Add double precision support for World Triplanar Mapping #91380
Conversation
fire
force-pushed
the
implement_double_precision_for_world_triplanar
branch
from
1494ab2 to
422f514
Compare
|
The center which is the triplanar material is correct in both far and near. Screen.Recording.2024-04-30.at.3.47.50.PM.mov |
|
tested working here, thank you |
fire
force-pushed
the
implement_double_precision_for_world_triplanar
branch
2 times, most recently
from
25e78bf to
ae60904
Compare
|
Resolved a conflict in |
fire
force-pushed
the
implement_double_precision_for_world_triplanar
branch
3 times, most recently
from
fe5a9d0 to
aa45100
Compare
|
Updated to 4.3-beta2. |
This commit adds emulated doubles support for the World UV Triplanar Mapping.
It also adds some QoL features to both ease the "integration" with the double
precision pipeline and also to allow greater flexibility when dealing
with Triplanar UVs.
Features:
- New mat4 TRIPLANAR_MATRIX vertex-only property (the same as MODEL_MATRIX by default).
- New "hint_triplanar_mat" hint for mat4 uniforms, which allow users
to specify a custom TRIPLANAR_MATRIX. Naturally, only a single
instance of the hint is allowed per shader.
- New vec3 TRIPLANAR_POSITION vertex-only property, which holds the
VERTEX transformed by the TRIPLANAR_MATRIX (With emulated doubles, if
applicable).
- Updated Standard 3D Material to make use of the above features.
- Updated VisualShader to include the TRIPLANAR_MATRIX and
TRIPLANAR_POSITION as valid Inputs.
- Added a "hint_triplanar_enabled" toggle to the VisualShaderNodeTransformParameter to allow
the new "hint_triplanar_mat" hint to be set.
Note: The shader related changes were applied to all renderers (Forward+,
Mobile and Compatibility).
Note²: Given the lack of emulated doubles support for the Compatibility
renderer, only the single precision path was implemented.
Co-Authored-By: =?UTF-8?q?Jo=C3=A3o=20Pedro=20Braz?= <brazjoaopedro@ymail.com>
fire
force-pushed
the
implement_double_precision_for_world_triplanar
branch
from
aa45100 to
a9d22cd
Compare
|
Updated to master. Had a code conflict in servers. |
There was a problem hiding this comment.
I'm not so sure about this PR. Its seems way too hacky to me. It is making huge changes in our shader compiler and internal shader to support something in the standard material.
Also, It looks like the new hint isn't even used in the StandardMaterial3D? So won't it remain broken for everyone?
It also relies on users manually passing in another matrix as a uniform and then secretly inserting that matrix into the backend code.
I think its worthwhile to see if we can improve this behaviour with a much simpler approach
ALso keep in mind our best practices https://docs.godotengine.org/en/latest/contributing/development/best_practices_for_engine_contributors.html#prefer-local-solutions particularly "Prefer local solutions".
|
Will need a rethink then, was trying to salvage some older prs |
|
I am not able to fix this at this moment, but the pr is still wanted and is salvageable. |
|
I don't think I can salvage this pull request but it is still wanted. |
Salvages #75577
The conflicts are mostly on these lines in the shader.
Use https://github.com/V-Sekai/godot-double-precision-test
Bugsquad edit: