Welcome to mirror list, hosted at ThFree Co, Russian Federation.

git.blender.org/blender.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat (limited to 'source/blender/render/intern/source/bake.c')
-rw-r--r--source/blender/render/intern/source/bake.c1342
1 files changed, 1342 insertions, 0 deletions
diff --git a/source/blender/render/intern/source/bake.c b/source/blender/render/intern/source/bake.c
new file mode 100644
index 00000000000..4a7962b1776
--- /dev/null
+++ b/source/blender/render/intern/source/bake.c
@@ -0,0 +1,1342 @@
+/*
+ * ***** BEGIN GPL LICENSE BLOCK *****
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ *
+ * Contributors: 2004/2005/2006 Blender Foundation, full recode
+ * Contributors: Vertex color baking, Copyright 2011 AutoCRC
+ *
+ * ***** END GPL LICENSE BLOCK *****
+ */
+
+/** \file blender/render/intern/source/bake.c
+ * \ingroup render
+ */
+
+
+/* system includes */
+#include <stdio.h>
+#include <string.h>
+
+/* External modules: */
+#include "MEM_guardedalloc.h"
+
+#include "BLI_math.h"
+#include "BLI_rand.h"
+#include "BLI_threads.h"
+#include "BLI_utildefines.h"
+
+#include "DNA_image_types.h"
+#include "DNA_material_types.h"
+#include "DNA_mesh_types.h"
+#include "DNA_meshdata_types.h"
+
+#include "BKE_customdata.h"
+#include "BKE_global.h"
+#include "BKE_image.h"
+#include "BKE_main.h"
+#include "BKE_node.h"
+#include "BKE_scene.h"
+#include "BKE_library.h"
+
+#include "IMB_imbuf_types.h"
+#include "IMB_imbuf.h"
+#include "IMB_colormanagement.h"
+
+/* local include */
+#include "rayintersection.h"
+#include "rayobject.h"
+#include "render_types.h"
+#include "renderdatabase.h"
+#include "shading.h"
+#include "zbuf.h"
+
+#include "PIL_time.h"
+
+/* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */
+/* defined in pipeline.c, is hardcopy of active dynamic allocated Render */
+/* only to be used here in this file, it's for speed */
+extern struct Render R;
+/* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */
+
+
+/* ************************* bake ************************ */
+
+
+typedef struct BakeShade {
+ int thread;
+
+ ShadeSample ssamp;
+ ObjectInstanceRen *obi;
+ VlakRen *vlr;
+
+ ZSpan *zspan;
+ Image *ima;
+ ImBuf *ibuf;
+
+ int rectx, recty, quad, type, vdone;
+ bool ready;
+
+ float dir[3];
+ Object *actob;
+
+ /* Output: vertex color or image data. If vcol is not NULL, rect and
+ * rect_float should be NULL. */
+ MPoly *mpoly;
+ MLoop *mloop;
+ MLoopCol *vcol;
+
+ unsigned int *rect;
+ float *rect_float;
+
+ /* displacement buffer used for normalization with unknown maximal distance */
+ bool use_displacement_buffer;
+ float *displacement_buffer;
+ float displacement_min, displacement_max;
+
+ bool use_mask;
+ char *rect_mask; /* bake pixel mask */
+
+ float dxco[3], dyco[3];
+
+ short *do_update;
+
+ struct ColorSpace *rect_colorspace;
+} BakeShade;
+
+static void bake_set_shade_input(ObjectInstanceRen *obi, VlakRen *vlr, ShadeInput *shi, int quad, int UNUSED(isect), int x, int y, float u, float v)
+{
+ if (quad)
+ shade_input_set_triangle_i(shi, obi, vlr, 0, 2, 3);
+ else
+ shade_input_set_triangle_i(shi, obi, vlr, 0, 1, 2);
+
+ /* cache for shadow */
+ shi->samplenr = R.shadowsamplenr[shi->thread]++;
+
+ shi->mask = 0xFFFF; /* all samples */
+
+ shi->u = -u;
+ shi->v = -v;
+ shi->xs = x;
+ shi->ys = y;
+
+ shade_input_set_uv(shi);
+ shade_input_set_normals(shi);
+
+ /* no normal flip */
+ if (shi->flippednor)
+ shade_input_flip_normals(shi);
+
+ /* set up view vector to look right at the surface (note that the normal
+ * is negated in the renderer so it does not need to be done here) */
+ shi->view[0] = shi->vn[0];
+ shi->view[1] = shi->vn[1];
+ shi->view[2] = shi->vn[2];
+}
+
+static void bake_shade(void *handle, Object *ob, ShadeInput *shi, int UNUSED(quad), int x, int y, float UNUSED(u), float UNUSED(v), float *tvn, float *ttang)
+{
+ BakeShade *bs = handle;
+ ShadeSample *ssamp = &bs->ssamp;
+ ShadeResult shr;
+ VlakRen *vlr = shi->vlr;
+
+ shade_input_init_material(shi);
+
+ if (bs->type == RE_BAKE_AO) {
+ ambient_occlusion(shi);
+
+ if (R.r.bake_flag & R_BAKE_NORMALIZE) {
+ copy_v3_v3(shr.combined, shi->ao);
+ }
+ else {
+ zero_v3(shr.combined);
+ environment_lighting_apply(shi, &shr);
+ }
+ }
+ else {
+ if (bs->type == RE_BAKE_SHADOW) /* Why do shadows set the color anyhow?, ignore material color for baking */
+ shi->r = shi->g = shi->b = 1.0f;
+
+ shade_input_set_shade_texco(shi);
+
+ /* only do AO for a full bake (and obviously AO bakes)
+ * AO for light bakes is a leftover and might not be needed */
+ if (ELEM(bs->type, RE_BAKE_ALL, RE_BAKE_AO, RE_BAKE_LIGHT))
+ shade_samples_do_AO(ssamp);
+
+ if (shi->mat->nodetree && shi->mat->use_nodes) {
+ ntreeShaderExecTree(shi->mat->nodetree, shi, &shr);
+ shi->mat = vlr->mat; /* shi->mat is being set in nodetree */
+ }
+ else
+ shade_material_loop(shi, &shr);
+
+ if (bs->type == RE_BAKE_NORMALS) {
+ float nor[3];
+
+ copy_v3_v3(nor, shi->vn);
+
+ if (R.r.bake_normal_space == R_BAKE_SPACE_CAMERA) {
+ /* pass */
+ }
+ else if (R.r.bake_normal_space == R_BAKE_SPACE_TANGENT) {
+ float mat[3][3], imat[3][3];
+
+ /* bitangent */
+ if (tvn && ttang) {
+ copy_v3_v3(mat[0], ttang);
+ cross_v3_v3v3(mat[1], tvn, ttang);
+ mul_v3_fl(mat[1], ttang[3]);
+ copy_v3_v3(mat[2], tvn);
+ }
+ else {
+ copy_v3_v3(mat[0], shi->nmaptang);
+ cross_v3_v3v3(mat[1], shi->nmapnorm, shi->nmaptang);
+ mul_v3_fl(mat[1], shi->nmaptang[3]);
+ copy_v3_v3(mat[2], shi->nmapnorm);
+ }
+
+ invert_m3_m3(imat, mat);
+ mul_m3_v3(imat, nor);
+ }
+ else if (R.r.bake_normal_space == R_BAKE_SPACE_OBJECT)
+ mul_mat3_m4_v3(ob->imat_ren, nor); /* ob->imat_ren includes viewinv! */
+ else if (R.r.bake_normal_space == R_BAKE_SPACE_WORLD)
+ mul_mat3_m4_v3(R.viewinv, nor);
+
+ normalize_v3(nor); /* in case object has scaling */
+
+ /* The invert of the red channel is to make
+ * the normal map compliant with the outside world.
+ * It needs to be done because in Blender
+ * the normal used in the renderer points inward. It is generated
+ * this way in calc_vertexnormals(). Should this ever change
+ * this negate must be removed.
+ *
+ * there is also a small 1e-5f bias for precision issues. otherwise
+ * we randomly get 127 or 128 for neutral colors. we choose 128
+ * because it is the convention flat color. * */
+ shr.combined[0] = (-nor[0]) / 2.0f + 0.5f + 1e-5f;
+ shr.combined[1] = nor[1] / 2.0f + 0.5f + 1e-5f;
+ shr.combined[2] = nor[2] / 2.0f + 0.5f + 1e-5f;
+ }
+ else if (bs->type == RE_BAKE_TEXTURE) {
+ copy_v3_v3(shr.combined, &shi->r);
+ shr.alpha = shi->alpha;
+ }
+ else if (bs->type == RE_BAKE_SHADOW) {
+ copy_v3_v3(shr.combined, shr.shad);
+ shr.alpha = shi->alpha;
+ }
+ else if (bs->type == RE_BAKE_SPEC_COLOR) {
+ copy_v3_v3(shr.combined, &shi->specr);
+ shr.alpha = 1.0f;
+ }
+ else if (bs->type == RE_BAKE_SPEC_INTENSITY) {
+ copy_v3_fl(shr.combined, shi->spec);
+ shr.alpha = 1.0f;
+ }
+ else if (bs->type == RE_BAKE_MIRROR_COLOR) {
+ copy_v3_v3(shr.combined, &shi->mirr);
+ shr.alpha = 1.0f;
+ }
+ else if (bs->type == RE_BAKE_MIRROR_INTENSITY) {
+ copy_v3_fl(shr.combined, shi->ray_mirror);
+ shr.alpha = 1.0f;
+ }
+ else if (bs->type == RE_BAKE_ALPHA) {
+ copy_v3_fl(shr.combined, shi->alpha);
+ shr.alpha = 1.0f;
+ }
+ else if (bs->type == RE_BAKE_EMIT) {
+ copy_v3_fl(shr.combined, shi->emit);
+ shr.alpha = 1.0f;
+ }
+ else if (bs->type == RE_BAKE_VERTEX_COLORS) {
+ copy_v3_v3(shr.combined, shi->vcol);
+ shr.alpha = shi->vcol[3];
+ }
+ }
+
+ if (bs->rect_float && !bs->vcol) {
+ float *col = bs->rect_float + 4 * (bs->rectx * y + x);
+ copy_v3_v3(col, shr.combined);
+ if (bs->type == RE_BAKE_ALL || bs->type == RE_BAKE_TEXTURE || bs->type == RE_BAKE_VERTEX_COLORS) {
+ col[3] = shr.alpha;
+ }
+ else {
+ col[3] = 1.0;
+ }
+ }
+ else {
+ /* Target is char (LDR). */
+ unsigned char col[4];
+
+ if (ELEM(bs->type, RE_BAKE_ALL, RE_BAKE_TEXTURE)) {
+ float rgb[3];
+
+ copy_v3_v3(rgb, shr.combined);
+ if (R.scene_color_manage) {
+ /* Vertex colors have no way to specify color space, so they
+ * default to sRGB. */
+ if (!bs->vcol)
+ IMB_colormanagement_scene_linear_to_colorspace_v3(rgb, bs->rect_colorspace);
+ else
+ linearrgb_to_srgb_v3_v3(rgb, rgb);
+ }
+ rgb_float_to_uchar(col, rgb);
+ }
+ else {
+ rgb_float_to_uchar(col, shr.combined);
+ }
+
+ if (ELEM(bs->type, RE_BAKE_ALL, RE_BAKE_TEXTURE, RE_BAKE_VERTEX_COLORS)) {
+ col[3] = unit_float_to_uchar_clamp(shr.alpha);
+ }
+ else {
+ col[3] = 255;
+ }
+
+ if (bs->vcol) {
+ /* Vertex color baking. Vcol has no useful alpha channel (it exists
+ * but is used only for vertex painting). */
+ bs->vcol->r = col[0];
+ bs->vcol->g = col[1];
+ bs->vcol->b = col[2];
+ }
+ else {
+ unsigned char *imcol = (unsigned char *)(bs->rect + bs->rectx * y + x);
+ copy_v4_v4_uchar(imcol, col);
+ }
+
+ }
+
+ if (bs->rect_mask) {
+ bs->rect_mask[bs->rectx * y + x] = FILTER_MASK_USED;
+ }
+
+ if (bs->do_update) {
+ *bs->do_update = true;
+ }
+}
+
+static void bake_displacement(void *handle, ShadeInput *UNUSED(shi), float dist, int x, int y)
+{
+ BakeShade *bs = handle;
+ float disp;
+
+ if (R.r.bake_flag & R_BAKE_NORMALIZE) {
+ if (R.r.bake_maxdist)
+ disp = (dist + R.r.bake_maxdist) / (R.r.bake_maxdist * 2); /* alter the range from [-bake_maxdist, bake_maxdist] to [0, 1]*/
+ else
+ disp = dist;
+ }
+ else {
+ disp = 0.5f + dist; /* alter the range from [-0.5,0.5] to [0,1]*/
+ }
+
+ if (bs->displacement_buffer) {
+ float *displacement = bs->displacement_buffer + (bs->rectx * y + x);
+ *displacement = disp;
+ bs->displacement_min = min_ff(bs->displacement_min, disp);
+ bs->displacement_max = max_ff(bs->displacement_max, disp);
+ }
+
+ if (bs->rect_float && !bs->vcol) {
+ float *col = bs->rect_float + 4 * (bs->rectx * y + x);
+ col[0] = col[1] = col[2] = disp;
+ col[3] = 1.0f;
+ }
+ else {
+ /* Target is char (LDR). */
+ unsigned char col[4];
+ col[0] = col[1] = col[2] = unit_float_to_uchar_clamp(disp);
+ col[3] = 255;
+
+ if (bs->vcol) {
+ /* Vertex color baking. Vcol has no useful alpha channel (it exists
+ * but is used only for vertex painting). */
+ bs->vcol->r = col[0];
+ bs->vcol->g = col[1];
+ bs->vcol->b = col[2];
+ }
+ else {
+ unsigned char *imcol = (unsigned char *)(bs->rect + bs->rectx * y + x);
+ copy_v4_v4_uchar(imcol, col);
+ }
+ }
+ if (bs->rect_mask) {
+ bs->rect_mask[bs->rectx * y + x] = FILTER_MASK_USED;
+ }
+}
+
+static int bake_intersect_tree(RayObject *raytree, Isect *isect, float *start, float *dir, float sign, float *hitco, float *dist)
+{
+ float maxdist;
+ int hit;
+
+ /* might be useful to make a user setting for maxsize*/
+ if (R.r.bake_maxdist > 0.0f)
+ maxdist = R.r.bake_maxdist;
+ else
+ maxdist = RE_RAYTRACE_MAXDIST + R.r.bake_biasdist;
+
+ /* 'dir' is always normalized */
+ madd_v3_v3v3fl(isect->start, start, dir, -R.r.bake_biasdist);
+
+ mul_v3_v3fl(isect->dir, dir, sign);
+
+ isect->dist = maxdist;
+
+ hit = RE_rayobject_raycast(raytree, isect);
+ if (hit) {
+ madd_v3_v3v3fl(hitco, isect->start, isect->dir, isect->dist);
+
+ *dist = isect->dist;
+ }
+
+ return hit;
+}
+
+static void bake_set_vlr_dxyco(BakeShade *bs, float *uv1, float *uv2, float *uv3)
+{
+ VlakRen *vlr = bs->vlr;
+ float A, d1, d2, d3, *v1, *v2, *v3;
+
+ if (bs->quad) {
+ v1 = vlr->v1->co;
+ v2 = vlr->v3->co;
+ v3 = vlr->v4->co;
+ }
+ else {
+ v1 = vlr->v1->co;
+ v2 = vlr->v2->co;
+ v3 = vlr->v3->co;
+ }
+
+ /* formula derived from barycentric coordinates:
+ * (uvArea1*v1 + uvArea2*v2 + uvArea3*v3)/uvArea
+ * then taking u and v partial derivatives to get dxco and dyco */
+ A = (uv2[0] - uv1[0]) * (uv3[1] - uv1[1]) - (uv3[0] - uv1[0]) * (uv2[1] - uv1[1]);
+
+ if (fabsf(A) > FLT_EPSILON) {
+ A = 0.5f / A;
+
+ d1 = uv2[1] - uv3[1];
+ d2 = uv3[1] - uv1[1];
+ d3 = uv1[1] - uv2[1];
+ bs->dxco[0] = (v1[0] * d1 + v2[0] * d2 + v3[0] * d3) * A;
+ bs->dxco[1] = (v1[1] * d1 + v2[1] * d2 + v3[1] * d3) * A;
+ bs->dxco[2] = (v1[2] * d1 + v2[2] * d2 + v3[2] * d3) * A;
+
+ d1 = uv3[0] - uv2[0];
+ d2 = uv1[0] - uv3[0];
+ d3 = uv2[0] - uv1[0];
+ bs->dyco[0] = (v1[0] * d1 + v2[0] * d2 + v3[0] * d3) * A;
+ bs->dyco[1] = (v1[1] * d1 + v2[1] * d2 + v3[1] * d3) * A;
+ bs->dyco[2] = (v1[2] * d1 + v2[2] * d2 + v3[2] * d3) * A;
+ }
+ else {
+ bs->dxco[0] = bs->dxco[1] = bs->dxco[2] = 0.0f;
+ bs->dyco[0] = bs->dyco[1] = bs->dyco[2] = 0.0f;
+ }
+
+ if (bs->obi->flag & R_TRANSFORMED) {
+ mul_m3_v3(bs->obi->nmat, bs->dxco);
+ mul_m3_v3(bs->obi->nmat, bs->dyco);
+ }
+}
+
+static void do_bake_shade(void *handle, int x, int y, float u, float v)
+{
+ BakeShade *bs = handle;
+ VlakRen *vlr = bs->vlr;
+ ObjectInstanceRen *obi = bs->obi;
+ Object *ob = obi->obr->ob;
+ float l, *v1, *v2, *v3, tvn[3], ttang[4];
+ int quad;
+ ShadeSample *ssamp = &bs->ssamp;
+ ShadeInput *shi = ssamp->shi;
+
+ /* fast threadsafe break test */
+ if (R.test_break(R.tbh))
+ return;
+
+ /* setup render coordinates */
+ if (bs->quad) {
+ v1 = vlr->v1->co;
+ v2 = vlr->v3->co;
+ v3 = vlr->v4->co;
+ }
+ else {
+ v1 = vlr->v1->co;
+ v2 = vlr->v2->co;
+ v3 = vlr->v3->co;
+ }
+
+ l = 1.0f - u - v;
+
+ /* shrink barycentric coordinates inwards slightly to avoid some issues
+ * where baking selected to active might just miss the other face at the
+ * near the edge of a face */
+ if (bs->actob) {
+ const float eps = 1.0f - 1e-4f;
+ float invsum;
+
+ u = (u - 0.5f) * eps + 0.5f;
+ v = (v - 0.5f) * eps + 0.5f;
+ l = (l - 0.5f) * eps + 0.5f;
+
+ invsum = 1.0f / (u + v + l);
+
+ u *= invsum;
+ v *= invsum;
+ l *= invsum;
+ }
+
+ /* renderco */
+ shi->co[0] = l * v3[0] + u * v1[0] + v * v2[0];
+ shi->co[1] = l * v3[1] + u * v1[1] + v * v2[1];
+ shi->co[2] = l * v3[2] + u * v1[2] + v * v2[2];
+
+ /* avoid self shadow with vertex bake from adjacent faces [#33729] */
+ if ((bs->vcol != NULL) && (bs->actob == NULL)) {
+ madd_v3_v3fl(shi->co, vlr->n, 0.0001f);
+ }
+
+ if (obi->flag & R_TRANSFORMED)
+ mul_m4_v3(obi->mat, shi->co);
+
+ copy_v3_v3(shi->dxco, bs->dxco);
+ copy_v3_v3(shi->dyco, bs->dyco);
+
+ quad = bs->quad;
+ bake_set_shade_input(obi, vlr, shi, quad, 0, x, y, u, v);
+
+ if (bs->type == RE_BAKE_NORMALS && R.r.bake_normal_space == R_BAKE_SPACE_TANGENT) {
+ shade_input_set_shade_texco(shi);
+ copy_v3_v3(tvn, shi->nmapnorm);
+ copy_v4_v4(ttang, shi->nmaptang);
+ }
+
+ /* if we are doing selected to active baking, find point on other face */
+ if (bs->actob) {
+ Isect isec, minisec;
+ float co[3], minco[3], dist, mindist = 0.0f;
+ int hit, sign, dir = 1;
+
+ /* intersect with ray going forward and backward*/
+ hit = 0;
+ memset(&minisec, 0, sizeof(minisec));
+ minco[0] = minco[1] = minco[2] = 0.0f;
+
+ copy_v3_v3(bs->dir, shi->vn);
+
+ for (sign = -1; sign <= 1; sign += 2) {
+ memset(&isec, 0, sizeof(isec));
+ isec.mode = RE_RAY_MIRROR;
+
+ isec.orig.ob = obi;
+ isec.orig.face = vlr;
+ isec.userdata = bs->actob;
+ isec.check = RE_CHECK_VLR_BAKE;
+ isec.skip = RE_SKIP_VLR_NEIGHBOUR;
+
+ if (bake_intersect_tree(R.raytree, &isec, shi->co, shi->vn, sign, co, &dist)) {
+ if (!hit || len_squared_v3v3(shi->co, co) < len_squared_v3v3(shi->co, minco)) {
+ minisec = isec;
+ mindist = dist;
+ copy_v3_v3(minco, co);
+ hit = 1;
+ dir = sign;
+ }
+ }
+ }
+
+ if (ELEM(bs->type, RE_BAKE_DISPLACEMENT, RE_BAKE_DERIVATIVE)) {
+ if (hit)
+ bake_displacement(handle, shi, (dir == -1) ? mindist : -mindist, x, y);
+ else
+ bake_displacement(handle, shi, 0.0f, x, y);
+ return;
+ }
+
+ /* if hit, we shade from the new point, otherwise from point one starting face */
+ if (hit) {
+ obi = (ObjectInstanceRen *)minisec.hit.ob;
+ vlr = (VlakRen *)minisec.hit.face;
+ quad = (minisec.isect == 2);
+ copy_v3_v3(shi->co, minco);
+
+ u = -minisec.u;
+ v = -minisec.v;
+ bake_set_shade_input(obi, vlr, shi, quad, 1, x, y, u, v);
+ }
+ }
+
+ if (bs->type == RE_BAKE_NORMALS && R.r.bake_normal_space == R_BAKE_SPACE_TANGENT)
+ bake_shade(handle, ob, shi, quad, x, y, u, v, tvn, ttang);
+ else
+ bake_shade(handle, ob, shi, quad, x, y, u, v, NULL, NULL);
+}
+
+static int get_next_bake_face(BakeShade *bs)
+{
+ ObjectRen *obr;
+ VlakRen *vlr;
+ MTFace *tface;
+ static int v = 0, vdone = false;
+ static ObjectInstanceRen *obi = NULL;
+
+ if (bs == NULL) {
+ vlr = NULL;
+ v = vdone = false;
+ obi = R.instancetable.first;
+ return 0;
+ }
+
+ BLI_thread_lock(LOCK_CUSTOM1);
+
+ for (; obi; obi = obi->next, v = 0) {
+ obr = obi->obr;
+
+ /* only allow non instances here */
+ if (obr->flag & R_INSTANCEABLE)
+ continue;
+
+ for (; v < obr->totvlak; v++) {
+ vlr = RE_findOrAddVlak(obr, v);
+
+ if ((bs->actob && bs->actob == obr->ob) || (!bs->actob && (obr->ob->flag & SELECT))) {
+ if (R.r.bake_flag & R_BAKE_VCOL) {
+ /* Gather face data for vertex color bake */
+ Mesh *me;
+ int *origindex, vcollayer;
+ CustomDataLayer *cdl;
+
+ if (obr->ob->type != OB_MESH)
+ continue;
+ me = obr->ob->data;
+
+ origindex = RE_vlakren_get_origindex(obr, vlr, 0);
+ if (origindex == NULL)
+ continue;
+ if (*origindex >= me->totpoly) {
+ /* Small hack for Array modifier, which gives false
+ * original indices - z0r */
+ continue;
+ }
+#if 0
+ /* Only shade selected faces. */
+ if ((me->mface[*origindex].flag & ME_FACE_SEL) == 0)
+ continue;
+#endif
+
+ vcollayer = CustomData_get_render_layer_index(&me->ldata, CD_MLOOPCOL);
+ if (vcollayer == -1)
+ continue;
+
+ cdl = &me->ldata.layers[vcollayer];
+ bs->mpoly = me->mpoly + *origindex;
+ bs->vcol = ((MLoopCol *)cdl->data) + bs->mpoly->loopstart;
+ bs->mloop = me->mloop + bs->mpoly->loopstart;
+
+ /* Tag mesh for reevaluation. */
+ me->id.tag |= LIB_TAG_DOIT;
+ }
+ else {
+ Image *ima = NULL;
+ ImBuf *ibuf = NULL;
+ const float vec_alpha[4] = {0.0f, 0.0f, 0.0f, 0.0f};
+ const float vec_solid[4] = {0.0f, 0.0f, 0.0f, 1.0f};
+ const float nor_alpha[4] = {0.5f, 0.5f, 1.0f, 0.0f};
+ const float nor_solid[4] = {0.5f, 0.5f, 1.0f, 1.0f};
+ const float disp_alpha[4] = {0.5f, 0.5f, 0.5f, 0.0f};
+ const float disp_solid[4] = {0.5f, 0.5f, 0.5f, 1.0f};
+
+ tface = RE_vlakren_get_tface(obr, vlr, obr->bakemtface, NULL, 0);
+
+ if (!tface || !tface->tpage)
+ continue;
+
+ ima = tface->tpage;
+ ibuf = BKE_image_acquire_ibuf(ima, NULL, NULL);
+
+ if (ibuf == NULL)
+ continue;
+
+ if (ibuf->rect == NULL && ibuf->rect_float == NULL) {
+ BKE_image_release_ibuf(ima, ibuf, NULL);
+ continue;
+ }
+
+ if (ibuf->rect_float && !(ibuf->channels == 0 || ibuf->channels == 4)) {
+ BKE_image_release_ibuf(ima, ibuf, NULL);
+ continue;
+ }
+
+ if (ima->flag & IMA_USED_FOR_RENDER) {
+ ima->id.tag &= ~LIB_TAG_DOIT;
+ BKE_image_release_ibuf(ima, ibuf, NULL);
+ continue;
+ }
+
+ /* find the image for the first time? */
+ if (ima->id.tag & LIB_TAG_DOIT) {
+ ima->id.tag &= ~LIB_TAG_DOIT;
+
+ /* we either fill in float or char, this ensures things go fine */
+ if (ibuf->rect_float)
+ imb_freerectImBuf(ibuf);
+ /* clear image */
+ if (R.r.bake_flag & R_BAKE_CLEAR) {
+ if (R.r.bake_mode == RE_BAKE_NORMALS && R.r.bake_normal_space == R_BAKE_SPACE_TANGENT)
+ IMB_rectfill(ibuf, (ibuf->planes == R_IMF_PLANES_RGBA) ? nor_alpha : nor_solid);
+ else if (ELEM(R.r.bake_mode, RE_BAKE_DISPLACEMENT, RE_BAKE_DERIVATIVE))
+ IMB_rectfill(ibuf, (ibuf->planes == R_IMF_PLANES_RGBA) ? disp_alpha : disp_solid);
+ else
+ IMB_rectfill(ibuf, (ibuf->planes == R_IMF_PLANES_RGBA) ? vec_alpha : vec_solid);
+ }
+ /* might be read by UI to set active image for display */
+ R.bakebuf = ima;
+ }
+
+ /* Tag image for redraw. */
+ ibuf->userflags |= IB_DISPLAY_BUFFER_INVALID;
+ BKE_image_release_ibuf(ima, ibuf, NULL);
+ }
+
+ bs->obi = obi;
+ bs->vlr = vlr;
+ bs->vdone++; /* only for error message if nothing was rendered */
+ v++;
+ BLI_thread_unlock(LOCK_CUSTOM1);
+ return 1;
+ }
+ }
+ }
+
+ BLI_thread_unlock(LOCK_CUSTOM1);
+ return 0;
+}
+
+static void bake_single_vertex(BakeShade *bs, VertRen *vert, float u, float v)
+{
+ int *origindex, i;
+ MLoopCol *basevcol;
+ MLoop *mloop;
+
+ /* per vertex fixed seed */
+ BLI_thread_srandom(bs->thread, vert->index);
+
+ origindex = RE_vertren_get_origindex(bs->obi->obr, vert, 0);
+ if (!origindex || *origindex == ORIGINDEX_NONE)
+ return;
+
+ /* Search for matching vertex index and apply shading. */
+ for (i = 0; i < bs->mpoly->totloop; i++) {
+ mloop = bs->mloop + i;
+ if (mloop->v != *origindex)
+ continue;
+ basevcol = bs->vcol;
+ bs->vcol = basevcol + i;
+ do_bake_shade(bs, 0, 0, u, v);
+ bs->vcol = basevcol;
+ break;
+ }
+}
+
+/* Bake all vertices of a face. Actually, this still works on a face-by-face
+ * basis, and each vertex on each face is shaded. Vertex colors are a property
+ * of loops, not vertices. */
+static void shade_verts(BakeShade *bs)
+{
+ VlakRen *vlr = bs->vlr;
+
+ /* Disable baking to image; write to vcol instead. vcol pointer is set in
+ * bake_single_vertex. */
+ bs->ima = NULL;
+ bs->rect = NULL;
+ bs->rect_float = NULL;
+ bs->displacement_buffer = NULL;
+ bs->displacement_min = FLT_MAX;
+ bs->displacement_max = -FLT_MAX;
+
+ bs->quad = 0;
+
+ /* No anti-aliasing for vertices. */
+ zero_v3(bs->dxco);
+ zero_v3(bs->dyco);
+
+ /* Shade each vertex of the face. u and v are barycentric coordinates; since
+ * we're only interested in vertices, these will be 0 or 1. */
+ if ((vlr->flag & R_FACE_SPLIT) == 0) {
+ /* Processing triangle face, whole quad, or first half of split quad. */
+
+ bake_single_vertex(bs, bs->vlr->v1, 1.0f, 0.0f);
+ bake_single_vertex(bs, bs->vlr->v2, 0.0f, 1.0f);
+ bake_single_vertex(bs, bs->vlr->v3, 0.0f, 0.0f);
+
+ if (vlr->v4) {
+ bs->quad = 1;
+ bake_single_vertex(bs, bs->vlr->v4, 0.0f, 0.0f);
+ }
+ }
+ else {
+ /* Processing second half of split quad. Only one vertex to go. */
+ if (vlr->flag & R_DIVIDE_24) {
+ bake_single_vertex(bs, bs->vlr->v2, 0.0f, 1.0f);
+ }
+ else {
+ bake_single_vertex(bs, bs->vlr->v3, 0.0f, 0.0f);
+ }
+ }
+}
+
+/* already have tested for tface and ima and zspan */
+static void shade_tface(BakeShade *bs)
+{
+ VlakRen *vlr = bs->vlr;
+ ObjectInstanceRen *obi = bs->obi;
+ ObjectRen *obr = obi->obr;
+ MTFace *tface = RE_vlakren_get_tface(obr, vlr, obr->bakemtface, NULL, 0);
+ Image *ima = tface->tpage;
+ float vec[4][2];
+ int a, i1, i2, i3;
+
+ /* per face fixed seed */
+ BLI_thread_srandom(bs->thread, vlr->index);
+
+ /* check valid zspan */
+ if (ima != bs->ima) {
+ BKE_image_release_ibuf(bs->ima, bs->ibuf, NULL);
+
+ bs->ima = ima;
+ bs->ibuf = BKE_image_acquire_ibuf(ima, NULL, NULL);
+ /* note, these calls only free/fill contents of zspan struct, not zspan itself */
+ zbuf_free_span(bs->zspan);
+ zbuf_alloc_span(bs->zspan, bs->ibuf->x, bs->ibuf->y, R.clipcrop);
+ }
+
+ bs->rectx = bs->ibuf->x;
+ bs->recty = bs->ibuf->y;
+ bs->rect = bs->ibuf->rect;
+ bs->rect_colorspace = bs->ibuf->rect_colorspace;
+ bs->rect_float = bs->ibuf->rect_float;
+ bs->vcol = NULL;
+ bs->quad = 0;
+ bs->rect_mask = NULL;
+ bs->displacement_buffer = NULL;
+
+ if (bs->use_mask || bs->use_displacement_buffer) {
+ BakeImBufuserData *userdata = bs->ibuf->userdata;
+ if (userdata == NULL) {
+ BLI_thread_lock(LOCK_CUSTOM1);
+ userdata = bs->ibuf->userdata;
+ if (userdata == NULL) /* since the thread was locked, its possible another thread alloced the value */
+ userdata = MEM_callocN(sizeof(BakeImBufuserData), "BakeImBufuserData");
+
+ if (bs->use_mask) {
+ if (userdata->mask_buffer == NULL) {
+ userdata->mask_buffer = MEM_callocN(sizeof(char) * bs->rectx * bs->recty, "BakeMask");
+ }
+ }
+
+ if (bs->use_displacement_buffer) {
+ if (userdata->displacement_buffer == NULL) {
+ userdata->displacement_buffer = MEM_callocN(sizeof(float) * bs->rectx * bs->recty, "BakeDisp");
+ }
+ }
+
+ bs->ibuf->userdata = userdata;
+
+ BLI_thread_unlock(LOCK_CUSTOM1);
+ }
+
+ bs->rect_mask = userdata->mask_buffer;
+ bs->displacement_buffer = userdata->displacement_buffer;
+ }
+
+ /* get pixel level vertex coordinates */
+ for (a = 0; a < 4; a++) {
+ /* Note, workaround for pixel aligned UVs which are common and can screw up our intersection tests
+ * where a pixel gets in between 2 faces or the middle of a quad,
+ * camera aligned quads also have this problem but they are less common.
+ * Add a small offset to the UVs, fixes bug #18685 - Campbell */
+ vec[a][0] = tface->uv[a][0] * (float)bs->rectx - (0.5f + 0.001f);
+ vec[a][1] = tface->uv[a][1] * (float)bs->recty - (0.5f + 0.002f);
+ }
+
+ /* UV indices have to be corrected for possible quad->tria splits */
+ i1 = 0; i2 = 1; i3 = 2;
+ vlr_set_uv_indices(vlr, &i1, &i2, &i3);
+ bake_set_vlr_dxyco(bs, vec[i1], vec[i2], vec[i3]);
+ zspan_scanconvert(bs->zspan, bs, vec[i1], vec[i2], vec[i3], do_bake_shade);
+
+ if (vlr->v4) {
+ bs->quad = 1;
+ bake_set_vlr_dxyco(bs, vec[0], vec[2], vec[3]);
+ zspan_scanconvert(bs->zspan, bs, vec[0], vec[2], vec[3], do_bake_shade);
+ }
+}
+
+static void *do_bake_thread(void *bs_v)
+{
+ BakeShade *bs = bs_v;
+
+ while (get_next_bake_face(bs)) {
+ if (R.r.bake_flag & R_BAKE_VCOL) {
+ shade_verts(bs);
+ }
+ else {
+ shade_tface(bs);
+ }
+
+ /* fast threadsafe break test */
+ if (R.test_break(R.tbh))
+ break;
+
+ /* access is not threadsafe but since its just true/false probably ok
+ * only used for interactive baking */
+ if (bs->do_update) {
+ *bs->do_update = true;
+ }
+ }
+ bs->ready = true;
+
+ BKE_image_release_ibuf(bs->ima, bs->ibuf, NULL);
+
+ return NULL;
+}
+
+void RE_bake_ibuf_filter(ImBuf *ibuf, char *mask, const int filter)
+{
+ /* must check before filtering */
+ const bool is_new_alpha = (ibuf->planes != R_IMF_PLANES_RGBA) && BKE_imbuf_alpha_test(ibuf);
+
+ /* Margin */
+ if (filter) {
+ IMB_filter_extend(ibuf, mask, filter);
+ }
+
+ /* if the bake results in new alpha then change the image setting */
+ if (is_new_alpha) {
+ ibuf->planes = R_IMF_PLANES_RGBA;
+ }
+ else {
+ if (filter && ibuf->planes != R_IMF_PLANES_RGBA) {
+ /* clear alpha added by filtering */
+ IMB_rectfill_alpha(ibuf, 1.0f);
+ }
+ }
+}
+
+void RE_bake_ibuf_normalize_displacement(ImBuf *ibuf, float *displacement, char *mask, float displacement_min, float displacement_max)
+{
+ int i;
+ const float *current_displacement = displacement;
+ const char *current_mask = mask;
+ float max_distance;
+
+ max_distance = max_ff(fabsf(displacement_min), fabsf(displacement_max));
+
+ for (i = 0; i < ibuf->x * ibuf->y; i++) {
+ if (*current_mask == FILTER_MASK_USED) {
+ float normalized_displacement;
+
+ if (max_distance > 1e-5f)
+ normalized_displacement = (*current_displacement + max_distance) / (max_distance * 2);
+ else
+ normalized_displacement = 0.5f;
+
+ if (ibuf->rect_float) {
+ /* currently baking happens to RGBA only */
+ float *fp = ibuf->rect_float + i * 4;
+ fp[0] = fp[1] = fp[2] = normalized_displacement;
+ fp[3] = 1.0f;
+ }
+
+ if (ibuf->rect) {
+ unsigned char *cp = (unsigned char *) (ibuf->rect + i);
+ cp[0] = cp[1] = cp[2] = unit_float_to_uchar_clamp(normalized_displacement);
+ cp[3] = 255;
+ }
+ }
+
+ current_displacement++;
+ current_mask++;
+ }
+}
+
+/* using object selection tags, the faces with UV maps get baked */
+/* render should have been setup */
+/* returns 0 if nothing was handled */
+int RE_bake_shade_all_selected(Render *re, int type, Object *actob, short *do_update, float *progress)
+{
+ BakeShade *handles;
+ ListBase threads;
+ Image *ima;
+ int a, vdone = false, result = BAKE_RESULT_OK;
+ bool use_mask = false;
+ bool use_displacement_buffer = false;
+ bool do_manage = false;
+
+ if (ELEM(type, RE_BAKE_ALL, RE_BAKE_TEXTURE)) {
+ do_manage = BKE_scene_check_color_management_enabled(re->scene);
+ }
+
+ re->scene_color_manage = BKE_scene_check_color_management_enabled(re->scene);
+
+ /* initialize render global */
+ R = *re;
+ R.bakebuf = NULL;
+
+ /* initialize static vars */
+ get_next_bake_face(NULL);
+
+ /* do we need a mask? */
+ if (re->r.bake_filter)
+ use_mask = true;
+
+ /* do we need buffer to store displacements */
+ if (ELEM(type, RE_BAKE_DISPLACEMENT, RE_BAKE_DERIVATIVE)) {
+ if (((R.r.bake_flag & R_BAKE_NORMALIZE) && R.r.bake_maxdist == 0.0f) ||
+ (type == RE_BAKE_DERIVATIVE))
+ {
+ use_displacement_buffer = true;
+ use_mask = true;
+ }
+ }
+
+ /* baker uses this flag to detect if image was initialized */
+ if ((R.r.bake_flag & R_BAKE_VCOL) == 0) {
+ for (ima = G.main->image.first; ima; ima = ima->id.next) {
+ ImBuf *ibuf = BKE_image_acquire_ibuf(ima, NULL, NULL);
+ ima->id.tag |= LIB_TAG_DOIT;
+ ima->flag &= ~IMA_USED_FOR_RENDER;
+ if (ibuf) {
+ ibuf->userdata = NULL; /* use for masking if needed */
+ }
+ BKE_image_release_ibuf(ima, ibuf, NULL);
+ }
+ }
+
+ if (R.r.bake_flag & R_BAKE_VCOL) {
+ /* untag all meshes */
+ BKE_main_id_tag_listbase(&G.main->mesh, LIB_TAG_DOIT, false);
+ }
+
+ BLI_threadpool_init(&threads, do_bake_thread, re->r.threads);
+
+ handles = MEM_callocN(sizeof(BakeShade) * re->r.threads, "BakeShade");
+
+ /* get the threads running */
+ for (a = 0; a < re->r.threads; a++) {
+ handles[a].thread = a;
+
+ /* set defaults in handles */
+ handles[a].ssamp.shi[0].lay = re->lay;
+
+ if (type == RE_BAKE_SHADOW) {
+ handles[a].ssamp.shi[0].passflag = SCE_PASS_SHADOW;
+ }
+ else {
+ handles[a].ssamp.shi[0].passflag = SCE_PASS_COMBINED;
+ }
+ handles[a].ssamp.shi[0].combinedflag = ~(SCE_PASS_SPEC);
+ handles[a].ssamp.shi[0].thread = a;
+ handles[a].ssamp.shi[0].do_manage = do_manage;
+ handles[a].ssamp.tot = 1;
+
+ handles[a].type = type;
+ handles[a].actob = actob;
+ if (R.r.bake_flag & R_BAKE_VCOL)
+ handles[a].zspan = NULL;
+ else
+ handles[a].zspan = MEM_callocN(sizeof(ZSpan), "zspan for bake");
+
+ handles[a].use_mask = use_mask;
+ handles[a].use_displacement_buffer = use_displacement_buffer;
+
+ handles[a].do_update = do_update; /* use to tell the view to update */
+
+ handles[a].displacement_min = FLT_MAX;
+ handles[a].displacement_max = -FLT_MAX;
+
+ BLI_threadpool_insert(&threads, &handles[a]);
+ }
+
+ /* wait for everything to be done */
+ a = 0;
+ while (a != re->r.threads) {
+ PIL_sleep_ms(50);
+
+ /* calculate progress */
+ for (vdone = false, a = 0; a < re->r.threads; a++)
+ vdone += handles[a].vdone;
+ if (progress)
+ *progress = (float)(vdone / (float)re->totvlak);
+
+ for (a = 0; a < re->r.threads; a++) {
+ if (handles[a].ready == false) {
+ break;
+ }
+ }
+ }
+
+ /* filter and refresh images */
+ if ((R.r.bake_flag & R_BAKE_VCOL) == 0) {
+ float displacement_min = FLT_MAX, displacement_max = -FLT_MAX;
+
+ if (use_displacement_buffer) {
+ for (a = 0; a < re->r.threads; a++) {
+ displacement_min = min_ff(displacement_min, handles[a].displacement_min);
+ displacement_max = max_ff(displacement_max, handles[a].displacement_max);
+ }
+ }
+
+ for (ima = G.main->image.first; ima; ima = ima->id.next) {
+ if ((ima->id.tag & LIB_TAG_DOIT) == 0) {
+ ImBuf *ibuf = BKE_image_acquire_ibuf(ima, NULL, NULL);
+ BakeImBufuserData *userdata;
+
+ if (ima->flag & IMA_USED_FOR_RENDER)
+ result = BAKE_RESULT_FEEDBACK_LOOP;
+
+ if (!ibuf)
+ continue;
+
+ userdata = (BakeImBufuserData *)ibuf->userdata;
+ if (userdata) {
+ if (use_displacement_buffer) {
+ if (type == RE_BAKE_DERIVATIVE) {
+ float user_scale = (R.r.bake_flag & R_BAKE_USERSCALE) ? R.r.bake_user_scale : -1.0f;
+ RE_bake_make_derivative(ibuf, userdata->displacement_buffer, userdata->mask_buffer,
+ displacement_min, displacement_max, user_scale);
+ }
+ else {
+ RE_bake_ibuf_normalize_displacement(ibuf, userdata->displacement_buffer, userdata->mask_buffer,
+ displacement_min, displacement_max);
+ }
+ }
+
+ RE_bake_ibuf_filter(ibuf, userdata->mask_buffer, re->r.bake_filter);
+ }
+
+ ibuf->userflags |= IB_BITMAPDIRTY;
+ BKE_image_release_ibuf(ima, ibuf, NULL);
+ }
+ }
+
+ /* calculate return value */
+ for (a = 0; a < re->r.threads; a++) {
+ zbuf_free_span(handles[a].zspan);
+ MEM_freeN(handles[a].zspan);
+ }
+ }
+
+ MEM_freeN(handles);
+
+ BLI_threadpool_end(&threads);
+
+ if (vdone == 0) {
+ result = BAKE_RESULT_NO_OBJECTS;
+ }
+
+ return result;
+}
+
+struct Image *RE_bake_shade_get_image(void)
+{
+ return R.bakebuf;
+}
+
+/* **************** Derivative Maps Baker **************** */
+
+static void add_single_heights_margin(const ImBuf *ibuf, const char *mask, float *heights_buffer)
+{
+ int x, y;
+
+ for (y = 0; y < ibuf->y; y++) {
+ for (x = 0; x < ibuf->x; x++) {
+ int index = ibuf->x * y + x;
+
+ /* If unassigned pixel, look for neighbors. */
+ if (mask[index] != FILTER_MASK_USED) {
+ float height_acc = 0;
+ int denom = 0;
+ int i, j;
+
+ for (j = -1; j <= 1; j++)
+ for (i = -1; i <= 1; i++) {
+ int w = (i == 0 ? 1 : 0) + (j == 0 ? 1 : 0) + 1;
+
+ if (i != 0 || j != 0) {
+ int index2 = 0;
+ int x0 = x + i;
+ int y0 = y + j;
+
+ CLAMP(x0, 0, ibuf->x - 1);
+ CLAMP(y0, 0, ibuf->y - 1);
+
+ index2 = ibuf->x * y0 + x0;
+
+ if (mask[index2] == FILTER_MASK_USED) {
+ height_acc += w * heights_buffer[index2];
+ denom += w;
+ }
+ }
+ }
+
+ /* Insert final value. */
+ if (denom > 0) {
+ heights_buffer[index] = height_acc / denom;
+ }
+ }
+ }
+ }
+}
+
+/* returns user-scale */
+float RE_bake_make_derivative(ImBuf *ibuf, float *heights_buffer, const char *mask,
+ const float height_min, const float height_max,
+ const float fmult)
+{
+ const float delta_height = height_max - height_min;
+ const float denom = delta_height > 0.0f ? (8 * delta_height) : 1.0f;
+ bool auto_range_fit = fmult <= 0.0f;
+ float max_num_deriv = -1.0f;
+ int x, y, index;
+
+ /* Need a single margin to calculate good derivatives. */
+ add_single_heights_margin(ibuf, mask, heights_buffer);
+
+ if (auto_range_fit) {
+ /* If automatic range fitting is enabled. */
+ for (y = 0; y < ibuf->y; y++) {
+ const int Yu = y == (ibuf->y - 1) ? (ibuf->y - 1) : (y + 1);
+ const int Yc = y;
+ const int Yd = y == 0 ? 0 : (y - 1);
+
+ for (x = 0; x < ibuf->x; x++) {
+ const int Xl = x == 0 ? 0 : (x - 1);
+ const int Xc = x;
+ const int Xr = x == (ibuf->x - 1) ? (ibuf->x - 1) : (x + 1);
+
+ const float Hcy = heights_buffer[Yc * ibuf->x + Xr] - heights_buffer[Yc * ibuf->x + Xl];
+ const float Hu = heights_buffer[Yu * ibuf->x + Xr] - heights_buffer[Yu * ibuf->x + Xl];
+ const float Hd = heights_buffer[Yd * ibuf->x + Xr] - heights_buffer[Yd * ibuf->x + Xl];
+
+ const float Hl = heights_buffer[Yu * ibuf->x + Xl] - heights_buffer[Yd * ibuf->x + Xl];
+ const float Hcx = heights_buffer[Yu * ibuf->x + Xc] - heights_buffer[Yd * ibuf->x + Xc];
+ const float Hr = heights_buffer[Yu * ibuf->x + Xr] - heights_buffer[Yd * ibuf->x + Xr];
+
+ /* This corresponds to using the sobel kernel on the heights buffer
+ * to obtain the derivative multiplied by 8.
+ */
+ const float deriv_x = Hu + 2 * Hcy + Hd;
+ const float deriv_y = Hr + 2 * Hcx + Hl;
+
+ /* early out */
+ index = ibuf->x * y + x;
+ if (mask[index] != FILTER_MASK_USED) {
+ continue;
+ }
+
+ /* Widen bound. */
+ if (fabsf(deriv_x) > max_num_deriv) {
+ max_num_deriv = fabsf(deriv_x);
+ }
+
+ if (fabsf(deriv_y) > max_num_deriv) {
+ max_num_deriv = fabsf(deriv_y);
+ }
+ }
+ }
+ }
+
+ /* Output derivatives. */
+ auto_range_fit &= (max_num_deriv > 0);
+ for (y = 0; y < ibuf->y; y++) {
+ const int Yu = y == (ibuf->y - 1) ? (ibuf->y - 1) : (y + 1);
+ const int Yc = y;
+ const int Yd = y == 0 ? 0 : (y - 1);
+
+ for (x = 0; x < ibuf->x; x++) {
+ const int Xl = x == 0 ? 0 : (x - 1);
+ const int Xc = x;
+ const int Xr = x == (ibuf->x - 1) ? (ibuf->x - 1) : (x + 1);
+
+ const float Hcy = heights_buffer[Yc * ibuf->x + Xr] - heights_buffer[Yc * ibuf->x + Xl];
+ const float Hu = heights_buffer[Yu * ibuf->x + Xr] - heights_buffer[Yu * ibuf->x + Xl];
+ const float Hd = heights_buffer[Yd * ibuf->x + Xr] - heights_buffer[Yd * ibuf->x + Xl];
+
+ const float Hl = heights_buffer[Yu * ibuf->x + Xl] - heights_buffer[Yd * ibuf->x + Xl];
+ const float Hcx = heights_buffer[Yu * ibuf->x + Xc] - heights_buffer[Yd * ibuf->x + Xc];
+ const float Hr = heights_buffer[Yu * ibuf->x + Xr] - heights_buffer[Yd * ibuf->x + Xr];
+
+ /* This corresponds to using the sobel kernel on the heights buffer
+ * to obtain the derivative multiplied by 8.
+ */
+ float deriv_x = Hu + 2 * Hcy + Hd;
+ float deriv_y = Hr + 2 * Hcx + Hl;
+
+ /* Early out. */
+ index = ibuf->x * y + x;
+ if (mask[index] != FILTER_MASK_USED) {
+ continue;
+ }
+
+ if (auto_range_fit) {
+ deriv_x /= max_num_deriv;
+ deriv_y /= max_num_deriv;
+ }
+ else {
+ deriv_x *= (fmult / denom);
+ deriv_y *= (fmult / denom);
+ }
+
+ deriv_x = deriv_x * 0.5f + 0.5f;
+ deriv_y = deriv_y * 0.5f + 0.5f;
+
+ /* Clamp. */
+ CLAMP(deriv_x, 0.0f, 1.0f);
+ CLAMP(deriv_y, 0.0f, 1.0f);
+
+ /* Write out derivatives. */
+ if (ibuf->rect_float) {
+ float *rrgbf = ibuf->rect_float + index * 4;
+
+ rrgbf[0] = deriv_x;
+ rrgbf[1] = deriv_y;
+ rrgbf[2] = 0.0f;
+ rrgbf[3] = 1.0f;
+ }
+ else {
+ char *rrgb = (char *)ibuf->rect + index * 4;
+
+ rrgb[0] = unit_float_to_uchar_clamp(deriv_x);
+ rrgb[1] = unit_float_to_uchar_clamp(deriv_y);
+ rrgb[2] = 0;
+ rrgb[3] = 255;
+ }
+ }
+ }
+
+ /* Eeturn user-scale (for rendering). */
+ return auto_range_fit ? (max_num_deriv / denom) : (fmult > 0.0f ? (1.0f / fmult) : 0.0f);
+}