Opted to keep includes if they are used indirectly (even if removing is possible).
1496 lines
37 KiB
C
1496 lines
37 KiB
C
/*
|
|
* ***** 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.
|
|
*
|
|
* The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
|
|
* All rights reserved.
|
|
*
|
|
* Contributor(s): 2004-2006, Blender Foundation, full recode
|
|
*
|
|
* ***** END GPL/BL DUAL LICENSE BLOCK *****
|
|
*/
|
|
|
|
/** \file blender/render/intern/source/renderdatabase.c
|
|
* \ingroup render
|
|
*/
|
|
|
|
|
|
/*
|
|
* Storage, retrieval and query of render specific data.
|
|
*
|
|
* All data from a Blender scene is converted by the renderconverter/
|
|
* into a special format that is used by the render module to make
|
|
* images out of. These functions interface to the render-specific
|
|
* database.
|
|
*
|
|
* The blo{ha/ve/vl} arrays store pointers to blocks of 256 data
|
|
* entries each.
|
|
*
|
|
* The index of an entry is >>8 (the highest 24 * bits), to find an
|
|
* offset in a 256-entry block.
|
|
*
|
|
* - If the 256-entry block entry has an entry in the
|
|
* vertnodes/vlaknodes/bloha array of the current block, the i-th entry in
|
|
* that block is allocated to this entry.
|
|
*
|
|
* - If the entry has no block allocated for it yet, memory is
|
|
* allocated.
|
|
*
|
|
* The pointer to the correct entry is returned. Memory is guaranteed
|
|
* to exist (as long as the malloc does not break). Since guarded
|
|
* allocation is used, memory _must_ be available. Otherwise, an
|
|
* exit(0) would occur.
|
|
*
|
|
*/
|
|
|
|
#include <limits.h>
|
|
#include <math.h>
|
|
#include <string.h>
|
|
|
|
#include "MEM_guardedalloc.h"
|
|
|
|
|
|
#include "BLI_math.h"
|
|
#include "BLI_blenlib.h"
|
|
#include "BLI_utildefines.h"
|
|
|
|
#include "DNA_material_types.h"
|
|
#include "DNA_meshdata_types.h"
|
|
#include "DNA_texture_types.h"
|
|
|
|
#include "BKE_customdata.h"
|
|
#include "BKE_DerivedMesh.h"
|
|
|
|
#include "RE_render_ext.h" /* externtex */
|
|
|
|
#include "rayintersection.h"
|
|
#include "rayobject.h"
|
|
#include "renderpipeline.h"
|
|
#include "render_types.h"
|
|
#include "renderdatabase.h"
|
|
#include "texture.h"
|
|
#include "strand.h"
|
|
#include "zbuf.h"
|
|
|
|
/* ------------------------------------------------------------------------- */
|
|
|
|
/* More dynamic allocation of options for render vertices and faces, so we don't
|
|
* have to reserve this space inside vertices.
|
|
* Important; vertices and faces, should have been created already (to get tables
|
|
* checked) that's a reason why the calls demand VertRen/VlakRen * as arg, not
|
|
* the index */
|
|
|
|
/* NOTE! the hardcoded table size 256 is used still in code for going quickly over vertices/faces */
|
|
#define RE_STRESS_ELEMS 1
|
|
#define RE_RAD_ELEMS 4
|
|
#define RE_STRAND_ELEMS 1
|
|
#define RE_TANGENT_ELEMS 3
|
|
#define RE_WINSPEED_ELEMS 4
|
|
#define RE_MTFACE_ELEMS 1
|
|
#define RE_MCOL_ELEMS 4
|
|
#define RE_UV_ELEMS 2
|
|
#define RE_VLAK_ORIGINDEX_ELEMS 1
|
|
#define RE_VERT_ORIGINDEX_ELEMS 1
|
|
#define RE_SURFNOR_ELEMS 3
|
|
#define RE_RADFACE_ELEMS 1
|
|
#define RE_SIMPLIFY_ELEMS 2
|
|
#define RE_FACE_ELEMS 1
|
|
#define RE_NMAP_TANGENT_ELEMS 16
|
|
|
|
float *RE_vertren_get_stress(ObjectRen *obr, VertRen *ver, int verify)
|
|
{
|
|
float *stress;
|
|
int nr= ver->index>>8;
|
|
|
|
stress= obr->vertnodes[nr].stress;
|
|
if (stress==NULL) {
|
|
if (verify)
|
|
stress= obr->vertnodes[nr].stress= MEM_mallocN(256*RE_STRESS_ELEMS*sizeof(float), "stress table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return stress + (ver->index & 255)*RE_STRESS_ELEMS;
|
|
}
|
|
|
|
/* this one callocs! */
|
|
float *RE_vertren_get_rad(ObjectRen *obr, VertRen *ver, int verify)
|
|
{
|
|
float *rad;
|
|
int nr= ver->index>>8;
|
|
|
|
rad= obr->vertnodes[nr].rad;
|
|
if (rad==NULL) {
|
|
if (verify)
|
|
rad= obr->vertnodes[nr].rad= MEM_callocN(256*RE_RAD_ELEMS*sizeof(float), "rad table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return rad + (ver->index & 255)*RE_RAD_ELEMS;
|
|
}
|
|
|
|
float *RE_vertren_get_strand(ObjectRen *obr, VertRen *ver, int verify)
|
|
{
|
|
float *strand;
|
|
int nr= ver->index>>8;
|
|
|
|
strand= obr->vertnodes[nr].strand;
|
|
if (strand==NULL) {
|
|
if (verify)
|
|
strand= obr->vertnodes[nr].strand= MEM_mallocN(256*RE_STRAND_ELEMS*sizeof(float), "strand table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return strand + (ver->index & 255)*RE_STRAND_ELEMS;
|
|
}
|
|
|
|
/* needs calloc */
|
|
float *RE_vertren_get_tangent(ObjectRen *obr, VertRen *ver, int verify)
|
|
{
|
|
float *tangent;
|
|
int nr= ver->index>>8;
|
|
|
|
tangent= obr->vertnodes[nr].tangent;
|
|
if (tangent==NULL) {
|
|
if (verify)
|
|
tangent= obr->vertnodes[nr].tangent= MEM_callocN(256*RE_TANGENT_ELEMS*sizeof(float), "tangent table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return tangent + (ver->index & 255)*RE_TANGENT_ELEMS;
|
|
}
|
|
|
|
/* needs calloc! not all renderverts have them */
|
|
/* also winspeed is exception, it is stored per instance */
|
|
float *RE_vertren_get_winspeed(ObjectInstanceRen *obi, VertRen *ver, int verify)
|
|
{
|
|
float *winspeed;
|
|
int totvector;
|
|
|
|
winspeed= obi->vectors;
|
|
if (winspeed==NULL) {
|
|
if (verify) {
|
|
totvector= obi->obr->totvert + obi->obr->totstrand;
|
|
winspeed= obi->vectors= MEM_callocN(totvector*RE_WINSPEED_ELEMS*sizeof(float), "winspeed table");
|
|
}
|
|
else
|
|
return NULL;
|
|
}
|
|
return winspeed + ver->index*RE_WINSPEED_ELEMS;
|
|
}
|
|
|
|
int *RE_vertren_get_origindex(ObjectRen *obr, VertRen *ver, int verify)
|
|
{
|
|
int *origindex;
|
|
int nr= ver->index>>8;
|
|
|
|
origindex= obr->vertnodes[nr].origindex;
|
|
if (origindex==NULL) {
|
|
if (verify)
|
|
origindex= obr->vertnodes[nr].origindex= MEM_mallocN(256*RE_VERT_ORIGINDEX_ELEMS*sizeof(int), "origindex table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return origindex + (ver->index & 255)*RE_VERT_ORIGINDEX_ELEMS;
|
|
}
|
|
|
|
VertRen *RE_vertren_copy(ObjectRen *obr, VertRen *ver)
|
|
{
|
|
VertRen *v1= RE_findOrAddVert(obr, obr->totvert++);
|
|
float *fp1, *fp2;
|
|
int *int1, *int2;
|
|
int index= v1->index;
|
|
|
|
*v1= *ver;
|
|
v1->index= index;
|
|
|
|
fp1= RE_vertren_get_stress(obr, ver, 0);
|
|
if (fp1) {
|
|
fp2= RE_vertren_get_stress(obr, v1, 1);
|
|
memcpy(fp2, fp1, RE_STRESS_ELEMS*sizeof(float));
|
|
}
|
|
fp1= RE_vertren_get_rad(obr, ver, 0);
|
|
if (fp1) {
|
|
fp2= RE_vertren_get_rad(obr, v1, 1);
|
|
memcpy(fp2, fp1, RE_RAD_ELEMS*sizeof(float));
|
|
}
|
|
fp1= RE_vertren_get_strand(obr, ver, 0);
|
|
if (fp1) {
|
|
fp2= RE_vertren_get_strand(obr, v1, 1);
|
|
memcpy(fp2, fp1, RE_STRAND_ELEMS*sizeof(float));
|
|
}
|
|
fp1= RE_vertren_get_tangent(obr, ver, 0);
|
|
if (fp1) {
|
|
fp2= RE_vertren_get_tangent(obr, v1, 1);
|
|
memcpy(fp2, fp1, RE_TANGENT_ELEMS*sizeof(float));
|
|
}
|
|
int1= RE_vertren_get_origindex(obr, ver, 0);
|
|
if (int1) {
|
|
int2= RE_vertren_get_origindex(obr, v1, 1);
|
|
memcpy(int2, int1, RE_VERT_ORIGINDEX_ELEMS*sizeof(int));
|
|
}
|
|
return v1;
|
|
}
|
|
|
|
VertRen *RE_findOrAddVert(ObjectRen *obr, int nr)
|
|
{
|
|
VertTableNode *temp;
|
|
VertRen *v;
|
|
int a;
|
|
|
|
if (nr<0) {
|
|
printf("error in findOrAddVert: %d\n", nr);
|
|
return NULL;
|
|
}
|
|
a= nr>>8;
|
|
|
|
if (a>=obr->vertnodeslen-1) { /* Need to allocate more columns..., and keep last element NULL for free loop */
|
|
temp= obr->vertnodes;
|
|
|
|
obr->vertnodes= MEM_mallocN(sizeof(VertTableNode)*(obr->vertnodeslen+TABLEINITSIZE), "vertnodes");
|
|
if (temp) memcpy(obr->vertnodes, temp, obr->vertnodeslen*sizeof(VertTableNode));
|
|
memset(obr->vertnodes+obr->vertnodeslen, 0, TABLEINITSIZE*sizeof(VertTableNode));
|
|
|
|
obr->vertnodeslen+=TABLEINITSIZE;
|
|
if (temp) MEM_freeN(temp);
|
|
}
|
|
|
|
v= obr->vertnodes[a].vert;
|
|
if (v==NULL) {
|
|
int i;
|
|
|
|
v= (VertRen *)MEM_callocN(256*sizeof(VertRen), "findOrAddVert");
|
|
obr->vertnodes[a].vert= v;
|
|
|
|
for (i= (nr & 0xFFFFFF00), a=0; a<256; a++, i++) {
|
|
v[a].index= i;
|
|
}
|
|
}
|
|
v+= (nr & 255);
|
|
return v;
|
|
}
|
|
|
|
/* ------------------------------------------------------------------------ */
|
|
|
|
MTFace *RE_vlakren_get_tface(ObjectRen *obr, VlakRen *vlr, int n, char **name, int verify)
|
|
{
|
|
VlakTableNode *node;
|
|
int nr= vlr->index>>8, vlakindex= (vlr->index&255);
|
|
int index= (n<<8) + vlakindex;
|
|
|
|
node= &obr->vlaknodes[nr];
|
|
|
|
if (verify) {
|
|
if (n>=node->totmtface) {
|
|
MTFace *mtface= node->mtface;
|
|
int size= (n+1)*256;
|
|
|
|
node->mtface= MEM_callocN(size*sizeof(MTFace), "Vlak mtface");
|
|
|
|
if (mtface) {
|
|
size= node->totmtface*256;
|
|
memcpy(node->mtface, mtface, size*sizeof(MTFace));
|
|
MEM_freeN(mtface);
|
|
}
|
|
|
|
node->totmtface= n+1;
|
|
}
|
|
}
|
|
else {
|
|
if (n>=node->totmtface)
|
|
return NULL;
|
|
|
|
if (name) *name= obr->mtface[n];
|
|
}
|
|
|
|
return node->mtface + index;
|
|
}
|
|
|
|
MCol *RE_vlakren_get_mcol(ObjectRen *obr, VlakRen *vlr, int n, char **name, int verify)
|
|
{
|
|
VlakTableNode *node;
|
|
int nr= vlr->index>>8, vlakindex= (vlr->index&255);
|
|
int index= (n<<8) + vlakindex;
|
|
|
|
node= &obr->vlaknodes[nr];
|
|
|
|
if (verify) {
|
|
if (n>=node->totmcol) {
|
|
MCol *mcol= node->mcol;
|
|
int size= (n+1)*256;
|
|
|
|
node->mcol= MEM_callocN(size*sizeof(MCol)*RE_MCOL_ELEMS, "Vlak mcol");
|
|
|
|
if (mcol) {
|
|
size= node->totmcol*256;
|
|
memcpy(node->mcol, mcol, size*sizeof(MCol)*RE_MCOL_ELEMS);
|
|
MEM_freeN(mcol);
|
|
}
|
|
|
|
node->totmcol= n+1;
|
|
}
|
|
}
|
|
else {
|
|
if (n>=node->totmcol)
|
|
return NULL;
|
|
|
|
if (name) *name= obr->mcol[n];
|
|
}
|
|
|
|
return node->mcol + index*RE_MCOL_ELEMS;
|
|
}
|
|
|
|
int *RE_vlakren_get_origindex(ObjectRen *obr, VlakRen *vlak, int verify)
|
|
{
|
|
int *origindex;
|
|
int nr= vlak->index>>8;
|
|
|
|
origindex= obr->vlaknodes[nr].origindex;
|
|
if (origindex==NULL) {
|
|
if (verify)
|
|
origindex= obr->vlaknodes[nr].origindex= MEM_callocN(256*RE_VLAK_ORIGINDEX_ELEMS*sizeof(int), "origindex table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return origindex + (vlak->index & 255)*RE_VLAK_ORIGINDEX_ELEMS;
|
|
}
|
|
|
|
float *RE_vlakren_get_surfnor(ObjectRen *obr, VlakRen *vlak, int verify)
|
|
{
|
|
float *surfnor;
|
|
int nr= vlak->index>>8;
|
|
|
|
surfnor= obr->vlaknodes[nr].surfnor;
|
|
if (surfnor==NULL) {
|
|
if (verify)
|
|
surfnor= obr->vlaknodes[nr].surfnor= MEM_callocN(256*RE_SURFNOR_ELEMS*sizeof(float), "surfnor table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return surfnor + (vlak->index & 255)*RE_SURFNOR_ELEMS;
|
|
}
|
|
|
|
float *RE_vlakren_get_nmap_tangent(ObjectRen *obr, VlakRen *vlak, int verify)
|
|
{
|
|
float *tangent;
|
|
int nr= vlak->index>>8;
|
|
|
|
tangent= obr->vlaknodes[nr].tangent;
|
|
if (tangent==NULL) {
|
|
if (verify)
|
|
tangent= obr->vlaknodes[nr].tangent= MEM_callocN(256*RE_NMAP_TANGENT_ELEMS*sizeof(float), "tangent table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return tangent + (vlak->index & 255)*RE_NMAP_TANGENT_ELEMS;
|
|
}
|
|
|
|
RadFace **RE_vlakren_get_radface(ObjectRen *obr, VlakRen *vlak, int verify)
|
|
{
|
|
RadFace **radface;
|
|
int nr= vlak->index>>8;
|
|
|
|
radface= obr->vlaknodes[nr].radface;
|
|
if (radface==NULL) {
|
|
if (verify)
|
|
radface = obr->vlaknodes[nr].radface= MEM_callocN(256 * RE_RADFACE_ELEMS * sizeof(void *), "radface table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return radface + (vlak->index & 255)*RE_RADFACE_ELEMS;
|
|
}
|
|
|
|
VlakRen *RE_vlakren_copy(ObjectRen *obr, VlakRen *vlr)
|
|
{
|
|
VlakRen *vlr1 = RE_findOrAddVlak(obr, obr->totvlak++);
|
|
MTFace *mtface, *mtface1;
|
|
MCol *mcol, *mcol1;
|
|
float *surfnor, *surfnor1, *tangent, *tangent1;
|
|
int *origindex, *origindex1;
|
|
RadFace **radface, **radface1;
|
|
int i, index = vlr1->index;
|
|
char *name;
|
|
|
|
*vlr1= *vlr;
|
|
vlr1->index= index;
|
|
|
|
for (i=0; (mtface=RE_vlakren_get_tface(obr, vlr, i, &name, 0)) != NULL; i++) {
|
|
mtface1= RE_vlakren_get_tface(obr, vlr1, i, &name, 1);
|
|
memcpy(mtface1, mtface, sizeof(MTFace)*RE_MTFACE_ELEMS);
|
|
}
|
|
|
|
for (i=0; (mcol=RE_vlakren_get_mcol(obr, vlr, i, &name, 0)) != NULL; i++) {
|
|
mcol1= RE_vlakren_get_mcol(obr, vlr1, i, &name, 1);
|
|
memcpy(mcol1, mcol, sizeof(MCol)*RE_MCOL_ELEMS);
|
|
}
|
|
|
|
origindex= RE_vlakren_get_origindex(obr, vlr, 0);
|
|
if (origindex) {
|
|
origindex1= RE_vlakren_get_origindex(obr, vlr1, 1);
|
|
/* Just an int, but memcpy for consistency. */
|
|
memcpy(origindex1, origindex, sizeof(int)*RE_VLAK_ORIGINDEX_ELEMS);
|
|
}
|
|
|
|
surfnor= RE_vlakren_get_surfnor(obr, vlr, 0);
|
|
if (surfnor) {
|
|
surfnor1= RE_vlakren_get_surfnor(obr, vlr1, 1);
|
|
copy_v3_v3(surfnor1, surfnor);
|
|
}
|
|
|
|
tangent= RE_vlakren_get_nmap_tangent(obr, vlr, 0);
|
|
if (tangent) {
|
|
tangent1= RE_vlakren_get_nmap_tangent(obr, vlr1, 1);
|
|
memcpy(tangent1, tangent, sizeof(float)*RE_NMAP_TANGENT_ELEMS);
|
|
}
|
|
|
|
radface= RE_vlakren_get_radface(obr, vlr, 0);
|
|
if (radface) {
|
|
radface1= RE_vlakren_get_radface(obr, vlr1, 1);
|
|
*radface1= *radface;
|
|
}
|
|
|
|
return vlr1;
|
|
}
|
|
|
|
void RE_vlakren_get_normal(Render *UNUSED(re), ObjectInstanceRen *obi, VlakRen *vlr, float r_nor[3])
|
|
{
|
|
float (*nmat)[3]= obi->nmat;
|
|
|
|
if (obi->flag & R_TRANSFORMED) {
|
|
mul_v3_m3v3(r_nor, nmat, vlr->n);
|
|
normalize_v3(r_nor);
|
|
}
|
|
else {
|
|
copy_v3_v3(r_nor, vlr->n);
|
|
}
|
|
}
|
|
|
|
void RE_set_customdata_names(ObjectRen *obr, CustomData *data)
|
|
{
|
|
/* CustomData layer names are stored per object here, because the
|
|
* DerivedMesh which stores the layers is freed */
|
|
|
|
CustomDataLayer *layer;
|
|
int numtf = 0, numcol = 0, i, mtfn, mcn;
|
|
|
|
if (CustomData_has_layer(data, CD_MTFACE)) {
|
|
numtf= CustomData_number_of_layers(data, CD_MTFACE);
|
|
obr->mtface= MEM_callocN(sizeof(*obr->mtface)*numtf, "mtfacenames");
|
|
}
|
|
|
|
if (CustomData_has_layer(data, CD_MCOL)) {
|
|
numcol= CustomData_number_of_layers(data, CD_MCOL);
|
|
obr->mcol= MEM_callocN(sizeof(*obr->mcol)*numcol, "mcolnames");
|
|
}
|
|
|
|
for (i=0, mtfn=0, mcn=0; i < data->totlayer; i++) {
|
|
layer= &data->layers[i];
|
|
|
|
if (layer->type == CD_MTFACE) {
|
|
BLI_strncpy(obr->mtface[mtfn++], layer->name, sizeof(layer->name));
|
|
obr->actmtface= CLAMPIS(layer->active_rnd, 0, numtf);
|
|
obr->bakemtface= layer->active;
|
|
}
|
|
else if (layer->type == CD_MCOL) {
|
|
BLI_strncpy(obr->mcol[mcn++], layer->name, sizeof(layer->name));
|
|
obr->actmcol= CLAMPIS(layer->active_rnd, 0, numcol);
|
|
}
|
|
}
|
|
}
|
|
|
|
VlakRen *RE_findOrAddVlak(ObjectRen *obr, int nr)
|
|
{
|
|
VlakTableNode *temp;
|
|
VlakRen *v;
|
|
int a;
|
|
|
|
if (nr<0) {
|
|
printf("error in findOrAddVlak: %d\n", nr);
|
|
return obr->vlaknodes[0].vlak;
|
|
}
|
|
a= nr>>8;
|
|
|
|
if (a>=obr->vlaknodeslen-1) { /* Need to allocate more columns..., and keep last element NULL for free loop */
|
|
temp= obr->vlaknodes;
|
|
|
|
obr->vlaknodes= MEM_mallocN(sizeof(VlakTableNode)*(obr->vlaknodeslen+TABLEINITSIZE), "vlaknodes");
|
|
if (temp) memcpy(obr->vlaknodes, temp, obr->vlaknodeslen*sizeof(VlakTableNode));
|
|
memset(obr->vlaknodes+obr->vlaknodeslen, 0, TABLEINITSIZE*sizeof(VlakTableNode));
|
|
|
|
obr->vlaknodeslen+=TABLEINITSIZE; /*Does this really need to be power of 2?*/
|
|
if (temp) MEM_freeN(temp);
|
|
}
|
|
|
|
v= obr->vlaknodes[a].vlak;
|
|
|
|
if (v==NULL) {
|
|
int i;
|
|
|
|
v= (VlakRen *)MEM_callocN(256*sizeof(VlakRen), "findOrAddVlak");
|
|
obr->vlaknodes[a].vlak= v;
|
|
|
|
for (i= (nr & 0xFFFFFF00), a=0; a<256; a++, i++)
|
|
v[a].index= i;
|
|
}
|
|
v+= (nr & 255);
|
|
return v;
|
|
}
|
|
|
|
/* ------------------------------------------------------------------------ */
|
|
|
|
float *RE_strandren_get_surfnor(ObjectRen *obr, StrandRen *strand, int verify)
|
|
{
|
|
float *surfnor;
|
|
int nr= strand->index>>8;
|
|
|
|
surfnor= obr->strandnodes[nr].surfnor;
|
|
if (surfnor==NULL) {
|
|
if (verify)
|
|
surfnor= obr->strandnodes[nr].surfnor= MEM_callocN(256*RE_SURFNOR_ELEMS*sizeof(float), "surfnor strand table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return surfnor + (strand->index & 255)*RE_SURFNOR_ELEMS;
|
|
}
|
|
|
|
float *RE_strandren_get_uv(ObjectRen *obr, StrandRen *strand, int n, char **name, int verify)
|
|
{
|
|
StrandTableNode *node;
|
|
int nr= strand->index>>8, strandindex= (strand->index&255);
|
|
int index= (n<<8) + strandindex;
|
|
|
|
node= &obr->strandnodes[nr];
|
|
|
|
if (verify) {
|
|
if (n>=node->totuv) {
|
|
float *uv= node->uv;
|
|
int size= (n+1)*256;
|
|
|
|
node->uv= MEM_callocN(size*sizeof(float)*RE_UV_ELEMS, "strand uv table");
|
|
|
|
if (uv) {
|
|
size= node->totuv*256;
|
|
memcpy(node->uv, uv, size*sizeof(float)*RE_UV_ELEMS);
|
|
MEM_freeN(uv);
|
|
}
|
|
|
|
node->totuv= n+1;
|
|
}
|
|
}
|
|
else {
|
|
if (n>=node->totuv)
|
|
return NULL;
|
|
|
|
if (name) *name= obr->mtface[n];
|
|
}
|
|
|
|
return node->uv + index*RE_UV_ELEMS;
|
|
}
|
|
|
|
MCol *RE_strandren_get_mcol(ObjectRen *obr, StrandRen *strand, int n, char **name, int verify)
|
|
{
|
|
StrandTableNode *node;
|
|
int nr= strand->index>>8, strandindex= (strand->index&255);
|
|
int index= (n<<8) + strandindex;
|
|
|
|
node= &obr->strandnodes[nr];
|
|
|
|
if (verify) {
|
|
if (n>=node->totmcol) {
|
|
MCol *mcol= node->mcol;
|
|
int size= (n+1)*256;
|
|
|
|
node->mcol= MEM_callocN(size*sizeof(MCol)*RE_MCOL_ELEMS, "strand mcol table");
|
|
|
|
if (mcol) {
|
|
size= node->totmcol*256;
|
|
memcpy(node->mcol, mcol, size*sizeof(MCol)*RE_MCOL_ELEMS);
|
|
MEM_freeN(mcol);
|
|
}
|
|
|
|
node->totmcol= n+1;
|
|
}
|
|
}
|
|
else {
|
|
if (n>=node->totmcol)
|
|
return NULL;
|
|
|
|
if (name) *name= obr->mcol[n];
|
|
}
|
|
|
|
return node->mcol + index*RE_MCOL_ELEMS;
|
|
}
|
|
|
|
float *RE_strandren_get_simplify(struct ObjectRen *obr, struct StrandRen *strand, int verify)
|
|
{
|
|
float *simplify;
|
|
int nr= strand->index>>8;
|
|
|
|
simplify= obr->strandnodes[nr].simplify;
|
|
if (simplify==NULL) {
|
|
if (verify)
|
|
simplify= obr->strandnodes[nr].simplify= MEM_callocN(256*RE_SIMPLIFY_ELEMS*sizeof(float), "simplify strand table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return simplify + (strand->index & 255)*RE_SIMPLIFY_ELEMS;
|
|
}
|
|
|
|
int *RE_strandren_get_face(ObjectRen *obr, StrandRen *strand, int verify)
|
|
{
|
|
int *face;
|
|
int nr= strand->index>>8;
|
|
|
|
face= obr->strandnodes[nr].face;
|
|
if (face==NULL) {
|
|
if (verify)
|
|
face= obr->strandnodes[nr].face= MEM_callocN(256*RE_FACE_ELEMS*sizeof(int), "face strand table");
|
|
else
|
|
return NULL;
|
|
}
|
|
return face + (strand->index & 255)*RE_FACE_ELEMS;
|
|
}
|
|
|
|
/* winspeed is exception, it is stored per instance */
|
|
float *RE_strandren_get_winspeed(ObjectInstanceRen *obi, StrandRen *strand, int verify)
|
|
{
|
|
float *winspeed;
|
|
int totvector;
|
|
|
|
winspeed= obi->vectors;
|
|
if (winspeed==NULL) {
|
|
if (verify) {
|
|
totvector= obi->obr->totvert + obi->obr->totstrand;
|
|
winspeed= obi->vectors= MEM_callocN(totvector*RE_WINSPEED_ELEMS*sizeof(float), "winspeed strand table");
|
|
}
|
|
else
|
|
return NULL;
|
|
}
|
|
return winspeed + (obi->obr->totvert + strand->index)*RE_WINSPEED_ELEMS;
|
|
}
|
|
|
|
StrandRen *RE_findOrAddStrand(ObjectRen *obr, int nr)
|
|
{
|
|
StrandTableNode *temp;
|
|
StrandRen *v;
|
|
int a;
|
|
|
|
if (nr<0) {
|
|
printf("error in findOrAddStrand: %d\n", nr);
|
|
return obr->strandnodes[0].strand;
|
|
}
|
|
a= nr>>8;
|
|
|
|
if (a>=obr->strandnodeslen-1) { /* Need to allocate more columns..., and keep last element NULL for free loop */
|
|
temp= obr->strandnodes;
|
|
|
|
obr->strandnodes= MEM_mallocN(sizeof(StrandTableNode)*(obr->strandnodeslen+TABLEINITSIZE), "strandnodes");
|
|
if (temp) memcpy(obr->strandnodes, temp, obr->strandnodeslen*sizeof(StrandTableNode));
|
|
memset(obr->strandnodes+obr->strandnodeslen, 0, TABLEINITSIZE*sizeof(StrandTableNode));
|
|
|
|
obr->strandnodeslen+=TABLEINITSIZE; /*Does this really need to be power of 2?*/
|
|
if (temp) MEM_freeN(temp);
|
|
}
|
|
|
|
v= obr->strandnodes[a].strand;
|
|
|
|
if (v==NULL) {
|
|
int i;
|
|
|
|
v= (StrandRen *)MEM_callocN(256*sizeof(StrandRen), "findOrAddStrand");
|
|
obr->strandnodes[a].strand= v;
|
|
|
|
for (i= (nr & 0xFFFFFF00), a=0; a<256; a++, i++)
|
|
v[a].index= i;
|
|
}
|
|
v+= (nr & 255);
|
|
return v;
|
|
}
|
|
|
|
StrandBuffer *RE_addStrandBuffer(ObjectRen *obr, int totvert)
|
|
{
|
|
StrandBuffer *strandbuf;
|
|
|
|
strandbuf= MEM_callocN(sizeof(StrandBuffer), "StrandBuffer");
|
|
strandbuf->vert= MEM_callocN(sizeof(StrandVert)*totvert, "StrandVert");
|
|
strandbuf->totvert= totvert;
|
|
strandbuf->obr= obr;
|
|
|
|
obr->strandbuf= strandbuf;
|
|
|
|
return strandbuf;
|
|
}
|
|
|
|
/* ------------------------------------------------------------------------ */
|
|
|
|
ObjectRen *RE_addRenderObject(Render *re, Object *ob, Object *par, int index, int psysindex, int lay)
|
|
{
|
|
ObjectRen *obr= MEM_callocN(sizeof(ObjectRen), "object render struct");
|
|
|
|
BLI_addtail(&re->objecttable, obr);
|
|
obr->ob= ob;
|
|
obr->par= par;
|
|
obr->index= index;
|
|
obr->psysindex= psysindex;
|
|
obr->lay= lay;
|
|
|
|
return obr;
|
|
}
|
|
|
|
void free_renderdata_vertnodes(VertTableNode *vertnodes)
|
|
{
|
|
int a;
|
|
|
|
if (vertnodes==NULL) return;
|
|
|
|
for (a=0; vertnodes[a].vert; a++) {
|
|
MEM_freeN(vertnodes[a].vert);
|
|
|
|
if (vertnodes[a].rad)
|
|
MEM_freeN(vertnodes[a].rad);
|
|
if (vertnodes[a].strand)
|
|
MEM_freeN(vertnodes[a].strand);
|
|
if (vertnodes[a].tangent)
|
|
MEM_freeN(vertnodes[a].tangent);
|
|
if (vertnodes[a].stress)
|
|
MEM_freeN(vertnodes[a].stress);
|
|
if (vertnodes[a].winspeed)
|
|
MEM_freeN(vertnodes[a].winspeed);
|
|
if (vertnodes[a].origindex)
|
|
MEM_freeN(vertnodes[a].origindex);
|
|
}
|
|
|
|
MEM_freeN(vertnodes);
|
|
}
|
|
|
|
void free_renderdata_vlaknodes(VlakTableNode *vlaknodes)
|
|
{
|
|
int a;
|
|
|
|
if (vlaknodes==NULL) return;
|
|
|
|
for (a=0; vlaknodes[a].vlak; a++) {
|
|
MEM_freeN(vlaknodes[a].vlak);
|
|
|
|
if (vlaknodes[a].mtface)
|
|
MEM_freeN(vlaknodes[a].mtface);
|
|
if (vlaknodes[a].mcol)
|
|
MEM_freeN(vlaknodes[a].mcol);
|
|
if (vlaknodes[a].origindex)
|
|
MEM_freeN(vlaknodes[a].origindex);
|
|
if (vlaknodes[a].surfnor)
|
|
MEM_freeN(vlaknodes[a].surfnor);
|
|
if (vlaknodes[a].tangent)
|
|
MEM_freeN(vlaknodes[a].tangent);
|
|
if (vlaknodes[a].radface)
|
|
MEM_freeN(vlaknodes[a].radface);
|
|
}
|
|
|
|
MEM_freeN(vlaknodes);
|
|
}
|
|
|
|
static void free_renderdata_strandnodes(StrandTableNode *strandnodes)
|
|
{
|
|
int a;
|
|
|
|
if (strandnodes==NULL) return;
|
|
|
|
for (a=0; strandnodes[a].strand; a++) {
|
|
MEM_freeN(strandnodes[a].strand);
|
|
|
|
if (strandnodes[a].uv)
|
|
MEM_freeN(strandnodes[a].uv);
|
|
if (strandnodes[a].mcol)
|
|
MEM_freeN(strandnodes[a].mcol);
|
|
if (strandnodes[a].winspeed)
|
|
MEM_freeN(strandnodes[a].winspeed);
|
|
if (strandnodes[a].surfnor)
|
|
MEM_freeN(strandnodes[a].surfnor);
|
|
if (strandnodes[a].simplify)
|
|
MEM_freeN(strandnodes[a].simplify);
|
|
if (strandnodes[a].face)
|
|
MEM_freeN(strandnodes[a].face);
|
|
}
|
|
|
|
MEM_freeN(strandnodes);
|
|
}
|
|
|
|
void free_renderdata_tables(Render *re)
|
|
{
|
|
ObjectInstanceRen *obi;
|
|
ObjectRen *obr;
|
|
StrandBuffer *strandbuf;
|
|
int a=0;
|
|
|
|
for (obr=re->objecttable.first; obr; obr=obr->next) {
|
|
if (obr->vertnodes) {
|
|
free_renderdata_vertnodes(obr->vertnodes);
|
|
obr->vertnodes= NULL;
|
|
obr->vertnodeslen= 0;
|
|
}
|
|
|
|
if (obr->vlaknodes) {
|
|
free_renderdata_vlaknodes(obr->vlaknodes);
|
|
obr->vlaknodes= NULL;
|
|
obr->vlaknodeslen= 0;
|
|
obr->totvlak= 0;
|
|
}
|
|
|
|
if (obr->bloha) {
|
|
for (a=0; obr->bloha[a]; a++)
|
|
MEM_freeN(obr->bloha[a]);
|
|
|
|
MEM_freeN(obr->bloha);
|
|
obr->bloha= NULL;
|
|
obr->blohalen= 0;
|
|
}
|
|
|
|
if (obr->strandnodes) {
|
|
free_renderdata_strandnodes(obr->strandnodes);
|
|
obr->strandnodes= NULL;
|
|
obr->strandnodeslen= 0;
|
|
}
|
|
|
|
strandbuf= obr->strandbuf;
|
|
if (strandbuf) {
|
|
if (strandbuf->vert) MEM_freeN(strandbuf->vert);
|
|
if (strandbuf->bound) MEM_freeN(strandbuf->bound);
|
|
MEM_freeN(strandbuf);
|
|
}
|
|
|
|
if (obr->mtface)
|
|
MEM_freeN(obr->mtface);
|
|
|
|
if (obr->mcol)
|
|
MEM_freeN(obr->mcol);
|
|
|
|
if (obr->rayfaces) {
|
|
MEM_freeN(obr->rayfaces);
|
|
obr->rayfaces = NULL;
|
|
}
|
|
|
|
if (obr->rayprimitives) {
|
|
MEM_freeN(obr->rayprimitives);
|
|
obr->rayprimitives = NULL;
|
|
}
|
|
|
|
if (obr->raytree) {
|
|
RE_rayobject_free(obr->raytree);
|
|
obr->raytree = NULL;
|
|
}
|
|
}
|
|
|
|
if (re->objectinstance) {
|
|
for (obi=re->instancetable.first; obi; obi=obi->next) {
|
|
if (obi->vectors)
|
|
MEM_freeN(obi->vectors);
|
|
|
|
if (obi->raytree)
|
|
RE_rayobject_free(obi->raytree);
|
|
}
|
|
|
|
MEM_freeN(re->objectinstance);
|
|
re->objectinstance= NULL;
|
|
re->totinstance= 0;
|
|
re->instancetable.first= re->instancetable.last= NULL;
|
|
}
|
|
|
|
if (re->sortedhalos) {
|
|
MEM_freeN(re->sortedhalos);
|
|
re->sortedhalos= NULL;
|
|
}
|
|
|
|
BLI_freelistN(&re->customdata_names);
|
|
BLI_freelistN(&re->objecttable);
|
|
BLI_freelistN(&re->instancetable);
|
|
}
|
|
|
|
/* ------------------------------------------------------------------------ */
|
|
|
|
HaloRen *RE_findOrAddHalo(ObjectRen *obr, int nr)
|
|
{
|
|
HaloRen *h, **temp;
|
|
int a;
|
|
|
|
if (nr<0) {
|
|
printf("error in findOrAddHalo: %d\n", nr);
|
|
return NULL;
|
|
}
|
|
a= nr>>8;
|
|
|
|
if (a>=obr->blohalen-1) { /* Need to allocate more columns..., and keep last element NULL for free loop */
|
|
//printf("Allocating %i more halo groups. %i total.\n",
|
|
// TABLEINITSIZE, obr->blohalen+TABLEINITSIZE );
|
|
temp=obr->bloha;
|
|
|
|
obr->bloha = (HaloRen **)MEM_callocN(sizeof(void *) * (obr->blohalen + TABLEINITSIZE), "Bloha");
|
|
if (temp) memcpy(obr->bloha, temp, obr->blohalen*sizeof(void *));
|
|
memset(&(obr->bloha[obr->blohalen]), 0, TABLEINITSIZE * sizeof(void *));
|
|
obr->blohalen+=TABLEINITSIZE; /*Does this really need to be power of 2?*/
|
|
if (temp) MEM_freeN(temp);
|
|
}
|
|
|
|
h= obr->bloha[a];
|
|
if (h==NULL) {
|
|
h= (HaloRen *)MEM_callocN(256*sizeof(HaloRen), "findOrAdHalo");
|
|
obr->bloha[a]= h;
|
|
}
|
|
h+= (nr & 255);
|
|
return h;
|
|
}
|
|
|
|
/* ------------------------------------------------------------------------- */
|
|
|
|
HaloRen *RE_inithalo(Render *re, ObjectRen *obr, Material *ma,
|
|
const float vec[3], const float vec1[3],
|
|
const float *orco, float hasize, float vectsize, int seed)
|
|
{
|
|
HaloRen *har;
|
|
MTex *mtex;
|
|
float tin, tr, tg, tb, ta;
|
|
float xn, yn, zn, texvec[3], hoco[4], hoco1[4];
|
|
|
|
if (hasize==0.0f) return NULL;
|
|
|
|
projectverto(vec, re->winmat, hoco);
|
|
if (hoco[3]==0.0f) return NULL;
|
|
if (vec1) {
|
|
projectverto(vec1, re->winmat, hoco1);
|
|
if (hoco1[3]==0.0f) return NULL;
|
|
}
|
|
|
|
har= RE_findOrAddHalo(obr, obr->tothalo++);
|
|
copy_v3_v3(har->co, vec);
|
|
har->hasize= hasize;
|
|
|
|
/* actual projectvert is done in function project_renderdata() because of parts/border/pano */
|
|
/* we do it here for sorting of halos */
|
|
zn= hoco[3];
|
|
har->xs= 0.5f*re->winx*(hoco[0]/zn);
|
|
har->ys= 0.5f*re->winy*(hoco[1]/zn);
|
|
har->zs= 0x7FFFFF*(hoco[2]/zn);
|
|
|
|
har->zBufDist = 0x7FFFFFFF*(hoco[2]/zn);
|
|
|
|
/* halovect */
|
|
if (vec1) {
|
|
|
|
har->type |= HA_VECT;
|
|
|
|
xn= har->xs - 0.5f*re->winx*(hoco1[0]/hoco1[3]);
|
|
yn= har->ys - 0.5f*re->winy*(hoco1[1]/hoco1[3]);
|
|
if (xn==0.0f || (xn==0.0f && yn==0.0f)) zn= 0.0f;
|
|
else zn= atan2(yn, xn);
|
|
|
|
har->sin= sin(zn);
|
|
har->cos= cos(zn);
|
|
zn= len_v3v3(vec1, vec);
|
|
|
|
har->hasize= vectsize*zn + (1.0f-vectsize)*hasize;
|
|
|
|
sub_v3_v3v3(har->no, vec, vec1);
|
|
normalize_v3(har->no);
|
|
}
|
|
|
|
if (ma->mode & MA_HALO_XALPHA) har->type |= HA_XALPHA;
|
|
|
|
har->alfa= ma->alpha;
|
|
har->r= ma->r;
|
|
har->g= ma->g;
|
|
har->b= ma->b;
|
|
har->add= (255.0f*ma->add);
|
|
har->mat= ma;
|
|
har->hard= ma->har;
|
|
har->seed= seed % 256;
|
|
|
|
if (ma->mode & MA_STAR) har->starpoints= ma->starc;
|
|
if (ma->mode & MA_HALO_LINES) har->linec= ma->linec;
|
|
if (ma->mode & MA_HALO_RINGS) har->ringc= ma->ringc;
|
|
if (ma->mode & MA_HALO_FLARE) har->flarec= ma->flarec;
|
|
|
|
|
|
if (ma->mtex[0]) {
|
|
|
|
if (ma->mode & MA_HALOTEX) {
|
|
har->tex = 1;
|
|
}
|
|
else if (har->mat->septex & (1 << 0)) {
|
|
/* only 1 level textures */
|
|
}
|
|
else {
|
|
mtex= ma->mtex[0];
|
|
copy_v3_v3(texvec, vec);
|
|
|
|
if (mtex->texco & TEXCO_NORM) {
|
|
;
|
|
}
|
|
else if (mtex->texco & TEXCO_OBJECT) {
|
|
/* texvec[0]+= imatbase->ivec[0]; */
|
|
/* texvec[1]+= imatbase->ivec[1]; */
|
|
/* texvec[2]+= imatbase->ivec[2]; */
|
|
/* mul_m3_v3(imatbase->imat, texvec); */
|
|
}
|
|
else {
|
|
if (orco) {
|
|
copy_v3_v3(texvec, orco);
|
|
}
|
|
}
|
|
|
|
externtex(mtex, texvec, &tin, &tr, &tg, &tb, &ta, 0, re->pool);
|
|
|
|
yn= tin*mtex->colfac;
|
|
//zn= tin*mtex->alphafac;
|
|
|
|
if (mtex->mapto & MAP_COL) {
|
|
zn= 1.0f-yn;
|
|
har->r= (yn*tr+ zn*ma->r);
|
|
har->g= (yn*tg+ zn*ma->g);
|
|
har->b= (yn*tb+ zn*ma->b);
|
|
}
|
|
if (mtex->texco & TEXCO_UV) {
|
|
har->alfa= tin;
|
|
}
|
|
if (mtex->mapto & MAP_ALPHA)
|
|
har->alfa= tin;
|
|
}
|
|
}
|
|
|
|
har->pool = re->pool;
|
|
|
|
return har;
|
|
}
|
|
|
|
HaloRen *RE_inithalo_particle(Render *re, ObjectRen *obr, DerivedMesh *dm, Material *ma,
|
|
const float vec[3], const float vec1[3],
|
|
const float *orco, const float *uvco, float hasize, float vectsize, int seed, const float pa_co[3])
|
|
{
|
|
HaloRen *har;
|
|
MTex *mtex;
|
|
float tin, tr, tg, tb, ta;
|
|
float xn, yn, zn, texvec[3], hoco[4], hoco1[4], in[3], tex[3], out[3];
|
|
int i, hasrgb;
|
|
|
|
if (hasize==0.0f) return NULL;
|
|
|
|
projectverto(vec, re->winmat, hoco);
|
|
if (hoco[3]==0.0f) return NULL;
|
|
if (vec1) {
|
|
projectverto(vec1, re->winmat, hoco1);
|
|
if (hoco1[3]==0.0f) return NULL;
|
|
}
|
|
|
|
har= RE_findOrAddHalo(obr, obr->tothalo++);
|
|
copy_v3_v3(har->co, vec);
|
|
har->hasize= hasize;
|
|
|
|
/* actual projectvert is done in function project_renderdata() because of parts/border/pano */
|
|
/* we do it here for sorting of halos */
|
|
zn= hoco[3];
|
|
har->xs= 0.5f*re->winx*(hoco[0]/zn);
|
|
har->ys= 0.5f*re->winy*(hoco[1]/zn);
|
|
har->zs= 0x7FFFFF*(hoco[2]/zn);
|
|
|
|
har->zBufDist = 0x7FFFFFFF*(hoco[2]/zn);
|
|
|
|
/* halovect */
|
|
if (vec1) {
|
|
|
|
har->type |= HA_VECT;
|
|
|
|
xn= har->xs - 0.5f*re->winx*(hoco1[0]/hoco1[3]);
|
|
yn= har->ys - 0.5f*re->winy*(hoco1[1]/hoco1[3]);
|
|
if (xn==0.0f || (xn==0.0f && yn==0.0f)) zn= 0.0;
|
|
else zn= atan2(yn, xn);
|
|
|
|
har->sin= sin(zn);
|
|
har->cos= cos(zn);
|
|
zn= len_v3v3(vec1, vec)*0.5f;
|
|
|
|
har->hasize= vectsize*zn + (1.0f-vectsize)*hasize;
|
|
|
|
sub_v3_v3v3(har->no, vec, vec1);
|
|
normalize_v3(har->no);
|
|
}
|
|
|
|
if (ma->mode & MA_HALO_XALPHA) har->type |= HA_XALPHA;
|
|
|
|
har->alfa= ma->alpha;
|
|
har->r= ma->r;
|
|
har->g= ma->g;
|
|
har->b= ma->b;
|
|
har->add= (255.0f*ma->add);
|
|
har->mat= ma;
|
|
har->hard= ma->har;
|
|
har->seed= seed % 256;
|
|
|
|
if (ma->mode & MA_STAR) har->starpoints= ma->starc;
|
|
if (ma->mode & MA_HALO_LINES) har->linec= ma->linec;
|
|
if (ma->mode & MA_HALO_RINGS) har->ringc= ma->ringc;
|
|
if (ma->mode & MA_HALO_FLARE) har->flarec= ma->flarec;
|
|
|
|
if ((ma->mode & MA_HALOTEX) && ma->mtex[0])
|
|
har->tex= 1;
|
|
|
|
for (i=0; i<MAX_MTEX; i++)
|
|
if (ma->mtex[i] && (ma->septex & (1<<i))==0) {
|
|
mtex= ma->mtex[i];
|
|
copy_v3_v3(texvec, vec);
|
|
|
|
if (mtex->texco & TEXCO_NORM) {
|
|
;
|
|
}
|
|
else if (mtex->texco & TEXCO_OBJECT) {
|
|
if (mtex->object)
|
|
mul_m4_v3(mtex->object->imat_ren, texvec);
|
|
}
|
|
else if (mtex->texco & TEXCO_GLOB) {
|
|
copy_v3_v3(texvec, vec);
|
|
}
|
|
else if (mtex->texco & TEXCO_UV && uvco) {
|
|
int uv_index=CustomData_get_named_layer_index(&dm->faceData, CD_MTFACE, mtex->uvname);
|
|
if (uv_index<0)
|
|
uv_index=CustomData_get_active_layer_index(&dm->faceData, CD_MTFACE);
|
|
|
|
uv_index-=CustomData_get_layer_index(&dm->faceData, CD_MTFACE);
|
|
|
|
texvec[0]=2.0f*uvco[2*uv_index]-1.0f;
|
|
texvec[1]=2.0f*uvco[2*uv_index+1]-1.0f;
|
|
texvec[2]=0.0f;
|
|
}
|
|
else if (mtex->texco & TEXCO_PARTICLE) {
|
|
/* particle coordinates in range [0, 1] */
|
|
texvec[0] = 2.f * pa_co[0] - 1.f;
|
|
texvec[1] = 2.f * pa_co[1] - 1.f;
|
|
texvec[2] = pa_co[2];
|
|
}
|
|
else if (orco) {
|
|
copy_v3_v3(texvec, orco);
|
|
}
|
|
|
|
hasrgb = externtex(mtex, texvec, &tin, &tr, &tg, &tb, &ta, 0, re->pool);
|
|
|
|
//yn= tin*mtex->colfac;
|
|
//zn= tin*mtex->alphafac;
|
|
if (mtex->mapto & MAP_COL) {
|
|
tex[0]=tr;
|
|
tex[1]=tg;
|
|
tex[2]=tb;
|
|
out[0]=har->r;
|
|
out[1]=har->g;
|
|
out[2]=har->b;
|
|
|
|
texture_rgb_blend(in, tex, out, tin, mtex->colfac, mtex->blendtype);
|
|
// zn= 1.0-yn;
|
|
//har->r= (yn*tr+ zn*ma->r);
|
|
//har->g= (yn*tg+ zn*ma->g);
|
|
//har->b= (yn*tb+ zn*ma->b);
|
|
har->r= in[0];
|
|
har->g= in[1];
|
|
har->b= in[2];
|
|
}
|
|
|
|
/* alpha returned, so let's use it instead of intensity */
|
|
if (hasrgb)
|
|
tin = ta;
|
|
|
|
if (mtex->mapto & MAP_ALPHA)
|
|
har->alfa = texture_value_blend(mtex->def_var, har->alfa, tin, mtex->alphafac, mtex->blendtype);
|
|
if (mtex->mapto & MAP_HAR)
|
|
har->hard = 1.0f+126.0f*texture_value_blend(mtex->def_var, ((float)har->hard)/127.0f, tin, mtex->hardfac, mtex->blendtype);
|
|
if (mtex->mapto & MAP_RAYMIRR)
|
|
har->hasize = 100.0f*texture_value_blend(mtex->def_var, har->hasize/100.0f, tin, mtex->raymirrfac, mtex->blendtype);
|
|
if (mtex->mapto & MAP_TRANSLU) {
|
|
float add = texture_value_blend(mtex->def_var, (float)har->add/255.0f, tin, mtex->translfac, mtex->blendtype);
|
|
CLAMP(add, 0.f, 1.f);
|
|
har->add = 255.0f*add;
|
|
}
|
|
/* now what on earth is this good for?? */
|
|
//if (mtex->texco & 16) {
|
|
// har->alfa= tin;
|
|
//}
|
|
}
|
|
|
|
har->pool = re->pool;
|
|
|
|
return har;
|
|
}
|
|
|
|
/* -------------------------- operations on entire database ----------------------- */
|
|
|
|
/* ugly function for halos in panorama */
|
|
static int panotestclip(Render *re, bool do_pano, float v[4])
|
|
{
|
|
/* part size (ensure we run RE_parts_clamp first) */
|
|
BLI_assert(re->partx == min_ii(re->r.tilex, re->rectx));
|
|
BLI_assert(re->party == min_ii(re->r.tiley, re->recty));
|
|
|
|
if (do_pano == false) {
|
|
return testclip(v);
|
|
}
|
|
else {
|
|
/* to be used for halos en infos */
|
|
float abs4;
|
|
short c = 0;
|
|
|
|
int xparts = (re->rectx + re->partx - 1) / re->partx;
|
|
|
|
abs4= fabsf(v[3]);
|
|
|
|
if (v[2]< -abs4) c=16; /* this used to be " if (v[2]<0) ", see clippz() */
|
|
else if (v[2]> abs4) c+= 32;
|
|
|
|
if ( v[1]>abs4) c+=4;
|
|
else if ( v[1]< -abs4) c+=8;
|
|
|
|
abs4*= xparts;
|
|
if ( v[0]>abs4) c+=2;
|
|
else if ( v[0]< -abs4) c+=1;
|
|
|
|
return c;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* This adds the hcs coordinates to vertices. It iterates over all
|
|
* vertices, halos and faces. After the conversion, we clip in hcs.
|
|
*
|
|
* Elsewhere, all primites are converted to vertices.
|
|
* Called in
|
|
* - envmapping (envmap.c)
|
|
* - shadow buffering (shadbuf.c)
|
|
*/
|
|
|
|
void project_renderdata(Render *re,
|
|
void (*projectfunc)(const float *, float mat[4][4], float *),
|
|
bool do_pano, float xoffs, bool UNUSED(do_buckets))
|
|
{
|
|
ObjectRen *obr;
|
|
HaloRen *har = NULL;
|
|
float zn, vec[3], hoco[4];
|
|
int a;
|
|
|
|
if (do_pano) {
|
|
float panophi= xoffs;
|
|
|
|
re->panosi= sin(panophi);
|
|
re->panoco= cos(panophi);
|
|
}
|
|
|
|
for (obr=re->objecttable.first; obr; obr=obr->next) {
|
|
/* calculate view coordinates (and zbuffer value) */
|
|
for (a=0; a<obr->tothalo; a++) {
|
|
if ((a & 255)==0) har= obr->bloha[a>>8];
|
|
else har++;
|
|
|
|
if (do_pano) {
|
|
vec[0]= re->panoco*har->co[0] + re->panosi*har->co[2];
|
|
vec[1]= har->co[1];
|
|
vec[2]= -re->panosi*har->co[0] + re->panoco*har->co[2];
|
|
}
|
|
else {
|
|
copy_v3_v3(vec, har->co);
|
|
}
|
|
|
|
projectfunc(vec, re->winmat, hoco);
|
|
|
|
/* we clip halos less critical, but not for the Z */
|
|
hoco[0]*= 0.5f;
|
|
hoco[1]*= 0.5f;
|
|
|
|
if ( panotestclip(re, do_pano, hoco) ) {
|
|
har->miny= har->maxy= -10000; /* that way render clips it */
|
|
}
|
|
else if (hoco[3]<0.0f) {
|
|
har->miny= har->maxy= -10000; /* render clips it */
|
|
}
|
|
else { /* do the projection...*/
|
|
/* bring back hocos */
|
|
hoco[0]*= 2.0f;
|
|
hoco[1]*= 2.0f;
|
|
|
|
zn= hoco[3];
|
|
har->xs= 0.5f*re->winx*(1.0f+hoco[0]/zn); /* the 0.5 negates the previous 2...*/
|
|
har->ys= 0.5f*re->winy*(1.0f+hoco[1]/zn);
|
|
|
|
/* this should be the zbuffer coordinate */
|
|
har->zs= 0x7FFFFF*(hoco[2]/zn);
|
|
/* taking this from the face clip functions? seems ok... */
|
|
har->zBufDist = 0x7FFFFFFF*(hoco[2]/zn);
|
|
|
|
vec[0]+= har->hasize;
|
|
projectfunc(vec, re->winmat, hoco);
|
|
vec[0]-= har->hasize;
|
|
zn= hoco[3];
|
|
har->rad= fabsf(har->xs- 0.5f*re->winx*(1.0f+hoco[0]/zn));
|
|
|
|
/* this clip is not really OK, to prevent stars to become too large */
|
|
if (har->type & HA_ONLYSKY) {
|
|
if (har->rad>3.0f) har->rad= 3.0f;
|
|
}
|
|
|
|
har->radsq= har->rad*har->rad;
|
|
|
|
har->miny= har->ys - har->rad/re->ycor;
|
|
har->maxy= har->ys + har->rad/re->ycor;
|
|
|
|
/* the Zd value is still not really correct for pano */
|
|
|
|
vec[2] -= har->hasize; /* z negative, otherwise it's clipped */
|
|
projectfunc(vec, re->winmat, hoco);
|
|
zn = hoco[3];
|
|
zn = fabsf((float)har->zs - 0x7FFFFF * (hoco[2] / zn));
|
|
har->zd = CLAMPIS(zn, 0, INT_MAX);
|
|
|
|
}
|
|
|
|
}
|
|
}
|
|
}
|
|
|
|
/* ------------------------------------------------------------------------- */
|
|
|
|
ObjectInstanceRen *RE_addRenderInstance(Render *re, ObjectRen *obr, Object *ob, Object *par, int index, int psysindex, float mat[4][4], int lay)
|
|
{
|
|
ObjectInstanceRen *obi;
|
|
float mat3[3][3];
|
|
|
|
obi= MEM_callocN(sizeof(ObjectInstanceRen), "ObjectInstanceRen");
|
|
obi->obr= obr;
|
|
obi->ob= ob;
|
|
obi->par= par;
|
|
obi->index= index;
|
|
obi->psysindex= psysindex;
|
|
obi->lay= lay;
|
|
|
|
if (mat) {
|
|
copy_m4_m4(obi->mat, mat);
|
|
copy_m3_m4(mat3, mat);
|
|
invert_m3_m3(obi->nmat, mat3);
|
|
transpose_m3(obi->nmat);
|
|
obi->flag |= R_DUPLI_TRANSFORMED;
|
|
}
|
|
|
|
BLI_addtail(&re->instancetable, obi);
|
|
|
|
return obi;
|
|
}
|
|
|
|
void RE_makeRenderInstances(Render *re)
|
|
{
|
|
ObjectInstanceRen *obi, *oldobi;
|
|
ListBase newlist;
|
|
int tot;
|
|
|
|
/* convert list of object instances to an array for index based lookup */
|
|
tot= BLI_countlist(&re->instancetable);
|
|
re->objectinstance= MEM_callocN(sizeof(ObjectInstanceRen)*tot, "ObjectInstance");
|
|
re->totinstance= tot;
|
|
newlist.first= newlist.last= NULL;
|
|
|
|
obi= re->objectinstance;
|
|
for (oldobi=re->instancetable.first; oldobi; oldobi=oldobi->next) {
|
|
*obi= *oldobi;
|
|
|
|
if (obi->obr) {
|
|
obi->prev= obi->next= NULL;
|
|
BLI_addtail(&newlist, obi);
|
|
obi++;
|
|
}
|
|
else
|
|
re->totinstance--;
|
|
}
|
|
|
|
BLI_freelistN(&re->instancetable);
|
|
re->instancetable= newlist;
|
|
}
|
|
|
|
/* four functions to facilitate envmap rotation for raytrace */
|
|
void RE_instance_rotate_ray_start(ObjectInstanceRen *obi, Isect *is)
|
|
{
|
|
if (obi && (obi->flag & R_ENV_TRANSFORMED)) {
|
|
copy_v3_v3(is->origstart, is->start);
|
|
mul_m4_v3(obi->imat, is->start);
|
|
}
|
|
}
|
|
|
|
void RE_instance_rotate_ray_dir(ObjectInstanceRen *obi, Isect *is)
|
|
{
|
|
if (obi && (obi->flag & R_ENV_TRANSFORMED)) {
|
|
float end[3];
|
|
|
|
copy_v3_v3(is->origdir, is->dir);
|
|
add_v3_v3v3(end, is->origstart, is->dir);
|
|
|
|
mul_m4_v3(obi->imat, end);
|
|
sub_v3_v3v3(is->dir, end, is->start);
|
|
}
|
|
}
|
|
|
|
void RE_instance_rotate_ray(ObjectInstanceRen *obi, Isect *is)
|
|
{
|
|
RE_instance_rotate_ray_start(obi, is);
|
|
RE_instance_rotate_ray_dir(obi, is);
|
|
}
|
|
|
|
void RE_instance_rotate_ray_restore(ObjectInstanceRen *obi, Isect *is)
|
|
{
|
|
if (obi && (obi->flag & R_ENV_TRANSFORMED)) {
|
|
copy_v3_v3(is->start, is->origstart);
|
|
copy_v3_v3(is->dir, is->origdir);
|
|
}
|
|
}
|
|
|
|
int clip_render_object(float boundbox[2][3], float bounds[4], float winmat[4][4])
|
|
{
|
|
float mat[4][4], vec[4];
|
|
int a, fl, flag = -1;
|
|
|
|
copy_m4_m4(mat, winmat);
|
|
|
|
for (a=0; a < 8; a++) {
|
|
vec[0]= (a & 1)? boundbox[0][0]: boundbox[1][0];
|
|
vec[1]= (a & 2)? boundbox[0][1]: boundbox[1][1];
|
|
vec[2]= (a & 4)? boundbox[0][2]: boundbox[1][2];
|
|
vec[3]= 1.0;
|
|
mul_m4_v4(mat, vec);
|
|
|
|
fl = 0;
|
|
if (bounds) {
|
|
if (vec[0] < bounds[0] * vec[3]) fl |= 1;
|
|
else if (vec[0] > bounds[1] * vec[3]) fl |= 2;
|
|
|
|
if (vec[1] > bounds[3] * vec[3]) fl |= 4;
|
|
else if (vec[1] < bounds[2] * vec[3]) fl |= 8;
|
|
}
|
|
else {
|
|
if (vec[0] < -vec[3]) fl |= 1;
|
|
else if (vec[0] > vec[3]) fl |= 2;
|
|
|
|
if (vec[1] > vec[3]) fl |= 4;
|
|
else if (vec[1] < -vec[3]) fl |= 8;
|
|
}
|
|
if (vec[2] < -vec[3]) fl |= 16;
|
|
else if (vec[2] > vec[3]) fl |= 32;
|
|
|
|
flag &= fl;
|
|
if (flag == 0) {
|
|
return 0;
|
|
}
|
|
}
|
|
|
|
return flag;
|
|
}
|
|
|