asgardius/virtualx-engine
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
50084 commits 1 branch 0 tags 2.2 GiB
Commit graph

207 commits

Author SHA1 Message Date
Rémi Verschelde
2e07477751
Merge pull request #70214 from Ansraer/speed-maybe 
[4.x] Use depth prepass to increase opaque render performance
 2022-12-23 23:43:50 +01:00
Ansraer
d44a91c2f7 use depth prepass to increase performance 2022-12-23 17:15:46 +01:00
Rémi Verschelde
1cab6c91e9
Merge pull request #69998 from BastiaanOlij/sorting-pivot-4 
Added options for sorting transparent objects (port of PR #63040)
 2022-12-23 09:47:24 +01:00
Bastiaan Olij
6f4f38db07 Added options for sorting transparent objects (port of PR 63040) 2022-12-23 19:30:33 +11:00
clayjohn
41021b07be Avoid updating particles during 2D rendering 
In 2D the particle uniform set isn't obtained until the render pass. So in 2D just avoid rendering if not updated yet.
 2022-12-22 12:04:34 -07:00
Rémi Verschelde
676f60b0cc
Merge pull request #70132 from clayjohn/RT-update 
Implement render_target_was_used API so that Viewports can properly check if they have been used.
 2022-12-17 12:26:51 +01:00
clayjohn
1b330820bf Implement render_target_was_used API so that Viewports can properly check if they have been used. 
For the RD renderer, this does not work for Viewports used in scene shaders yet
 2022-12-16 09:50:45 -08:00
Lyuma
80dea918c4 Flip culling when rendering a camera with negative scale 2022-12-15 16:12:10 -08:00
Yuri Rubinsky
56954485ed Refactor ShaderData & fix the sorting of shader uniforms 2022-12-15 10:02:59 +03:00
NumbuhFour
1b09fd5410 Implement CAMERA_VISIBLE_LAYERS as built-in shader variable 2022-12-03 22:41:27 -08:00
Yuri Rubinsky
acaf38cfbc Expose BarrierMask as flags enum in RenderingDevice 2022-11-22 09:45:20 +03:00
clayjohn
21ac6d7d8e Finish implementing Canvas Background mode 2022-11-17 12:51:26 -08:00
Rémi Verschelde
64d7ce2a6e
Merge pull request #68628 from clayjohn/shadows 
Implement is_animated and casts_shadows
 2022-11-15 10:27:29 +01:00
Markus Sauermann
3b14f0334c Remove redundant Variant-types initializations 2022-11-14 19:35:19 +01:00
Rémi Verschelde
63cded6a85
Merge pull request #61884 from Calinou/alpha-hash-use-opaque-pipeline 
Use opaque rendering pipeline for alpha hash materials
 2022-11-14 11:08:35 +01:00
clayjohn
4abf47f407 Implement is_animated and casts_shadows 
This allows the renderer to correctly decide when to update shadow maps

This PR also adds TIME to a few missing places
 2022-11-13 14:09:21 -08:00
clayjohn
aaaa4b8d68 Properly set TIME shader uniform when rendering shadows 2022-11-12 10:51:02 -08:00
clayjohn
a7adc7ab37 Ensure user shader is used in depth pass when point size is used 2022-11-10 15:26:07 -08:00
Rémi Verschelde
3312d0a87b
Merge pull request #67907 from clayjohn/RD-model-matrix 
Correctly rename to read_model_matrix when using shader builtins that rely on model matrix
 2022-10-31 11:07:24 +01:00
Gordon MacPherson
56df8d5f19 Fix EXE_BAD_ACCESS caused by optional argument 
This argument is now non optional, but this never hits the same bad access.
I voted to simplify the code here since the argument is never used optionally in our codebase.
 2022-10-26 19:33:35 +01:00
clayjohn
020ac29fc3 Correctly rename to read_model_matrix when using shader builtins that rely on model matrix 2022-10-26 09:45:35 -07:00
TechnoPorg
b04350828e Use radial distance for making LOD decisions. 
Previously, only forward basis distance from the camera was used.
This means that unnecessarily high LOD levels were used for objects located to the side of the camera.
The distance from the camera origin is now used, independently of direction.
 2022-10-12 07:21:42 -06:00
Robbie Cooper
79d3e090cd Vulkan Clustered: Fix culling of negatively-scaled objects 
Negatively scaled objects should be mirrored. This is already implemented, but it breaks when mirrored and non-mirrored instances of the same object are visible together.

It turns out that the code that skips-over repeats in `RenderForwardClustered::_render_list_template` also skips the code that accounts for the culling mode of mirrored objects.

The solution here is to consider the `mirror` flag when determining repeats. This might result in more draw commands than necessary since a mirrored object can split a group of non-mirrored instances in two.

This problem doesn't appear in the mobile renderer because the repeat optimization isn't implemented there yet.

The problem still appears in MultiMeshInstance3D in *all* renderers.

Fixes #62879 and #58546.
 2022-10-10 03:20:59 -04:00
Rémi Verschelde
17c62a692e
Merge pull request #66756 from BastiaanOlij/fix_ssr 
Fixing artifacts in SSR
 2022-10-06 22:27:49 +02:00
Rémi Verschelde
d935235e9c Merge pull request #66922 from BastiaanOlij/fix_direct_shadow 
Make sure atlas rect for directional lights is calculated using floats
 2022-10-05 13:38:24 +02:00
Bastiaan Olij
aba356e882 Make sure atlas rect for directional lights is calculated using floats 2022-10-05 22:12:57 +11:00
Rémi Verschelde
ecd4ab1e14 Merge pull request #66683 from clayjohn/SRGB-canvasitem 
Default CanvasItem materials to use sRGB space for uniform colors
 2022-10-05 11:44:52 +02:00
Bastiaan Olij
c7656978ba Adding getters to RenderTarget and implementing override functionality for XR 2022-10-05 11:37:49 +11:00
Bastiaan Olij
d09ad714bf Fixing artifacts in SSR 2022-10-04 23:47:13 +11:00
Bastiaan Olij
02b0c22e38 Moving SSEffects settings into class 2022-10-04 11:03:36 +11:00
Bastiaan Olij
ddc4ae1175 Move cluster builder, sdfgi and gi structures to clustered renderer, move light and probe elements into storage and reorganise our render_scene method. 2022-10-04 11:03:32 +11:00
clayjohn
3126619aa5 Update Instance flags in shaders to match instance flags in engine 2022-10-02 21:37:06 -07:00
clayjohn
06d0e56559 Default CanvasItem materials to not convert uniform colors to linear space. 
The 2D renderer in Godot is totally in sRGB space so it is appropriate
to keep 2D uniform colors in sRGB space
 2022-09-30 15:15:11 -07:00
Rémi Verschelde
67961d875d Merge pull request #66178 from clayjohn/double-precision-rendering 
Emulate double precision for regular rendering operation when REAL_T_IS_DOUBLE
 2022-09-30 09:59:45 +02:00
bruvzg
ea1848ce0a
Use constexpr in the conditions with template parameters and sizeofs to suppress C4127 warnings. 2022-09-29 10:38:21 +03:00
clayjohn
aac8d5c406 Properly initialize motion vectors offset when motion vectors are disabled. 
This fixes an uninitialized memory bug that caused particles to fail in non-LTO builds
 2022-09-22 20:36:01 -07:00
clayjohn
27a3014f50 Emulate double precision for regular rendering operation. 
We calculate the lost precision on the CPU and pass it into the GPU
so that it can calculate an error-corrected version of the vertex position
 2022-09-20 23:40:01 -07:00
Rémi Verschelde
7da532275b Merge pull request #65541 from clayjohn/renderer-setting 
Split rendering driver project setting into renderer_name and rendering_driver
 2022-09-20 09:43:59 +02:00
clayjohn
4a1c7de57c Split rendering driver project setting into renderer_name and rendering_driver. To differentiate between a driver (e.g. Vulkan or D3D12) and a renderer (e.g. clustered or mobile renderer). 2022-09-19 10:26:10 -07:00
Rémi Verschelde
bda63e1b5a Merge pull request #65833 from JFonS/taa_fix_particles_errors 
Fix error spam in the renderer when using GPUParticles3D
 2022-09-16 11:59:00 +02:00
JFonS
b668268772 Fix error spam in the renderer when using GPUParticles3D 
The flag INSTANCE_DATA_FLAG_MULTIMESH is used for both multimesh and particles instances, this commit adds a new INSTANCE_DATA_FLAG_PARTICLES flag to discriminate between them.

This flag will also be used in the future to properly support TAA in particles.
 2022-09-16 09:44:22 +02:00
Bastiaan Olij
02ea1de7d0 Extract shared scene data into a separate class 2022-09-15 12:09:57 +10:00
Clay John
f709596631
Merge pull request #65418 from JFonS/taa_multimesh 
Add motion vectors support for MultiMeshInstance
 2022-09-14 09:06:27 -07:00
jfons
74cdee4cb5 Add motion vectors support for MultiMeshInstance 
Transparently handle changes in multimesh transforms so they get reflected in the motion vectors buffer.
 2022-09-12 15:28:12 +02:00
Rémi Verschelde
f211869530
Merge pull request #65594 from clayjohn/SSR-dialectric 2022-09-12 14:23:04 +02:00
Bastiaan Olij
b519d7619c Fixed MSAA initialisation in clustered forward renderer 2022-09-12 11:07:51 +10:00
clayjohn
65f0113bc6 Properly scale SSR reflection based on metallic value for dielectric materials 2022-09-09 16:52:02 -07:00
Rémi Verschelde
c82bbc38a5 Merge pull request #64952 from Chaosus/vs_rename_uniform_to_param 2022-09-02 13:49:53 +02:00
Bastiaan Olij
2cd84be64d Extracting render buffers and changing it to a more generic solution 2022-09-01 20:01:45 +10:00
Yuri Rubinsky
8191b3c110 Rename uniform to parameter across the engine 2022-09-01 11:42:57 +03:00