static void DeringBlockWeak(unsigned char *SrcPtr,
unsigned char *DstPtr,
ogg_int32_t Pitch,
ogg_uint32_t FragQIndex,
const ogg_uint32_t *QuantScale){
ogg_int16_t UDMod[72];
ogg_int16_t LRMod[72];
unsigned int j,k;
const unsigned char * Src;
unsigned int QValue = QuantScale[FragQIndex];
unsigned char p;
unsigned char pl;
unsigned char pr;
unsigned char pu;
unsigned char pd;
int al;
int ar;
int au;
int ad;
int atot;
int B;
int newVal;
const unsigned char *curRow = SrcPtr-1;
unsigned char *dstRow = DstPtr;
const unsigned char *lastRow = SrcPtr-Pitch;
const unsigned char *nextRow = SrcPtr+Pitch;
unsigned int rowOffset = 0;
unsigned int round = (1<<6);
int High;
int Low;
int TmpMod;
int Sharpen = SharpenModifier[FragQIndex];
High = 3 * QValue;
if(High>24)
High=24;
Low = 0 ;
/* Initialize the Mod Data */
Src=SrcPtr-Pitch;
for(k=0;k<9;k++) {
for(j=0;j<8;j++) {
TmpMod = 32 + QValue - 2*(abs(Src[j+Pitch]-Src[j]));
if(TmpMod< -64)
TmpMod = Sharpen;
else if(TmpMod<Low)
TmpMod = Low;
else if(TmpMod>High)
TmpMod = High;
UDMod[k*8+j] = (ogg_int16_t)TmpMod;
}
Src +=Pitch;
}
Src = SrcPtr-1;
for(k=0;k<8;k++){
for(j=0;j<9;j++){
TmpMod = 32 + QValue - 2*(abs(Src[j+1]-Src[j]));
if(TmpMod< -64 )
TmpMod = Sharpen;
else if(TmpMod<Low)
TmpMod = Low;
else if(TmpMod>High)
TmpMod = High;
LRMod[k*9+j] = (ogg_int16_t)TmpMod;
}
Src+=Pitch;
}
for(k=0;k<8;k++) {
for(j=0;j<8;j++){
atot = 128;
B = round;
p = curRow[ rowOffset +j+1];
pl = curRow[ rowOffset +j];
al = LRMod[k*9+j];
atot -= al;
B += al * pl;
pu = lastRow[ rowOffset +j];
au = UDMod[k*8+j];
atot -= au;
B += au * pu;
pd = nextRow[ rowOffset +j];
ad = UDMod[(k+1)*8+j];
atot -= ad;
B += ad * pd;
pr = curRow[ rowOffset +j+2];
ar = LRMod[k*9+j+1];
atot -= ar;
B += ar * pr;
newVal = ( atot * p + B) >> 7;
dstRow[ rowOffset +j] = clamp255( newVal );
}
rowOffset += Pitch;
}
}
static void DeringBlockStrong(unsigned char *SrcPtr,
unsigned char *DstPtr,
ogg_int32_t Pitch,
ogg_uint32_t FragQIndex,
const ogg_uint32_t *QuantScale){
ogg_int16_t UDMod[72];
ogg_int16_t LRMod[72];
unsigned int j,k,l;
const unsigned char * Src;
unsigned int QValue = QuantScale[FragQIndex];
unsigned char p;
unsigned char pl;
unsigned char pr;
unsigned char pu;
unsigned char pd;
int al;
int ar;
int au;
int ad;
int atot;
int B;
int newVal;
const unsigned char *curRow = SrcPtr - 1; /* avoid negative array indexes */
unsigned char *dstRow = DstPtr;
const unsigned char *lastRow = SrcPtr-Pitch;
const unsigned char *nextRow = SrcPtr+Pitch;
unsigned int rowOffset = 0;
unsigned int round = (1<<6);
int High;
int Low;
int TmpMod;
int Sharpen = SharpenModifier[FragQIndex];
High = 3 * QValue;
if(High>32)High=32;
Low = 0;
/* Initialize the Mod Data */
Src = SrcPtr-Pitch;
for(k=0;k<9;k++){
for(j=0;j<8;j++){
TmpMod = 32 + QValue - (abs(Src[j+Pitch]-Src[j]));
if(TmpMod< -64)
TmpMod = Sharpen;
else if(TmpMod<Low)
TmpMod = Low;
else if(TmpMod>High)
TmpMod = High;
UDMod[k*8+j] = (ogg_int16_t)TmpMod;
}
Src +=Pitch;
}
Src = SrcPtr-1;
for(k=0;k<8;k++){
for(j=0;j<9;j++){
TmpMod = 32 + QValue - (abs(Src[j+1]-Src[j]));
if(TmpMod< -64 )
TmpMod = Sharpen;
else if(TmpMod<0)
TmpMod = Low;
else if(TmpMod>High)
TmpMod = High;
LRMod[k*9+j] = (ogg_int16_t)TmpMod;
}
Src+=Pitch;
}
for(k=0;k<8;k++){
/* In the case that this function called with same buffer for
source and destination, To keep the c and the mmx version to have
consistant results, intermediate buffer is used to store the
eight pixel value before writing them to destination
(i.e. Overwriting souce for the speical case) */
for(l=0;l<8;l++){
atot = 128;
B = round;
p = curRow[ rowOffset +l +1];
pl = curRow[ rowOffset +l];
al = LRMod[k*9+l];
atot -= al;
B += al * pl;
pu = lastRow[ rowOffset +l];
au = UDMod[k*8+l];
atot -= au;
B += au * pu;
pd = nextRow[ rowOffset +l];
ad = UDMod[(k+1)*8+l];
atot -= ad;
B += ad * pd;
pr = curRow[ rowOffset +l+2];
ar = LRMod[k*9+l+1];
atot -= ar;
B += ar * pr;
newVal = ( atot * p + B) >> 7;
dstRow[ rowOffset +l]= clamp255( newVal );
}
rowOffset += Pitch;
}
}
int main(int argc, char *argv[])
{
byte *ibuf, *yibuf, *obuf, *blks, *yibuf2;
byte *tbuf, *tbuf1, *ct1, *ibuf2;
BTIC1H_Context *ctx;
FILE *tfd;
double f, g, h, e;
int xs, ys, xs1, ys1;
int cr, cg, cb, cy, cu, cv, cu1, cv1;
int t0, t1, t2, t3;
int n, n1, nf, ncf, qf, nwe, nwd;
int i, j, k, l;
nwe=4; nwd=4;
for(i=1; i<argc; i++)
{
if(!strncmp(argv[i], "--th=", 5))
{
nwe=atoi(argv[i]+5);
nwd=nwe;
btic1h_workqueue_defaultworkers=nwe;
// btic1h_workqueue_defaultworkers=atoi(argv[i]+5);
continue;
}
if(!strncmp(argv[i], "--thd=", 6))
{
nwd=atoi(argv[i]+6);
continue;
}
if(!strncmp(argv[i], "--the=", 6))
{
nwe=atoi(argv[i]+6);
continue;
}
}
// BTIC1H_InitCamera();
ibuf=BTIC1H_Img_LoadTGA("StreetHollandSD.tga", &xs, &ys);
ibuf2=BTIC1H_Img_LoadTGA("StreetHollandSD_q45.tga", &xs1, &ys1);
// ibuf=BTIC1H_Img_LoadTGA("3917969_f260.tga", &xs, &ys);
ibuf=BTIC1H_Img_LoadTGA("screencap0.tga", &xs, &ys);
ibuf2=BTIC1H_Img_LoadTGA("screencap0.tga", &xs1, &ys1);
// ibuf=BTIC1H_Img_LoadTGA("MLP_FIM1.tga", &xs, &ys);
// ibuf2=BTIC1H_Img_LoadTGA("MLP_FIM1_q95.tga", &xs1, &ys1);
ibuf=BTIC1H_Img_LoadTGA("MLP_FIM3.tga", &xs, &ys);
// ibuf2=BTIC1H_Img_LoadTGA("MLP_FIM3.tga", &xs1, &ys1);
ibuf2=BTIC1H_Img_LoadTGA("MLP_FIM3_q75.tga", &xs1, &ys1);
yibuf=malloc(xs*ys*2);
if(!ibuf || !ibuf2)
{
printf("failed load test images\n");
return(-1);
}
obuf=malloc(xs*ys*4);
// blks=malloc(xs*ys*2);
tbuf=malloc(1<<24);
tbuf1=malloc(1<<24);
#if 0
for(i=0; i<ys; i++)
for(j=0; j<xs; j++)
{
cr=ibuf[(i*xs+j)*4+0];
cg=ibuf[(i*xs+j)*4+1];
cb=ibuf[(i*xs+j)*4+2];
cy=(( 19595*cr + 38470*cg + 7471*cb + 32768)>>16)+0;
cu=((-11056*cr - 21712*cg +32768*cb + 32768)>>16)+128;
cv=(( 32768*cr - 27440*cg - 5328*cb + 32768)>>16)+128;
// cu=128; cv=128;
cy=clamp255(cy);
cu=clamp255(cu);
cv=clamp255(cv);
// k=j&(~1);
yibuf[(i*xs+j)*2+0]=cy;
// yibuf[(i*xs+k)*2+1]=cu;
// yibuf[(i*xs+k)*2+3]=cv;
// yibuf[(i*xs+j)*2+1]=(j&1)?cv:cu;
// yibuf[(i*xs+j)*2+1]=(yibuf[(i*xs+j)*2+1]+((j&1)?cv:cu))>>1;
// yibuf[(i*xs+j)*2+3]=(j&1)?cu:cv;
if(j&1)
{
k=j&(~1);
yibuf[(i*xs+k)*2+1]=(yibuf[(i*xs+k)*2+1]+cu)>>1;
yibuf[(i*xs+k)*2+3]=(yibuf[(i*xs+k)*2+3]+cv)>>1;
}else
{
static void
updatereflectmap(int slot)
{
float rf, r, g, b, t, dfact, kfact, rdir[3];
float rcol[3], ppos[3], norm[3], ldir[3], h[3], vdir[3], planepos[3];
int x, y;
glBindTexture(GL_TEXTURE_2D, reflectid);
for (y = slot * TEX_REFLECT_SLOT_SIZE;
y < (slot + 1) * TEX_REFLECT_SLOT_SIZE; y++)
for (x = 0; x < TEX_REFLECT_WIDTH; x++) {
ppos[0] = sphere_pos[y][x][0] + objpos[0];
ppos[1] = sphere_pos[y][x][1] + objpos[1];
ppos[2] = sphere_pos[y][x][2] + objpos[2];
vsub(norm, ppos, objpos);
vnormalize(norm, norm);
vsub(ldir, lightpos, ppos);
vnormalize(ldir, ldir);
vsub(vdir, obs, ppos);
vnormalize(vdir, vdir);
rf = 2.0f * dprod(norm, vdir);
if (rf > EPSILON) {
rdir[0] = rf * norm[0] - vdir[0];
rdir[1] = rf * norm[1] - vdir[1];
rdir[2] = rf * norm[2] - vdir[2];
t = -objpos[2] / rdir[2];
if (t > EPSILON) {
planepos[0] = objpos[0] + t * rdir[0];
planepos[1] = objpos[1] + t * rdir[1];
planepos[2] = 0.0f;
if (!colorcheckmap(planepos, rcol))
rcol[0] = rcol[1] = rcol[2] = 0.0f;
}
else
rcol[0] = rcol[1] = rcol[2] = 0.0f;
}
else
rcol[0] = rcol[1] = rcol[2] = 0.0f;
dfact = 0.1f * dprod(ldir, norm);
if (dfact < 0.0f) {
dfact = 0.0f;
kfact = 0.0f;
}
else {
h[0] = 0.5f * (vdir[0] + ldir[0]);
h[1] = 0.5f * (vdir[1] + ldir[1]);
h[2] = 0.5f * (vdir[2] + ldir[2]);
kfact = dprod(h, norm);
kfact = pow(kfact, 4.0);
if (kfact < 1.0e-10)
kfact = 0.0;
}
r = dfact + kfact;
g = dfact + kfact;
b = dfact + kfact;
r *= 255.0f;
g *= 255.0f;
b *= 255.0f;
r += rcol[0];
g += rcol[1];
b += rcol[2];
r = clamp255(r);
g = clamp255(g);
b = clamp255(b);
reflectmap[y][x][0] = (GLubyte) r;
reflectmap[y][x][1] = (GLubyte) g;
reflectmap[y][x][2] = (GLubyte) b;
}
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, slot * TEX_REFLECT_SLOT_SIZE,
TEX_REFLECT_WIDTH, TEX_REFLECT_SLOT_SIZE, GL_RGB,
GL_UNSIGNED_BYTE,
&reflectmap[slot * TEX_REFLECT_SLOT_SIZE][0][0]);
}
static int
colorcheckmap(float ppos[3], float c[3])
{
static float norm[3] = { 0.0f, 0.0f, 1.0f };
float ldir[3], vdir[3], h[3], dfact, kfact, r, g, b;
int x, y;
x = (int) ((ppos[0] + BASESIZE / 2) * (10.0f / BASESIZE));
if ((x < 0) || (x > 10))
return GL_FALSE;
y = (int) ((ppos[1] + BASESIZE / 2) * (10.0f / BASESIZE));
if ((y < 0) || (y > 10))
return GL_FALSE;
r = 255.0f;
if (y & 1) {
if (x & 1)
g = 255.0f;
else
g = 0.0f;
}
else {
if (x & 1)
g = 0.0f;
else
g = 255.0f;
}
b = 0.0f;
vsub(ldir, lightpos, ppos);
vnormalize(ldir, ldir);
if (seelight(ppos, ldir)) {
c[0] = r * 0.05f;
c[1] = g * 0.05f;
c[2] = b * 0.05f;
return GL_TRUE;
}
dfact = dprod(ldir, norm);
if (dfact < 0.0f)
dfact = 0.0f;
vsub(vdir, obs, ppos);
vnormalize(vdir, vdir);
h[0] = 0.5f * (vdir[0] + ldir[0]);
h[1] = 0.5f * (vdir[1] + ldir[1]);
h[2] = 0.5f * (vdir[2] + ldir[2]);
kfact = dprod(h, norm);
kfact = pow(kfact, 6.0) * 7.0 * 255.0;
r = r * dfact + kfact;
g = g * dfact + kfact;
b = b * dfact + kfact;
c[0] = clamp255(r);
c[1] = clamp255(g);
c[2] = clamp255(b);
return GL_TRUE;
}
void render_col(BITMAP* buffer, MAP* map, short idx, VEC2F vp, float cam_h, VEC2F ray_dir, int col)
{
if(idx & SHORT_SIGN_FLAG)
{
//printf("ASDF = %d %d\n", (short)(idx & ~SHORT_SIGN_FLAG), idx);
SSECTOR* ssector = &map->ssectors[(short)(idx & ~SHORT_SIGN_FLAG)];
int i;
for(i = 0; i < ssector->seg_num; ++i)
{
SEG* seg = &map->segs[ssector->first_seg_idx + i];
VEC2F v1 = get_vertex(map, seg->v1_idx);
VEC2F v2 = get_vertex(map, seg->v2_idx);
RAY_VS_SEGMENT_RESULT res = ray_vs_segment(v1, v2, vp, ray_dir);
VEC2F n = vec2f(v1.y - v2.y, v2.x - v1.x);
LINEDEF* linedef = &map->linedefs[seg->linedef_idx];
SIDEDEF *sidedef1, *sidedef2;
SECTOR *sector1, *sector2;
if(seg->dir)
{
sidedef1 = &map->sidedefs[linedef->neg_sidedef_idx];
sector1 = &map->sectors[sidedef1->sector_idx];
sidedef2 = linedef->pos_sidedef_idx == -1 ? NULL : &map->sidedefs[linedef->pos_sidedef_idx];
sector2 = linedef->pos_sidedef_idx == -1 ? NULL : &map->sectors[sidedef2->sector_idx];
}
else
{
sidedef1 = &map->sidedefs[linedef->pos_sidedef_idx];
sector1 = &map->sectors[sidedef1->sector_idx];
sidedef2 = linedef->neg_sidedef_idx == -1 ? NULL : &map->sidedefs[linedef->neg_sidedef_idx];
sector2 = linedef->neg_sidedef_idx == -1 ? NULL : &map->sectors[sidedef2->sector_idx];
}
//printf("%c%c%c%c%c%c%c%c\n", sidedef->mid_tex_name[0], sidedef->mid_tex_name[1],
//sidedef->mid_tex_name[2], sidedef->mid_tex_name[3], sidedef->mid_tex_name[4], sidedef->mid_tex_name[5],
//sidedef->mid_tex_name[6], sidedef->mid_tex_name[7]);
if(res.t > 10 && res.k >= 0 && res.k <= 1.0 && vec2f_dot(n, vec2f_diff(vp, v1)) < 0)
{
VEC2F d = vec2f_diff(v2, v1);
VEC2F wall_normal = vec2f(-d.y, d.x);
float wall_len = sqrt(d.x * d.x + d.y * d.y);
wall_normal.x /= wall_len;
wall_normal.y /= wall_len;
int r = clamp255((int)(fabs(wall_normal.x) * 4000 / sqrt(res.t)));
int g = clamp255((int)(fabs(wall_normal.y) * 4000 / sqrt(res.t)));
int color = makecol(r, g, 0);
if(sidedef1 && !sidedef2)
render_col_helper(col, sector1->ceil_height + cam_h, sector1->floor_height + cam_h, res.t, color);
else
{
render_col_helper(col, sector2->ceil_height + cam_h, sector1->ceil_height + cam_h, res.t, color);
render_col_helper(col, sector2->floor_height + cam_h, sector1->floor_height + cam_h, res.t, color);
if(sidedef1->mid_tex_name[0] != '-')
render_col_helper(col, sector1->floor_height + cam_h, sector1->ceil_height + cam_h, res.t, color);
}
}
}
// render subsector...
}
else
{
NODE* node = &map->nodes[idx];
VEC2F n = vec2f(-node->y2, node->x2);
if(vec2f_dot(n, vec2f(vp.x - node->x1, vp.y - node->y1)) >= 0)
{
render_col(buffer, map, node->left_child_idx, vp, cam_h, ray_dir, col);
render_col(buffer, map, node->right_child_idx, vp, cam_h, ray_dir, col);
}
else
{
render_col(buffer, map, node->right_child_idx, vp, cam_h, ray_dir, col);
render_col(buffer, map, node->left_child_idx, vp, cam_h, ray_dir, col);
}
}
}
