forked from godotengine/godot
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathlightmapper_rd.cpp
1735 lines (1491 loc) · 60.4 KB
/
lightmapper_rd.cpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*************************************************************************/
/* lightmapper_rd.cpp */
/*************************************************************************/
/* This file is part of: */
/* GODOT ENGINE */
/* https://godotengine.org */
/*************************************************************************/
/* Copyright (c) 2007-2022 Juan Linietsky, Ariel Manzur. */
/* Copyright (c) 2014-2022 Godot Engine contributors (cf. AUTHORS.md). */
/* */
/* Permission is hereby granted, free of charge, to any person obtaining */
/* a copy of this software and associated documentation files (the */
/* "Software"), to deal in the Software without restriction, including */
/* without limitation the rights to use, copy, modify, merge, publish, */
/* distribute, sublicense, and/or sell copies of the Software, and to */
/* permit persons to whom the Software is furnished to do so, subject to */
/* the following conditions: */
/* */
/* The above copyright notice and this permission notice shall be */
/* included in all copies or substantial portions of the Software. */
/* */
/* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, */
/* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF */
/* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.*/
/* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY */
/* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, */
/* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE */
/* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. */
/*************************************************************************/
#include "lightmapper_rd.h"
#include "core/config/project_settings.h"
#include "core/math/geometry_2d.h"
#include "lm_blendseams.glsl.gen.h"
#include "lm_compute.glsl.gen.h"
#include "lm_raster.glsl.gen.h"
#include "servers/rendering/rendering_device_binds.h"
//uncomment this if you want to see textures from all the process saved
//#define DEBUG_TEXTURES
void LightmapperRD::add_mesh(const MeshData &p_mesh) {
ERR_FAIL_COND(p_mesh.albedo_on_uv2.is_null() || p_mesh.albedo_on_uv2->is_empty());
ERR_FAIL_COND(p_mesh.emission_on_uv2.is_null() || p_mesh.emission_on_uv2->is_empty());
ERR_FAIL_COND(p_mesh.albedo_on_uv2->get_width() != p_mesh.emission_on_uv2->get_width());
ERR_FAIL_COND(p_mesh.albedo_on_uv2->get_height() != p_mesh.emission_on_uv2->get_height());
ERR_FAIL_COND(p_mesh.points.size() == 0);
MeshInstance mi;
mi.data = p_mesh;
mesh_instances.push_back(mi);
}
void LightmapperRD::add_directional_light(bool p_static, const Vector3 &p_direction, const Color &p_color, float p_energy, float p_angular_distance, float p_shadow_blur) {
Light l;
l.type = LIGHT_TYPE_DIRECTIONAL;
l.direction[0] = p_direction.x;
l.direction[1] = p_direction.y;
l.direction[2] = p_direction.z;
l.color[0] = p_color.r;
l.color[1] = p_color.g;
l.color[2] = p_color.b;
l.energy = p_energy;
l.static_bake = p_static;
l.size = Math::tan(Math::deg2rad(p_angular_distance));
l.shadow_blur = p_shadow_blur;
lights.push_back(l);
}
void LightmapperRD::add_omni_light(bool p_static, const Vector3 &p_position, const Color &p_color, float p_energy, float p_range, float p_attenuation, float p_size, float p_shadow_blur) {
Light l;
l.type = LIGHT_TYPE_OMNI;
l.position[0] = p_position.x;
l.position[1] = p_position.y;
l.position[2] = p_position.z;
l.range = p_range;
l.attenuation = p_attenuation;
l.color[0] = p_color.r;
l.color[1] = p_color.g;
l.color[2] = p_color.b;
l.energy = p_energy;
l.static_bake = p_static;
l.size = p_size;
l.shadow_blur = p_shadow_blur;
lights.push_back(l);
}
void LightmapperRD::add_spot_light(bool p_static, const Vector3 &p_position, const Vector3 p_direction, const Color &p_color, float p_energy, float p_range, float p_attenuation, float p_spot_angle, float p_spot_attenuation, float p_size, float p_shadow_blur) {
Light l;
l.type = LIGHT_TYPE_SPOT;
l.position[0] = p_position.x;
l.position[1] = p_position.y;
l.position[2] = p_position.z;
l.direction[0] = p_direction.x;
l.direction[1] = p_direction.y;
l.direction[2] = p_direction.z;
l.range = p_range;
l.attenuation = p_attenuation;
l.cos_spot_angle = Math::cos(Math::deg2rad(p_spot_angle));
l.inv_spot_attenuation = 1.0f / p_spot_attenuation;
l.color[0] = p_color.r;
l.color[1] = p_color.g;
l.color[2] = p_color.b;
l.energy = p_energy;
l.static_bake = p_static;
l.size = p_size;
l.shadow_blur = p_shadow_blur;
lights.push_back(l);
}
void LightmapperRD::add_probe(const Vector3 &p_position) {
Probe probe;
probe.position[0] = p_position.x;
probe.position[1] = p_position.y;
probe.position[2] = p_position.z;
probe.position[3] = 0;
probe_positions.push_back(probe);
}
void LightmapperRD::_plot_triangle_into_triangle_index_list(int p_size, const Vector3i &p_ofs, const AABB &p_bounds, const Vector3 p_points[3], uint32_t p_triangle_index, LocalVector<TriangleSort> &triangles, uint32_t p_grid_size) {
int half_size = p_size / 2;
for (int i = 0; i < 8; i++) {
AABB aabb = p_bounds;
aabb.size *= 0.5;
Vector3i n = p_ofs;
if (i & 1) {
aabb.position.x += aabb.size.x;
n.x += half_size;
}
if (i & 2) {
aabb.position.y += aabb.size.y;
n.y += half_size;
}
if (i & 4) {
aabb.position.z += aabb.size.z;
n.z += half_size;
}
{
Vector3 qsize = aabb.size * 0.5; //quarter size, for fast aabb test
if (!Geometry3D::triangle_box_overlap(aabb.position + qsize, qsize, p_points)) {
//does not fit in child, go on
continue;
}
}
if (half_size == 1) {
//got to the end
TriangleSort ts;
ts.cell_index = n.x + (n.y * p_grid_size) + (n.z * p_grid_size * p_grid_size);
ts.triangle_index = p_triangle_index;
triangles.push_back(ts);
} else {
_plot_triangle_into_triangle_index_list(half_size, n, aabb, p_points, p_triangle_index, triangles, p_grid_size);
}
}
}
Lightmapper::BakeError LightmapperRD::_blit_meshes_into_atlas(int p_max_texture_size, Vector<Ref<Image>> &albedo_images, Vector<Ref<Image>> &emission_images, AABB &bounds, Size2i &atlas_size, int &atlas_slices, BakeStepFunc p_step_function, void *p_bake_userdata) {
Vector<Size2i> sizes;
for (int m_i = 0; m_i < mesh_instances.size(); m_i++) {
MeshInstance &mi = mesh_instances.write[m_i];
Size2i s = Size2i(mi.data.albedo_on_uv2->get_width(), mi.data.albedo_on_uv2->get_height());
sizes.push_back(s);
atlas_size.width = MAX(atlas_size.width, s.width + 2);
atlas_size.height = MAX(atlas_size.height, s.height + 2);
}
int max = nearest_power_of_2_templated(atlas_size.width);
max = MAX(max, nearest_power_of_2_templated(atlas_size.height));
if (max > p_max_texture_size) {
return BAKE_ERROR_LIGHTMAP_TOO_SMALL;
}
if (p_step_function) {
p_step_function(0.1, RTR("Determining optimal atlas size"), p_bake_userdata, true);
}
atlas_size = Size2i(max, max);
Size2i best_atlas_size;
int best_atlas_slices = 0;
int best_atlas_memory = 0x7FFFFFFF;
Vector<Vector3i> best_atlas_offsets;
//determine best texture array atlas size by bruteforce fitting
while (atlas_size.x <= p_max_texture_size && atlas_size.y <= p_max_texture_size) {
Vector<Vector2i> source_sizes;
Vector<int> source_indices;
source_sizes.resize(sizes.size());
source_indices.resize(sizes.size());
for (int i = 0; i < source_indices.size(); i++) {
source_sizes.write[i] = sizes[i] + Vector2i(2, 2); // Add padding between lightmaps
source_indices.write[i] = i;
}
Vector<Vector3i> atlas_offsets;
atlas_offsets.resize(source_sizes.size());
int slices = 0;
while (source_sizes.size() > 0) {
Vector<Vector3i> offsets = Geometry2D::partial_pack_rects(source_sizes, atlas_size);
Vector<int> new_indices;
Vector<Vector2i> new_sources;
for (int i = 0; i < offsets.size(); i++) {
Vector3i ofs = offsets[i];
int sidx = source_indices[i];
if (ofs.z > 0) {
//valid
ofs.z = slices;
atlas_offsets.write[sidx] = ofs + Vector3i(1, 1, 0); // Center lightmap in the reserved oversized region
} else {
new_indices.push_back(sidx);
new_sources.push_back(source_sizes[i]);
}
}
source_sizes = new_sources;
source_indices = new_indices;
slices++;
}
int mem_used = atlas_size.x * atlas_size.y * slices;
if (mem_used < best_atlas_memory) {
best_atlas_size = atlas_size;
best_atlas_offsets = atlas_offsets;
best_atlas_slices = slices;
best_atlas_memory = mem_used;
}
if (atlas_size.width == atlas_size.height) {
atlas_size.width *= 2;
} else {
atlas_size.height *= 2;
}
}
atlas_size = best_atlas_size;
atlas_slices = best_atlas_slices;
// apply the offsets and slice to all images, and also blit albedo and emission
albedo_images.resize(atlas_slices);
emission_images.resize(atlas_slices);
if (p_step_function) {
p_step_function(0.2, RTR("Blitting albedo and emission"), p_bake_userdata, true);
}
for (int i = 0; i < atlas_slices; i++) {
Ref<Image> albedo;
albedo.instantiate();
albedo->create(atlas_size.width, atlas_size.height, false, Image::FORMAT_RGBA8);
albedo->set_as_black();
albedo_images.write[i] = albedo;
Ref<Image> emission;
emission.instantiate();
emission->create(atlas_size.width, atlas_size.height, false, Image::FORMAT_RGBAH);
emission->set_as_black();
emission_images.write[i] = emission;
}
//assign uv positions
for (int m_i = 0; m_i < mesh_instances.size(); m_i++) {
MeshInstance &mi = mesh_instances.write[m_i];
mi.offset.x = best_atlas_offsets[m_i].x;
mi.offset.y = best_atlas_offsets[m_i].y;
mi.slice = best_atlas_offsets[m_i].z;
albedo_images.write[mi.slice]->blit_rect(mi.data.albedo_on_uv2, Rect2i(Vector2i(), mi.data.albedo_on_uv2->get_size()), mi.offset);
emission_images.write[mi.slice]->blit_rect(mi.data.emission_on_uv2, Rect2(Vector2i(), mi.data.emission_on_uv2->get_size()), mi.offset);
}
return BAKE_OK;
}
void LightmapperRD::_create_acceleration_structures(RenderingDevice *rd, Size2i atlas_size, int atlas_slices, AABB &bounds, int grid_size, Vector<Probe> &probe_positions, GenerateProbes p_generate_probes, Vector<int> &slice_triangle_count, Vector<int> &slice_seam_count, RID &vertex_buffer, RID &triangle_buffer, RID &lights_buffer, RID &triangle_cell_indices_buffer, RID &probe_positions_buffer, RID &grid_texture, RID &seams_buffer, BakeStepFunc p_step_function, void *p_bake_userdata) {
HashMap<Vertex, uint32_t, VertexHash> vertex_map;
//fill triangles array and vertex array
LocalVector<Triangle> triangles;
LocalVector<Vertex> vertex_array;
LocalVector<Seam> seams;
slice_triangle_count.resize(atlas_slices);
slice_seam_count.resize(atlas_slices);
for (int i = 0; i < atlas_slices; i++) {
slice_triangle_count.write[i] = 0;
slice_seam_count.write[i] = 0;
}
bounds = AABB();
for (int m_i = 0; m_i < mesh_instances.size(); m_i++) {
if (p_step_function) {
float p = float(m_i + 1) / mesh_instances.size() * 0.1;
p_step_function(0.3 + p, vformat(RTR("Plotting mesh into acceleration structure %d/%d"), m_i + 1, mesh_instances.size()), p_bake_userdata, false);
}
HashMap<Edge, EdgeUV2, EdgeHash> edges;
MeshInstance &mi = mesh_instances.write[m_i];
Vector2 uv_scale = Vector2(mi.data.albedo_on_uv2->get_width(), mi.data.albedo_on_uv2->get_height()) / Vector2(atlas_size);
Vector2 uv_offset = Vector2(mi.offset) / Vector2(atlas_size);
if (m_i == 0) {
bounds.position = mi.data.points[0];
}
for (int i = 0; i < mi.data.points.size(); i += 3) {
Vector3 vtxs[3] = { mi.data.points[i + 0], mi.data.points[i + 1], mi.data.points[i + 2] };
Vector2 uvs[3] = { mi.data.uv2[i + 0] * uv_scale + uv_offset, mi.data.uv2[i + 1] * uv_scale + uv_offset, mi.data.uv2[i + 2] * uv_scale + uv_offset };
Vector3 normal[3] = { mi.data.normal[i + 0], mi.data.normal[i + 1], mi.data.normal[i + 2] };
AABB taabb;
Triangle t;
t.slice = mi.slice;
for (int k = 0; k < 3; k++) {
bounds.expand_to(vtxs[k]);
Vertex v;
v.position[0] = vtxs[k].x;
v.position[1] = vtxs[k].y;
v.position[2] = vtxs[k].z;
v.uv[0] = uvs[k].x;
v.uv[1] = uvs[k].y;
v.normal_xy[0] = normal[k].x;
v.normal_xy[1] = normal[k].y;
v.normal_z = normal[k].z;
uint32_t *indexptr = vertex_map.getptr(v);
if (indexptr) {
t.indices[k] = *indexptr;
} else {
uint32_t new_index = vertex_map.size();
t.indices[k] = new_index;
vertex_map[v] = new_index;
vertex_array.push_back(v);
}
if (k == 0) {
taabb.position = vtxs[k];
} else {
taabb.expand_to(vtxs[k]);
}
}
//compute seams that will need to be blended later
for (int k = 0; k < 3; k++) {
int n = (k + 1) % 3;
Edge edge(vtxs[k], vtxs[n], normal[k], normal[n]);
Vector2i edge_indices(t.indices[k], t.indices[n]);
EdgeUV2 uv2(uvs[k], uvs[n], edge_indices);
if (edge.b == edge.a) {
continue; //degenerate, somehow
}
if (edge.b < edge.a) {
SWAP(edge.a, edge.b);
SWAP(edge.na, edge.nb);
SWAP(uv2.a, uv2.b);
SWAP(edge_indices.x, edge_indices.y);
}
EdgeUV2 *euv2 = edges.getptr(edge);
if (!euv2) {
edges[edge] = uv2;
} else {
if (*euv2 == uv2) {
continue; // seam shared UV space, no need to blend
}
if (euv2->seam_found) {
continue; //bad geometry
}
Seam seam;
seam.a = edge_indices;
seam.b = euv2->indices;
seam.slice = mi.slice;
seams.push_back(seam);
slice_seam_count.write[mi.slice]++;
euv2->seam_found = true;
}
}
t.min_bounds[0] = taabb.position.x;
t.min_bounds[1] = taabb.position.y;
t.min_bounds[2] = taabb.position.z;
t.max_bounds[0] = taabb.position.x + MAX(taabb.size.x, 0.0001);
t.max_bounds[1] = taabb.position.y + MAX(taabb.size.y, 0.0001);
t.max_bounds[2] = taabb.position.z + MAX(taabb.size.z, 0.0001);
t.pad0 = t.pad1 = 0; //make valgrind not complain
triangles.push_back(t);
slice_triangle_count.write[t.slice]++;
}
}
//also consider probe positions for bounds
for (int i = 0; i < probe_positions.size(); i++) {
Vector3 pp(probe_positions[i].position[0], probe_positions[i].position[1], probe_positions[i].position[2]);
bounds.expand_to(pp);
}
bounds.grow_by(0.1); //grow a bit to avoid numerical error
triangles.sort(); //sort by slice
seams.sort();
if (p_step_function) {
p_step_function(0.4, RTR("Optimizing acceleration structure"), p_bake_userdata, true);
}
//fill list of triangles in grid
LocalVector<TriangleSort> triangle_sort;
for (uint32_t i = 0; i < triangles.size(); i++) {
const Triangle &t = triangles[i];
Vector3 face[3] = {
Vector3(vertex_array[t.indices[0]].position[0], vertex_array[t.indices[0]].position[1], vertex_array[t.indices[0]].position[2]),
Vector3(vertex_array[t.indices[1]].position[0], vertex_array[t.indices[1]].position[1], vertex_array[t.indices[1]].position[2]),
Vector3(vertex_array[t.indices[2]].position[0], vertex_array[t.indices[2]].position[1], vertex_array[t.indices[2]].position[2])
};
_plot_triangle_into_triangle_index_list(grid_size, Vector3i(), bounds, face, i, triangle_sort, grid_size);
}
//sort it
triangle_sort.sort();
Vector<uint32_t> triangle_indices;
triangle_indices.resize(triangle_sort.size());
Vector<uint32_t> grid_indices;
grid_indices.resize(grid_size * grid_size * grid_size * 2);
memset(grid_indices.ptrw(), 0, grid_indices.size() * sizeof(uint32_t));
Vector<bool> solid;
solid.resize(grid_size * grid_size * grid_size);
memset(solid.ptrw(), 0, solid.size() * sizeof(bool));
{
uint32_t *tiw = triangle_indices.ptrw();
uint32_t last_cell = 0xFFFFFFFF;
uint32_t *giw = grid_indices.ptrw();
bool *solidw = solid.ptrw();
for (uint32_t i = 0; i < triangle_sort.size(); i++) {
uint32_t cell = triangle_sort[i].cell_index;
if (cell != last_cell) {
//cell changed, update pointer to indices
giw[cell * 2 + 1] = i;
solidw[cell] = true;
}
tiw[i] = triangle_sort[i].triangle_index;
giw[cell * 2]++; //update counter
last_cell = cell;
}
}
#if 0
for (int i = 0; i < grid_size; i++) {
for (int j = 0; j < grid_size; j++) {
for (int k = 0; k < grid_size; k++) {
uint32_t index = i * (grid_size * grid_size) + j * grid_size + k;
grid_indices.write[index * 2] = float(i) / grid_size * 255;
grid_indices.write[index * 2 + 1] = float(j) / grid_size * 255;
}
}
}
#endif
#if 0
for (int i = 0; i < grid_size; i++) {
Vector<uint8_t> grid_usage;
grid_usage.resize(grid_size * grid_size);
for (int j = 0; j < grid_usage.size(); j++) {
uint32_t ofs = i * grid_size * grid_size + j;
uint32_t count = grid_indices[ofs * 2];
grid_usage.write[j] = count > 0 ? 255 : 0;
}
Ref<Image> img;
img.instantiate();
img->create(grid_size, grid_size, false, Image::FORMAT_L8, grid_usage);
img->save_png("res://grid_layer_" + itos(1000 + i).substr(1, 3) + ".png");
}
#endif
/*****************************/
/*** CREATE GPU STRUCTURES ***/
/*****************************/
lights.sort();
Vector<Vector2i> seam_buffer_vec;
seam_buffer_vec.resize(seams.size() * 2);
for (uint32_t i = 0; i < seams.size(); i++) {
seam_buffer_vec.write[i * 2 + 0] = seams[i].a;
seam_buffer_vec.write[i * 2 + 1] = seams[i].b;
}
{ //buffers
Vector<uint8_t> vb = vertex_array.to_byte_array();
vertex_buffer = rd->storage_buffer_create(vb.size(), vb);
Vector<uint8_t> tb = triangles.to_byte_array();
triangle_buffer = rd->storage_buffer_create(tb.size(), tb);
Vector<uint8_t> tib = triangle_indices.to_byte_array();
triangle_cell_indices_buffer = rd->storage_buffer_create(tib.size(), tib);
Vector<uint8_t> lb = lights.to_byte_array();
if (lb.size() == 0) {
lb.resize(sizeof(Light)); //even if no lights, the buffer must exist
}
lights_buffer = rd->storage_buffer_create(lb.size(), lb);
Vector<uint8_t> sb = seam_buffer_vec.to_byte_array();
if (sb.size() == 0) {
sb.resize(sizeof(Vector2i) * 2); //even if no seams, the buffer must exist
}
seams_buffer = rd->storage_buffer_create(sb.size(), sb);
Vector<uint8_t> pb = probe_positions.to_byte_array();
if (pb.size() == 0) {
pb.resize(sizeof(Probe));
}
probe_positions_buffer = rd->storage_buffer_create(pb.size(), pb);
}
{ //grid
RD::TextureFormat tf;
tf.width = grid_size;
tf.height = grid_size;
tf.depth = grid_size;
tf.texture_type = RD::TEXTURE_TYPE_3D;
tf.usage_bits = RD::TEXTURE_USAGE_SAMPLING_BIT | RD::TEXTURE_USAGE_CAN_UPDATE_BIT;
Vector<Vector<uint8_t>> texdata;
texdata.resize(1);
//grid and indices
tf.format = RD::DATA_FORMAT_R32G32_UINT;
texdata.write[0] = grid_indices.to_byte_array();
grid_texture = rd->texture_create(tf, RD::TextureView(), texdata);
}
}
void LightmapperRD::_raster_geometry(RenderingDevice *rd, Size2i atlas_size, int atlas_slices, int grid_size, AABB bounds, float p_bias, Vector<int> slice_triangle_count, RID position_tex, RID unocclude_tex, RID normal_tex, RID raster_depth_buffer, RID rasterize_shader, RID raster_base_uniform) {
Vector<RID> framebuffers;
for (int i = 0; i < atlas_slices; i++) {
RID slice_pos_tex = rd->texture_create_shared_from_slice(RD::TextureView(), position_tex, i, 0);
RID slice_unoc_tex = rd->texture_create_shared_from_slice(RD::TextureView(), unocclude_tex, i, 0);
RID slice_norm_tex = rd->texture_create_shared_from_slice(RD::TextureView(), normal_tex, i, 0);
Vector<RID> fb;
fb.push_back(slice_pos_tex);
fb.push_back(slice_norm_tex);
fb.push_back(slice_unoc_tex);
fb.push_back(raster_depth_buffer);
framebuffers.push_back(rd->framebuffer_create(fb));
}
RD::PipelineDepthStencilState ds;
ds.enable_depth_test = true;
ds.enable_depth_write = true;
ds.depth_compare_operator = RD::COMPARE_OP_LESS; //so it does render same pixel twice
RID raster_pipeline = rd->render_pipeline_create(rasterize_shader, rd->framebuffer_get_format(framebuffers[0]), RD::INVALID_FORMAT_ID, RD::RENDER_PRIMITIVE_TRIANGLES, RD::PipelineRasterizationState(), RD::PipelineMultisampleState(), ds, RD::PipelineColorBlendState::create_disabled(3), 0);
RID raster_pipeline_wire;
{
RD::PipelineRasterizationState rw;
rw.wireframe = true;
raster_pipeline_wire = rd->render_pipeline_create(rasterize_shader, rd->framebuffer_get_format(framebuffers[0]), RD::INVALID_FORMAT_ID, RD::RENDER_PRIMITIVE_TRIANGLES, rw, RD::PipelineMultisampleState(), ds, RD::PipelineColorBlendState::create_disabled(3), 0);
}
uint32_t triangle_offset = 0;
Vector<Color> clear_colors;
clear_colors.push_back(Color(0, 0, 0, 0));
clear_colors.push_back(Color(0, 0, 0, 0));
clear_colors.push_back(Color(0, 0, 0, 0));
for (int i = 0; i < atlas_slices; i++) {
RasterPushConstant raster_push_constant;
raster_push_constant.atlas_size[0] = atlas_size.x;
raster_push_constant.atlas_size[1] = atlas_size.y;
raster_push_constant.base_triangle = triangle_offset;
raster_push_constant.to_cell_offset[0] = bounds.position.x;
raster_push_constant.to_cell_offset[1] = bounds.position.y;
raster_push_constant.to_cell_offset[2] = bounds.position.z;
raster_push_constant.bias = p_bias;
raster_push_constant.to_cell_size[0] = (1.0 / bounds.size.x) * float(grid_size);
raster_push_constant.to_cell_size[1] = (1.0 / bounds.size.y) * float(grid_size);
raster_push_constant.to_cell_size[2] = (1.0 / bounds.size.z) * float(grid_size);
raster_push_constant.grid_size[0] = grid_size;
raster_push_constant.grid_size[1] = grid_size;
raster_push_constant.grid_size[2] = grid_size;
raster_push_constant.uv_offset[0] = 0;
raster_push_constant.uv_offset[1] = 0;
RD::DrawListID draw_list = rd->draw_list_begin(framebuffers[i], RD::INITIAL_ACTION_CLEAR, RD::FINAL_ACTION_READ, RD::INITIAL_ACTION_CLEAR, RD::FINAL_ACTION_DISCARD, clear_colors);
//draw opaque
rd->draw_list_bind_render_pipeline(draw_list, raster_pipeline);
rd->draw_list_bind_uniform_set(draw_list, raster_base_uniform, 0);
rd->draw_list_set_push_constant(draw_list, &raster_push_constant, sizeof(RasterPushConstant));
rd->draw_list_draw(draw_list, false, 1, slice_triangle_count[i] * 3);
//draw wire
rd->draw_list_bind_render_pipeline(draw_list, raster_pipeline_wire);
rd->draw_list_bind_uniform_set(draw_list, raster_base_uniform, 0);
rd->draw_list_set_push_constant(draw_list, &raster_push_constant, sizeof(RasterPushConstant));
rd->draw_list_draw(draw_list, false, 1, slice_triangle_count[i] * 3);
rd->draw_list_end();
triangle_offset += slice_triangle_count[i];
}
}
LightmapperRD::BakeError LightmapperRD::_dilate(RenderingDevice *rd, Ref<RDShaderFile> &compute_shader, RID &compute_base_uniform_set, PushConstant &push_constant, RID &source_light_tex, RID &dest_light_tex, const Size2i &atlas_size, int atlas_slices) {
Vector<RD::Uniform> uniforms;
{
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_IMAGE;
u.binding = 0;
u.append_id(dest_light_tex);
uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_TEXTURE;
u.binding = 1;
u.append_id(source_light_tex);
uniforms.push_back(u);
}
}
RID compute_shader_dilate = rd->shader_create_from_spirv(compute_shader->get_spirv_stages("dilate"));
ERR_FAIL_COND_V(compute_shader_dilate.is_null(), BAKE_ERROR_LIGHTMAP_CANT_PRE_BAKE_MESHES); //internal check, should not happen
RID compute_shader_dilate_pipeline = rd->compute_pipeline_create(compute_shader_dilate);
RID dilate_uniform_set = rd->uniform_set_create(uniforms, compute_shader_dilate, 1);
RD::ComputeListID compute_list = rd->compute_list_begin();
rd->compute_list_bind_compute_pipeline(compute_list, compute_shader_dilate_pipeline);
rd->compute_list_bind_uniform_set(compute_list, compute_base_uniform_set, 0);
rd->compute_list_bind_uniform_set(compute_list, dilate_uniform_set, 1);
push_constant.region_ofs[0] = 0;
push_constant.region_ofs[1] = 0;
Vector3i group_size((atlas_size.x - 1) / 8 + 1, (atlas_size.y - 1) / 8 + 1, 1); //restore group size
for (int i = 0; i < atlas_slices; i++) {
push_constant.atlas_slice = i;
rd->compute_list_set_push_constant(compute_list, &push_constant, sizeof(PushConstant));
rd->compute_list_dispatch(compute_list, group_size.x, group_size.y, group_size.z);
//no barrier, let them run all together
}
rd->compute_list_end();
rd->free(compute_shader_dilate);
#ifdef DEBUG_TEXTURES
for (int i = 0; i < atlas_slices; i++) {
Vector<uint8_t> s = rd->texture_get_data(light_accum_tex, i);
Ref<Image> img;
img.instantiate();
img->create(atlas_size.width, atlas_size.height, false, Image::FORMAT_RGBAH, s);
img->convert(Image::FORMAT_RGBA8);
img->save_png("res://5_dilated_" + itos(i) + ".png");
}
#endif
return BAKE_OK;
}
LightmapperRD::BakeError LightmapperRD::bake(BakeQuality p_quality, bool p_use_denoiser, int p_bounces, float p_bias, int p_max_texture_size, bool p_bake_sh, GenerateProbes p_generate_probes, const Ref<Image> &p_environment_panorama, const Basis &p_environment_transform, BakeStepFunc p_step_function, void *p_bake_userdata) {
if (p_step_function) {
p_step_function(0.0, RTR("Begin Bake"), p_bake_userdata, true);
}
bake_textures.clear();
int grid_size = 128;
/* STEP 1: Fetch material textures and compute the bounds */
AABB bounds;
Size2i atlas_size;
int atlas_slices;
Vector<Ref<Image>> albedo_images;
Vector<Ref<Image>> emission_images;
BakeError bake_error = _blit_meshes_into_atlas(p_max_texture_size, albedo_images, emission_images, bounds, atlas_size, atlas_slices, p_step_function, p_bake_userdata);
if (bake_error != BAKE_OK) {
return bake_error;
}
#ifdef DEBUG_TEXTURES
for (int i = 0; i < atlas_slices; i++) {
albedo_images[i]->save_png("res://0_albedo_" + itos(i) + ".png");
emission_images[i]->save_png("res://0_emission_" + itos(i) + ".png");
}
#endif
RenderingDevice *rd = RenderingDevice::get_singleton()->create_local_device();
RID albedo_array_tex;
RID emission_array_tex;
RID normal_tex;
RID position_tex;
RID unocclude_tex;
RID light_source_tex;
RID light_dest_tex;
RID light_accum_tex;
RID light_accum_tex2;
RID light_primary_dynamic_tex;
RID light_environment_tex;
#define FREE_TEXTURES \
rd->free(albedo_array_tex); \
rd->free(emission_array_tex); \
rd->free(normal_tex); \
rd->free(position_tex); \
rd->free(unocclude_tex); \
rd->free(light_source_tex); \
rd->free(light_accum_tex2); \
rd->free(light_accum_tex); \
rd->free(light_primary_dynamic_tex); \
rd->free(light_environment_tex);
{ // create all textures
Vector<Vector<uint8_t>> albedo_data;
Vector<Vector<uint8_t>> emission_data;
for (int i = 0; i < atlas_slices; i++) {
albedo_data.push_back(albedo_images[i]->get_data());
emission_data.push_back(emission_images[i]->get_data());
}
RD::TextureFormat tf;
tf.width = atlas_size.width;
tf.height = atlas_size.height;
tf.array_layers = atlas_slices;
tf.texture_type = RD::TEXTURE_TYPE_2D_ARRAY;
tf.usage_bits = RD::TEXTURE_USAGE_SAMPLING_BIT | RD::TEXTURE_USAGE_CAN_UPDATE_BIT;
tf.format = RD::DATA_FORMAT_R8G8B8A8_UNORM;
albedo_array_tex = rd->texture_create(tf, RD::TextureView(), albedo_data);
tf.format = RD::DATA_FORMAT_R16G16B16A16_SFLOAT;
emission_array_tex = rd->texture_create(tf, RD::TextureView(), emission_data);
//this will be rastered to
tf.usage_bits = RD::TEXTURE_USAGE_SAMPLING_BIT | RD::TEXTURE_USAGE_COLOR_ATTACHMENT_BIT | RD::TEXTURE_USAGE_CAN_COPY_FROM_BIT | RD::TEXTURE_USAGE_STORAGE_BIT;
normal_tex = rd->texture_create(tf, RD::TextureView());
tf.format = RD::DATA_FORMAT_R32G32B32A32_SFLOAT;
position_tex = rd->texture_create(tf, RD::TextureView());
unocclude_tex = rd->texture_create(tf, RD::TextureView());
tf.format = RD::DATA_FORMAT_R16G16B16A16_SFLOAT;
tf.usage_bits = RD::TEXTURE_USAGE_COLOR_ATTACHMENT_BIT | RD::TEXTURE_USAGE_SAMPLING_BIT | RD::TEXTURE_USAGE_STORAGE_BIT | RD::TEXTURE_USAGE_CAN_COPY_FROM_BIT | RD::TEXTURE_USAGE_CAN_COPY_TO_BIT | RD::TEXTURE_USAGE_CAN_UPDATE_BIT;
light_source_tex = rd->texture_create(tf, RD::TextureView());
rd->texture_clear(light_source_tex, Color(0, 0, 0, 0), 0, 1, 0, atlas_slices);
light_primary_dynamic_tex = rd->texture_create(tf, RD::TextureView());
rd->texture_clear(light_primary_dynamic_tex, Color(0, 0, 0, 0), 0, 1, 0, atlas_slices);
if (p_bake_sh) {
tf.array_layers *= 4;
}
light_accum_tex = rd->texture_create(tf, RD::TextureView());
rd->texture_clear(light_accum_tex, Color(0, 0, 0, 0), 0, 1, 0, tf.array_layers);
light_dest_tex = rd->texture_create(tf, RD::TextureView());
rd->texture_clear(light_dest_tex, Color(0, 0, 0, 0), 0, 1, 0, tf.array_layers);
light_accum_tex2 = light_dest_tex;
//env
{
Ref<Image> panorama_tex;
if (p_environment_panorama.is_valid()) {
panorama_tex = p_environment_panorama;
panorama_tex->convert(Image::FORMAT_RGBAF);
} else {
panorama_tex.instantiate();
panorama_tex->create(8, 8, false, Image::FORMAT_RGBAF);
panorama_tex->fill(Color(0, 0, 0, 1));
}
RD::TextureFormat tfp;
tfp.width = panorama_tex->get_width();
tfp.height = panorama_tex->get_height();
tfp.usage_bits = RD::TEXTURE_USAGE_SAMPLING_BIT | RD::TEXTURE_USAGE_CAN_UPDATE_BIT;
tfp.format = RD::DATA_FORMAT_R32G32B32A32_SFLOAT;
Vector<Vector<uint8_t>> tdata;
tdata.push_back(panorama_tex->get_data());
light_environment_tex = rd->texture_create(tfp, RD::TextureView(), tdata);
#ifdef DEBUG_TEXTURES
panorama_tex->save_exr("res://0_panorama.exr", false);
#endif
}
}
/* STEP 2: create the acceleration structure for the GPU*/
Vector<int> slice_triangle_count;
RID vertex_buffer;
RID triangle_buffer;
RID lights_buffer;
RID triangle_cell_indices_buffer;
RID grid_texture;
RID seams_buffer;
RID probe_positions_buffer;
Vector<int> slice_seam_count;
#define FREE_BUFFERS \
rd->free(vertex_buffer); \
rd->free(triangle_buffer); \
rd->free(lights_buffer); \
rd->free(triangle_cell_indices_buffer); \
rd->free(grid_texture); \
rd->free(seams_buffer); \
rd->free(probe_positions_buffer);
_create_acceleration_structures(rd, atlas_size, atlas_slices, bounds, grid_size, probe_positions, p_generate_probes, slice_triangle_count, slice_seam_count, vertex_buffer, triangle_buffer, lights_buffer, triangle_cell_indices_buffer, probe_positions_buffer, grid_texture, seams_buffer, p_step_function, p_bake_userdata);
if (p_step_function) {
p_step_function(0.47, RTR("Preparing shaders"), p_bake_userdata, true);
}
//shaders
Ref<RDShaderFile> raster_shader;
raster_shader.instantiate();
Error err = raster_shader->parse_versions_from_text(lm_raster_shader_glsl);
if (err != OK) {
raster_shader->print_errors("raster_shader");
FREE_TEXTURES
FREE_BUFFERS
memdelete(rd);
}
ERR_FAIL_COND_V(err != OK, BAKE_ERROR_LIGHTMAP_CANT_PRE_BAKE_MESHES);
RID rasterize_shader = rd->shader_create_from_spirv(raster_shader->get_spirv_stages());
ERR_FAIL_COND_V(rasterize_shader.is_null(), BAKE_ERROR_LIGHTMAP_CANT_PRE_BAKE_MESHES); //this is a bug check, though, should not happen
RID sampler;
{
RD::SamplerState s;
s.mag_filter = RD::SAMPLER_FILTER_LINEAR;
s.min_filter = RD::SAMPLER_FILTER_LINEAR;
s.max_lod = 0;
sampler = rd->sampler_create(s);
}
Vector<RD::Uniform> base_uniforms;
{
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_STORAGE_BUFFER;
u.binding = 1;
u.append_id(vertex_buffer);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_STORAGE_BUFFER;
u.binding = 2;
u.append_id(triangle_buffer);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_STORAGE_BUFFER;
u.binding = 3;
u.append_id(triangle_cell_indices_buffer);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_STORAGE_BUFFER;
u.binding = 4;
u.append_id(lights_buffer);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_STORAGE_BUFFER;
u.binding = 5;
u.append_id(seams_buffer);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_STORAGE_BUFFER;
u.binding = 6;
u.append_id(probe_positions_buffer);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_TEXTURE;
u.binding = 7;
u.append_id(grid_texture);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_TEXTURE;
u.binding = 8;
u.append_id(albedo_array_tex);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_TEXTURE;
u.binding = 9;
u.append_id(emission_array_tex);
base_uniforms.push_back(u);
}
{
RD::Uniform u;
u.uniform_type = RD::UNIFORM_TYPE_SAMPLER;
u.binding = 10;
u.append_id(sampler);
base_uniforms.push_back(u);
}
}
RID raster_base_uniform = rd->uniform_set_create(base_uniforms, rasterize_shader, 0);
RID raster_depth_buffer;
{
RD::TextureFormat tf;
tf.width = atlas_size.width;
tf.height = atlas_size.height;
tf.depth = 1;
tf.texture_type = RD::TEXTURE_TYPE_2D;
tf.usage_bits = RD::TEXTURE_USAGE_DEPTH_STENCIL_ATTACHMENT_BIT;
tf.format = RD::DATA_FORMAT_D32_SFLOAT;
raster_depth_buffer = rd->texture_create(tf, RD::TextureView());
}
rd->submit();
rd->sync();
/* STEP 3: Raster the geometry to UV2 coords in the atlas textures GPU*/
_raster_geometry(rd, atlas_size, atlas_slices, grid_size, bounds, p_bias, slice_triangle_count, position_tex, unocclude_tex, normal_tex, raster_depth_buffer, rasterize_shader, raster_base_uniform);
#ifdef DEBUG_TEXTURES
for (int i = 0; i < atlas_slices; i++) {
Vector<uint8_t> s = rd->texture_get_data(position_tex, i);
Ref<Image> img;
img.instantiate();
img->create(atlas_size.width, atlas_size.height, false, Image::FORMAT_RGBAF, s);
img->save_exr("res://1_position_" + itos(i) + ".exr", false);
s = rd->texture_get_data(normal_tex, i);
img->create(atlas_size.width, atlas_size.height, false, Image::FORMAT_RGBAH, s);
img->save_exr("res://1_normal_" + itos(i) + ".exr", false);
}
#endif
#define FREE_RASTER_RESOURCES \
rd->free(rasterize_shader); \
rd->free(sampler); \
rd->free(raster_depth_buffer);
/* Plot direct light */
Ref<RDShaderFile> compute_shader;
compute_shader.instantiate();
err = compute_shader->parse_versions_from_text(lm_compute_shader_glsl, p_bake_sh ? "\n#define USE_SH_LIGHTMAPS\n" : "");
if (err != OK) {
FREE_TEXTURES
FREE_BUFFERS
FREE_RASTER_RESOURCES
memdelete(rd);
compute_shader->print_errors("compute_shader");
}
ERR_FAIL_COND_V(err != OK, BAKE_ERROR_LIGHTMAP_CANT_PRE_BAKE_MESHES);
// Unoccluder
RID compute_shader_unocclude = rd->shader_create_from_spirv(compute_shader->get_spirv_stages("unocclude"));
ERR_FAIL_COND_V(compute_shader_unocclude.is_null(), BAKE_ERROR_LIGHTMAP_CANT_PRE_BAKE_MESHES); // internal check, should not happen
RID compute_shader_unocclude_pipeline = rd->compute_pipeline_create(compute_shader_unocclude);
// Direct light
RID compute_shader_primary = rd->shader_create_from_spirv(compute_shader->get_spirv_stages("primary"));
ERR_FAIL_COND_V(compute_shader_primary.is_null(), BAKE_ERROR_LIGHTMAP_CANT_PRE_BAKE_MESHES); // internal check, should not happen
RID compute_shader_primary_pipeline = rd->compute_pipeline_create(compute_shader_primary);
// Indirect light
RID compute_shader_secondary = rd->shader_create_from_spirv(compute_shader->get_spirv_stages("secondary"));
ERR_FAIL_COND_V(compute_shader_secondary.is_null(), BAKE_ERROR_LIGHTMAP_CANT_PRE_BAKE_MESHES); //internal check, should not happen
RID compute_shader_secondary_pipeline = rd->compute_pipeline_create(compute_shader_secondary);