seedlingattempt/godot
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge pull request #110028 from zeux/simplify-prune

Browse Source 
Enable component pruning during mesh simplification
This commit is contained in:
Thaddeus Crews
2025-10-01 13:12:36 -05:00
parent eab59f8057 ccbece58d8
commit f745685f54
1 changed files with 15 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
scene/resources/3d/importer_mesh.cpp
View File

@ -480,6 +480,7 @@ void ImporterMesh::generate_lods(float p_normal_merge_angle, Array p_bone_transf
LocalVector<int> current_indices = merged_indices;
float current_error = 0.0f;
bool allow_prune = true;
while (current_indices.size() > min_target_indices * 2) {
unsigned int current_index_count = current_indices.size();
@ -493,6 +494,11 @@ void ImporterMesh::generate_lods(float p_normal_merge_angle, Array p_bone_transf
// Lock geometric boundary in case the mesh is composed of multiple material subsets.
simplify_options |= SurfaceTool::SIMPLIFY_LOCK_BORDER;
if (allow_prune) {
// Remove small disconnected components.
simplify_options |= SurfaceTool::SIMPLIFY_PRUNE;
}
if (deformable) {
// Improves appearance of deformable objects after deformation by using more regular tessellation.
simplify_options |= SurfaceTool::SIMPLIFY_REGULARIZE;
@ -513,6 +519,15 @@ void ImporterMesh::generate_lods(float p_normal_merge_angle, Array p_bone_transf
simplify_options,
&step_error);
if (new_index_count == 0 && allow_prune) {
// If the best result the simplifier could arrive at with pruning enabled is 0 triangles, there might still be an opportunity
// to reduce the number of triangles further *without* completely decimating the mesh. It will be impossible to reach the target
// this way - if the target was reachable without going down to 0, the simplifier would have done it! - but we might still be able
// to get one more slightly lower level if we retry without pruning.
allow_prune = false;
continue;
}
// Accumulate error over iterations. Usually, it's correct to use step_error as is; however, on coarse LODs, we may start
// getting *smaller* relative error compared to the previous LOD. To make sure the error is monotonic and strictly increasing,
// and to limit the switching (pop) distance, we ensure the error grows by an arbitrary factor each iteration.