Commit graph

127 commits

Author SHA1 Message Date
Rémi Verschelde
4226d56ca9 Style: Enable clang-format on GLSL shaders
As of clang-format 6.0.1, putting the `/* clang-format off */` hint
around our "invalid" `[vertex]` and `[shader]` statements isn't enough
to prevent a bogus indent of the next comments and first valid statement,
so we need to enclose that first valid statement in the unformatted chunk.
2018-08-27 07:34:14 +02:00
Rémi Verschelde
1b6d75a599 Style: Fix code formatting in GLES3 shaders 2018-08-24 13:42:18 +02:00
Nick Hahn
66eba18969 Fix handling of normals that approach 1 2018-08-12 12:45:43 +02:00
Hugo Locurcio
31c12f05c4
Tweak the default SpatialMaterial properties
Roughness is now set to 1 by default and albedo is now white,
even on meshes that do not have any materials defined.

This means there is no longer a visual difference between a
mesh with no materials defined and a mesh with a default
SpatialMaterial defined.
2018-08-07 17:04:30 +02:00
Juan Linietsky
4b549faaab transmission was broken, fix was made. Also fixed treshold. 2018-07-22 14:04:11 -03:00
Juan Linietsky
ad88979841 Fix opaque pre pass not casting shadows 2018-07-22 13:18:27 -03:00
Alex Roman
2ce1118faa Add disable ambient light flag to shaders and materials 2018-07-17 21:30:43 +02:00
Rémi Verschelde
92bfde531c
Merge pull request #19786 from JFonS/correct_normal_scaling
Add render mode to ensure correct normals when using non-uniform scaling
2018-07-04 15:09:59 +02:00
Juan Linietsky
7fc2367508 Added ability for SSAO to affect AO textures too 2018-07-02 16:50:52 -03:00
JFonS
c8cf71753c Add render mode to ensure correct normals when using non-uniform scaling 2018-06-21 00:12:12 +02:00
JFonS
9c307d4b6c Added flag on SpatialMaterial to disable shadows 2018-03-29 18:46:42 +02:00
AndreaCatania
252344f706 Fixed graphic artifact caused by anisotropic filter 2018-01-11 12:55:18 +01:00
Bastiaan Olij
ebc96195d8 Clamp blend to fix screen space reflections 2018-01-10 23:23:00 +11:00
Juan Linietsky
4c23f94af9 properly blend interior and exterior ambient in reflection probes, fixes #14695 2018-01-06 19:34:03 -03:00
Juan Linietsky
d72a32bc54 Use better initialization value for normalmap, fixes #14720 2018-01-06 18:09:11 -03:00
Juan Linietsky
003f2dab78 Fix dual paraboloid shadow maps, closes #14487 2018-01-05 12:13:52 -03:00
Juan Linietsky
0c86c1ad15 Fixed GIProbe blending, closes #15164 2018-01-05 10:34:31 -03:00
Juan Linietsky
b411029e31
Merge pull request #14796 from AlmightyScientist/issue-14552
Shader Language: Fix Vertex Lighting artifacts.
2017-12-26 12:46:48 -03:00
Juan Linietsky
2b31155baa Fix sidedness check in material. Also remove SIDE built-in. 2017-12-26 11:09:52 -03:00
Juan Linietsky
7d0cabe9d3 Change skeleton processing to work on global coordinates, should help fix many import problems from Blender, GLTF2, etc. 2017-12-21 17:36:39 -03:00
Enzo Nocera
6c25eabbc5 Shader Language: Fix Vertex Lighting artifacts.
- When using Direction Lighting along with Vertex Lighting,
  putting a SpatialMaterial Roughness to 1.0 causes artifacts to appear.
  (#14552)

Fixes #14552.
2017-12-18 09:15:28 +01:00
Juan Linietsky
f3ad14224e -Add lightmapper
-Fixes to unwrapper (remove degenerates), makes Thekla not crash
-Added optional cancel button in EditorProgress
-Added function to force processing of events (needed for cancel button)
2017-12-14 09:01:27 -03:00
Rémi Verschelde
663f7e6a61 Fix usage of signed type in skeleton shader
The resulting primitive is drawn as GL_UNSIGNED_SHORT, so uvec4 should be used.
Fixes #12804.
2017-12-10 13:58:36 +01:00
Juan Linietsky
8c78ccb027 After a lot of trial and error, i think rim lighting implementation is now perfect, closes #13986 2017-12-08 11:09:03 -03:00
Juan Linietsky
8717afbfe1 Added small bit of bias depending on lenght for contact shadows, fixes #12726 2017-12-07 16:19:35 -03:00
Juan Linietsky
e8494f7ab7 Fixed rim lighting glitch, closes #13340, closes #13986 2017-12-06 14:30:49 -03:00
Rémi Verschelde
9b9fcb1977 Disable invariant gl_Position to workaround Mesa bug 100316
Fixes #13450, though it would be worth reverting if/when the Mesa bug is fixed.
2017-12-03 15:05:39 +01:00
Juan Linietsky
0243803117 Properly take into consideration that VERTEX must be written to in opaque pre pass, does some speed up to scenes using triplanar. 2017-12-01 08:45:36 -03:00
Rémi Verschelde
4a1d1cbbb4
Merge pull request #13290 from Chaosus/projectionfix
Fix invalid transform when skip_vertex_transform used
2017-11-26 21:31:02 +01:00
Chaosus
284f8f6d91 Fixed invalid transform when skip_vertex_transform used 2017-11-26 22:02:40 +03:00
volzhs
b97e87480e Fix shader compilation fail on Android
Error message from logcat

**ERROR**: SceneShaderGLES3: Fragment Program Compilation Failed:
50:471: S0032: no default precision defined for parameter 'probe'
   At: drivers/gles3/shader_gles3.cpp:167:_display_error_with_code() - SceneShaderGLES3: Fragment Program Compilation Failed:
2017-11-27 01:15:39 +09:00
Juan Linietsky
9738ebcda0 -Fixed height fog (was broken)
-Make sure materials are named in OBJ importer, so they can be saved outside as resources.
2017-11-25 12:32:41 -03:00
Ferenc Arn
6e68c86ba6 Add viewport to the other SceneData struct.
This was missed in #12288.
2017-10-30 20:48:33 -04:00
Juan Linietsky
4785e66eea
Merge pull request #12288 from tagcup/expose_viewport_size
Expose VIEWPORT_SIZE in shader language.
2017-10-30 19:32:21 -03:00
Ferenc Arn
a0fc641f25 Expose VIEWPORT_SIZE in shader language.
Fixes #11710.
2017-10-26 15:00:28 -04:00
Rémi Verschelde
abea816e54 Revert "Rename Schlick GGX to GGX." 2017-10-23 07:42:36 +02:00
Ferenc Arn
fd166d6fda Partially undo #11807.
Specular light at zero roughness needs some discussion. Until we do that, this fixes #12309.
2017-10-22 16:18:54 -04:00
Rémi Verschelde
d60b774924 Merge pull request #12297 from tagcup/clearcoat_fix
Don't add clearcoat BRDF to specular light as-is.
2017-10-22 13:18:24 +02:00
Ferenc Arn
06fcbe0092 Don't add clearcoat BRDF to specular light as-is.
BRDF needs to be multiplied with N.L, the incoming light and attenuation. Also specular_blob_intensity should affect clearcoat.
2017-10-21 20:25:17 -04:00
Ferenc Arn
cb0bf1edea Rename Schlick GGX to GGX.
Schlick's approximation and GGX are orthogonal concepts.

Furthermore, it's usage so far has been inconsistent: we don't even use it with anisotropic SchlickGGX, and Burley (Disney) diffuse does use it while its name doesn't indicate it.

The use of Schlick's approximation in Burley and GGX is an implementation detail and doesn't need to be reflected to the namig.
2017-10-21 19:35:54 -04:00
Ferenc Arn
6890245321 Corrections to #11807.
Added missing metallic parameter, also fixed a missed a case where wrong roughness condition was used.
2017-10-21 13:40:14 -04:00
Juan Linietsky
a2a4f9a62a Merge pull request #11807 from tagcup/roughness_metallic_limiting_cases
Fix the condition when specular light calculation is avoided (should …
2017-10-21 13:44:41 -03:00
Hiroshi Ogawa
59544c74fe Fix render mode specular schlick ggx 2017-10-14 19:32:13 +09:00
Ferenc Arn
9324555998 Fix the condition when specular light calculation is avoided (should be roughness == 1).
Also avoid calculating diffuse light calculation when metallic == 1, and test against clearcoat's own parameter to check whether to do clearcoat.
2017-10-11 17:22:44 -04:00
Ferenc Arn
12596cb5bc Fix anisotropic GGX D function, and introduce and use anistropic GGX G function.
Also fixed isotropic GGX G function with Schlick approximation, and added a commented out version without the approximation.
Added references.
2017-10-09 12:29:05 -04:00
Ferenc Arn
f48b162a6b Various clean ups and cosmetic changes in scene.glsl.
Use self-documenting names for variables which are otherwise confusing. Also avoid recalculating certain terms.
2017-10-03 11:02:39 -04:00
Andreas Haas
1e9465096d Merge pull request #11700 from tagcup/clearcoat_fix
Fix clearcoat without Schlick-GGX specular.
2017-10-02 22:51:51 +02:00
Juan Linietsky
66987d6878 missing PI division on vertex shader 2017-10-02 09:09:32 -03:00
Juan Linietsky
1b358783ce Slight proposed fixes to PBR. 2017-10-01 21:51:11 -03:00
Juan Linietsky
c864b782c0 Using advise from Matias, left the 0-1 range for lights, divisiom by PI, and multiplied lights by PI internally. 2017-10-01 20:33:38 -03:00