sandey/blender
1
0
Fork 0
forked from bartvdbraak/blender
Code Pull Requests Activity

- last time commit. It fixes bug reported by Kent Mein. Polygonisation of Metaball was broken. Test file:

Browse Source 
http://www.cs.umn.edu/~mein/blender/blend/metaball.blend
    http://www.cs.umn.edu/~mein/blender/blend/metaball.png

 - it should fill octal tree much more better ... polygonisation of MetaBalls is faster little bit now too
 - stiffness limit is 10.0
This commit is contained in:
Jiri Hnidek 2005-05-30 06:37:33 +00:00
parent 4b99e61cf2
commit 8a59480289
1 changed files with 70 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

93
source/blender/blenkernel/intern/mball.c
View File

@ -1485,6 +1485,16 @@ float init_meta(Object *ob) /* return totsize */
}
while(ml) {
if(!(ml->flag & MB_HIDE)) {
int i;
float max_x, max_y, max_z, min_x, min_y, min_z;
max_x = max_y = max_z = -3.4e38;
min_x = min_y = min_z = 3.4e38;
/* too big stiffness seems only ugly due to linear interpolation
* no need to have possibility for too big stiffness */
if(ml->s > 10.0) ml->s = 10.0;
/* Rotation of MetaElem is stored in quat */
QuatToMat4(ml->quat, temp3);
@ -1500,7 +1510,7 @@ float init_meta(Object *ob) /* return totsize */
mainb[a]= new_pgn_element(sizeof(MetaElem));
*(mainb[a])= *ml;
mainb[a]->bb = new_pgn_element(sizeof(BoundBox));
mat= new_pgn_element(4*4*sizeof(float));
imat= new_pgn_element(4*4*sizeof(float));
@ -1516,29 +1526,66 @@ float init_meta(Object *ob) /* return totsize */
mainb[a]->mat= (float*) mat;
mainb[a]->imat= (float*) imat;
if(ml->type==MB_BALL){
max= 0.0;
/* untransformed Bounding Box of MetaElem */
/* 0 */
mainb[a]->bb->vec[0][0]= -ml->expx;
mainb[a]->bb->vec[0][1]= -ml->expy;
mainb[a]->bb->vec[0][2]= -ml->expz;
/* 1 */
mainb[a]->bb->vec[1][0]= ml->expx;
mainb[a]->bb->vec[1][1]= -ml->expy;
mainb[a]->bb->vec[1][2]= -ml->expz;
/* 2 */
mainb[a]->bb->vec[2][0]= ml->expx;
mainb[a]->bb->vec[2][1]= ml->expy;
mainb[a]->bb->vec[2][2]= -ml->expz;
/* 3 */
mainb[a]->bb->vec[3][0]= -ml->expx;
mainb[a]->bb->vec[3][1]= ml->expy;
mainb[a]->bb->vec[3][2]= -ml->expz;
/* 4 */
mainb[a]->bb->vec[4][0]= -ml->expx;
mainb[a]->bb->vec[4][1]= -ml->expy;
mainb[a]->bb->vec[4][2]= ml->expz;
/* 5 */
mainb[a]->bb->vec[5][0]= ml->expx;
mainb[a]->bb->vec[5][1]= -ml->expy;
mainb[a]->bb->vec[5][2]= ml->expz;
/* 6 */
mainb[a]->bb->vec[6][0]= ml->expx;
mainb[a]->bb->vec[6][1]= ml->expy;
mainb[a]->bb->vec[6][2]= ml->expz;
/* 7 */
mainb[a]->bb->vec[7][0]= -ml->expx;
mainb[a]->bb->vec[7][1]= ml->expy;
mainb[a]->bb->vec[7][2]= ml->expz;
/* transformation of Metalem bb */
for(i=0; i<8; i++)
Mat4MulVecfl((float ( * )[4])mat, mainb[a]->bb->vec[i]);
/* find max and min of transformed bb */
for(i=0; i<8; i++){
/* find maximums */
if(mainb[a]->bb->vec[i][0] > max_x) max_x = mainb[a]->bb->vec[i][0];
if(mainb[a]->bb->vec[i][1] > max_y) max_y = mainb[a]->bb->vec[i][1];
if(mainb[a]->bb->vec[i][2] > max_z) max_z = mainb[a]->bb->vec[i][2];
/* find minimums */
if(mainb[a]->bb->vec[i][0] < min_x) min_x = mainb[a]->bb->vec[i][0];
if(mainb[a]->bb->vec[i][1] < min_y) min_y = mainb[a]->bb->vec[i][1];
if(mainb[a]->bb->vec[i][2] < min_z) min_z = mainb[a]->bb->vec[i][2];
}
else if((ml->type==MB_TUBE)){
max= bob->size[0]*ml->expx;
}
else if((ml->type==MB_PLANE)){
max= MAX2(bob->size[0]*ml->expx, bob->size[1]*ml->expy);
}
else if((ml->type==MB_CUBE)||(ml->type==MB_ELIPSOID)){
max= MAX3(bob->size[0]*ml->expx,
bob->size[1]*ml->expy,
bob->size[2]*ml->expz);
}
mainb[a]->bb->vec[0][0]= mat[3][0] - bob->size[0]*(max+ml->rad)/ob->size[0];
mainb[a]->bb->vec[0][1]= mat[3][1] - bob->size[1]*(max+ml->rad)/ob->size[1];
mainb[a]->bb->vec[0][2]= mat[3][2] - bob->size[2]*(max+ml->rad)/ob->size[2];
mainb[a]->bb->vec[6][0]= mat[3][0] + bob->size[0]*(max+ml->rad)/ob->size[0];
mainb[a]->bb->vec[6][1]= mat[3][1] + bob->size[1]*(max+ml->rad)/ob->size[1];
mainb[a]->bb->vec[6][2]= mat[3][2] + bob->size[2]*(max+ml->rad)/ob->size[2];
/* create "new" bb, only point 0 and 6, which are
* neccesary for octal tree filling */
mainb[a]->bb->vec[0][0] = min_x - ml->rad;
mainb[a]->bb->vec[0][1] = min_y - ml->rad;
mainb[a]->bb->vec[0][2] = min_z - ml->rad;
mainb[a]->bb->vec[6][0] = max_x + ml->rad;
mainb[a]->bb->vec[6][1] = max_y + ml->rad;
mainb[a]->bb->vec[6][2] = max_z + ml->rad;
a++;
}