
It would be nice to find the underlying cause so the shader potentially can be simple again..
| [mode; 3] | ||
| }, | ||
| mag_filter: match options.magnification { | ||
| egui::TextureFilter::Nearest => blade_graphics::FilterMode::Nearest, |
There was a problem hiding this comment.
we should have helper function to do this mapping, since we are doing it 3 times here
| egui::TextureFilter::Nearest => blade_graphics::FilterMode::Nearest, | ||
| egui::TextureFilter::Linear => blade_graphics::FilterMode::Linear, | ||
| }, | ||
| mipmap_filter: match options.mipmap_mode { |
There was a problem hiding this comment.
this would look like options.mipmap_mode.map_or(blade_graphics::FilterMode::Linear, my_mapping_function)
|
|
||
| use blade_graphics as gpu; | ||
| use winit::platform::x11::WindowAttributesExtX11; | ||
|
|
There was a problem hiding this comment.
do we need a separate example for this? we already have particle using egui, scene using it, and move as well.
There was a problem hiding this comment.
Nop, will remove, didn't intend to include it here, was just the one I used to debug
| @@ -144,7 +204,11 @@ impl GuiPainter { | |||
| ..Default::default() | |||
| }, | |||
| depth_stencil: None, //TODO? | |||
| fragment: shader.at("fs_main"), | |||
| fragment: if info.format.is_srgb() { | |||
There was a problem hiding this comment.
This looks highly suspicious to me.
Generally speaking, I expect the rendering pipeline to assume linear space render targets all the way. I don't think there is value in expanding this assumption any further.
There was a problem hiding this comment.
This is what egui-wgpu did so I just replicated all of it to see if it would help.
Egui examples use a non-srg render target whereas blade always uses one
There was a problem hiding this comment.
Minimizing changes is a good experiment for sure. Shipping it the way egui_wgpu did seems very concerning to me. I'll try to find some associated discussion there.
|
Just a note, my approach for the investigation was to minimize differences between blade and egui-wgpu, so I made sure to try and replicate their code. Once we find the issue we can go back and make only minimal changes in a own PR :) |
See notes in #215
Still not solved but colors are closer. Will keep looking into it