ViewVC Help

View File | Revision Log | Show Annotations | Download File | Root Listing

root/radiance/ray/src/cv/pabopto2xml.c

Comparing ray/src/cv/pabopto2xml.c (file contents):
Revision 2.1 by greg, Fri Aug 24 15:20:18 2012 UTC vs.
Revision 2.13 by greg, Fri Sep 21 05:17:22 2012 UTC

#	Line 16 | Line 16 | static const char RCSid[] = "$Id$";
16		#include <math.h>
17		#include "bsdf.h"
18
19	+	#define DEBUG           1
20	+
21		#ifndef GRIDRES
22	<	#define GRIDRES         200             /* max. grid resolution per side */
22	>	#define GRIDRES         200             /* grid resolution per side */
23		#endif
24
25	<	#define RSCA            1.9             /* radius scaling factor (empirical) */
24	<	#define MSCA            .12             /* magnitude scaling (empirical) */
25	>	#define RSCA            2.7             /* radius scaling factor (empirical) */
26
27	+	/* convert to/from coded radians */
28	+	#define ANG2R(r)        (int)((r)*((1<<16)/M_PI))
29	+	#define R2ANG(c)        (((c)+.5)*(M_PI/(1<<16)))
30	+
31		typedef struct {
32	<	float           vsum;           /* BSDF sum */
32	>	float           vsum;           /* DSF sum */
33		unsigned short  nval;           /* number of values in sum */
34	<	unsigned short  hrad2;          /* half radius squared */
34	>	unsigned short  crad;           /* radius (coded angle) */
35		} GRIDVAL;                      /* grid value */
36
37		typedef struct {
38	<	float           bsdf;           /* BSDF value at peak */
39	<	unsigned short  rad;            /* radius */
38	>	float           peak;           /* lobe value at peak */
39	>	unsigned short  crad;           /* radius (coded angle) */
40		unsigned char   gx, gy;         /* grid position */
41		} RBFVAL;                       /* radial basis function value */
42
43	<	typedef struct s_rbflist {
44	<	struct s_rbflist        *next;          /* next in our RBF list */
43	>	struct s_rbfnode;               /* forward declaration of RBF struct */
44	>
45	>	typedef struct s_migration {
46	>	struct s_migration      *next;          /* next in global edge list */
47	>	struct s_rbfnode        *rbfv[2];       /* from,to vertex */
48	>	struct s_migration      *enxt[2];       /* next from,to sibling */
49	>	float                   mtx[1];         /* matrix (extends struct) */
50	>	} MIGRATION;                    /* migration link (winged edge structure) */
51	>
52	>	typedef struct s_rbfnode {
53	>	struct s_rbfnode        *next;          /* next in global RBF list */
54	>	MIGRATION               *ejl;           /* edge list for this vertex */
55		FVECT                   invec;          /* incident vector direction */
56	+	double                  vtotal;         /* volume for normalization */
57		int                     nrbf;           /* number of RBFs */
58		RBFVAL                  rbfa[1];        /* RBF array (extends struct) */
59	<	} RBFLIST;                      /* RBF representation of BSDF @ 1 incidence */
59	>	} RBFNODE;                      /* RBF representation of DSF @ 1 incidence */
60
61		/* our loaded grid for this incident angle */
62	<	static double   theta_in_deg, phi_in_deg;
63	<	static GRIDVAL  bsdf_grid[GRIDRES][GRIDRES];
62	>	static double           theta_in_deg, phi_in_deg;
63	>	static GRIDVAL          dsf_grid[GRIDRES][GRIDRES];
64
65	<	/* processed incident BSDF measurements */
66	<	static RBFLIST  *bsdf_list = NULL;
65	>	/* all incident angles in-plane so far? */
66	>	static int              single_plane_incident = -1;
67
68	<	/* Count up non-empty nodes and build RBF representation from current grid */
69	<	static RBFLIST *
70	<	make_rbfrep(void)
55	<	{
56	<	int     nn = 0;
57	<	RBFLIST *newnode;
58	<	int     i, j;
59	<	/* count non-empty bins */
60	<	for (i = 0; i < GRIDRES; i++)
61	<	for (j = 0; j < GRIDRES; j++)
62	<	nn += (bsdf_grid[i][j].nval > 0);
63	<	/* allocate RBF array */
64	<	newnode = (RBFLIST *)malloc(sizeof(RBFLIST) + sizeof(RBFVAL)*(nn-1));
65	<	if (newnode == NULL) {
66	<	fputs("Out of memory in make_rbfrep\n", stderr);
67	<	exit(1);
68	<	}
69	<	newnode->invec[2] = sin(M_PI/180.*theta_in_deg);
70	<	newnode->invec[0] = cos(M_PI/180.*phi_in_deg)*newnode->invec[2];
71	<	newnode->invec[1] = sin(M_PI/180.*phi_in_deg)*newnode->invec[2];
72	<	newnode->invec[2] = sqrt(1. - newnode->invec[2]*newnode->invec[2]);
73	<	newnode->nrbf = nn;
74	<	nn = 0;                 /* fill RBF array */
75	<	for (i = 0; i < GRIDRES; i++)
76	<	for (j = 0; j < GRIDRES; j++)
77	<	if (bsdf_grid[i][j].nval) {
78	<	newnode->rbfa[nn].bsdf = MSCA*bsdf_grid[i][j].vsum /
79	<	(double)bsdf_grid[i][j].nval;
80	<	newnode->rbfa[nn].rad =
81	<	(int)(2.*RSCA*sqrt((double)bsdf_grid[i][j].hrad2) + .5);
82	<	newnode->rbfa[nn].gx = i;
83	<	newnode->rbfa[nn].gy = j;
84	<	++nn;
85	<	}
86	<	newnode->next = bsdf_list;
87	<	return(bsdf_list = newnode);
88	<	}
68	>	/* input/output orientations */
69	>	static int              input_orient = 0;
70	>	static int              output_orient = 0;
71
72	<	/* Compute grid position from normalized outgoing vector */
73	<	static void
92	<	pos_from_vec(int pos[2], const FVECT vec)
93	<	{
94	<	double  sq[2];          /* uniform hemispherical projection */
95	<	double  norm = 1./sqrt(1. + vec[2]);
72	>	/* processed incident DSF measurements */
73	>	static RBFNODE          *dsf_list = NULL;
74
75	<	SDdisk2square(sq, vec[0]*norm, vec[1]*norm);
75	>	/* RBF-linking matrices (edges) */
76	>	static MIGRATION        *mig_list = NULL;
77
78	<	pos[0] = (int)(sq[0]*GRIDRES);
79	<	pos[1] = (int)(sq[1]*GRIDRES);
78	>	/* migration edges drawn in raster fashion */
79	>	static MIGRATION        *mig_grid[GRIDRES][GRIDRES];
80	>
81	>	#define mtx_nrows(m)    ((m)->rbfv[0]->nrbf)
82	>	#define mtx_ncols(m)    ((m)->rbfv[1]->nrbf)
83	>	#define mtx_ndx(m,i,j)  ((i)*mtx_ncols(m) + (j))
84	>	#define is_src(rbf,m)   ((rbf) == (m)->rbfv[0])
85	>	#define is_dest(rbf,m)  ((rbf) == (m)->rbfv[1])
86	>	#define nextedge(rbf,m) (m)->enxt[is_dest(rbf,m)]
87	>	#define opp_rbf(rbf,m)  (m)->rbfv[is_src(rbf,m)]
88	>
89	>	#define round(v)        (int)((v) + .5 - ((v) < -.5))
90	>
91	>	char                    *progname;
92	>
93	>	#ifdef DEBUG                    /* percentage to cull (<0 to turn off) */
94	>	int                     pctcull = -1;
95	>	#else
96	>	int                     pctcull = 90;
97	>	#endif
98	>	/* sampling order (set by data density) */
99	>	int                     samp_order = 0;
100	>
101	>	/* Compute volume associated with Gaussian lobe */
102	>	static double
103	>	rbf_volume(const RBFVAL *rbfp)
104	>	{
105	>	double  rad = R2ANG(rbfp->crad);
106	>
107	>	return((2.*M_PI) * rbfp->peak * rad*rad);
108		}
109
110		/* Compute outgoing vector from grid position */
111		static void
112	<	vec_from_pos(FVECT vec, int xpos, int ypos)
112	>	ovec_from_pos(FVECT vec, int xpos, int ypos)
113		{
114		double  uv[2];
115		double  r2;
#	Line 113 | Line 120 | vec_from_pos(FVECT vec, int xpos, int ypos)
120		vec[0] = vec[1] = sqrt(2. - r2);
121		vec[0] *= uv[0];
122		vec[1] *= uv[1];
123	<	vec[2] = 1. - r2;
123	>	vec[2] = output_orient*(1. - r2);
124		}
125
126	<	/* Evaluate RBF at this grid position */
127	<	static double
128	<	eval_rbfrep2(const RBFLIST *rp, int xi, int yi)
126	>	/* Compute grid position from normalized input/output vector */
127	>	static void
128	>	pos_from_vec(int pos[2], const FVECT vec)
129		{
130	<	double          res = .0;
131	<	const RBFVAL    *rbfp;
125	<	double          sig2;
126	<	int             x2, y2;
127	<	int             n;
130	>	double  sq[2];          /* uniform hemispherical projection */
131	>	double  norm = 1./sqrt(1. + fabs(vec[2]));
132
133	<	rbfp = rp->rbfa;
134	<	for (n = rp->nrbf; n--; rbfp++) {
135	<	x2 = (signed)rbfp->gx - xi;
136	<	x2 *= x2;
133	<	y2 = (signed)rbfp->gy - yi;
134	<	y2 *= y2;
135	<	sig2 = -.5*(x2 + y2)/(double)(rbfp->rad*rbfp->rad);
136	<	if (sig2 > -19.)
137	<	res += rbfp->bsdf * exp(sig2);
138	<	}
139	<	return(res);
133	>	SDdisk2square(sq, vec[0]*norm, vec[1]*norm);
134	>
135	>	pos[0] = (int)(sq[0]*GRIDRES);
136	>	pos[1] = (int)(sq[1]*GRIDRES);
137		}
138
139	<	/* Evaluate RBF for BSDF at the given normalized outgoing direction */
139	>	/* Evaluate RBF for DSF at the given normalized outgoing direction */
140		static double
141	<	eval_rbfrep(const RBFLIST *rp, const FVECT outvec)
141	>	eval_rbfrep(const RBFNODE *rp, const FVECT outvec)
142		{
143		double          res = .0;
144		const RBFVAL    *rbfp;
#	Line 149 | Line 146 | eval_rbfrep(const RBFLIST *rp, const FVECT outvec)
146		double          sig2;
147		int             n;
148
149	+	if (rp == NULL)
150	+	return(.0);
151		rbfp = rp->rbfa;
152		for (n = rp->nrbf; n--; rbfp++) {
153	<	vec_from_pos(odir, rbfp->gx, rbfp->gy);
154	<	sig2 = (DOT(odir, outvec) - 1.) /
155	<	((M_PI*M_PI/(double)(GRIDRES*GRIDRES)) *
157	<	rbfp->rad*rbfp->rad);
153	>	ovec_from_pos(odir, rbfp->gx, rbfp->gy);
154	>	sig2 = R2ANG(rbfp->crad);
155	>	sig2 = (DOT(odir,outvec) - 1.) / (sig2*sig2);
156		if (sig2 > -19.)
157	<	res += rbfp->bsdf * exp(sig2);
157	>	res += rbfp->peak * exp(sig2);
158		}
159		return(res);
160		}
161
162	+	/* Insert a new directional scattering function in our global list */
163	+	static void
164	+	insert_dsf(RBFNODE *newrbf)
165	+	{
166	+	RBFNODE         *rbf, *rbf_last;
167	+	/* keep in ascending theta order */
168	+	for (rbf_last = NULL, rbf = dsf_list;
169	+	single_plane_incident & (rbf != NULL);
170	+	rbf_last = rbf, rbf = rbf->next)
171	+	if (input_orient*rbf->invec[2] < input_orient*newrbf->invec[2])
172	+	break;
173	+	if (rbf_last == NULL) {
174	+	newrbf->next = dsf_list;
175	+	dsf_list = newrbf;
176	+	return;
177	+	}
178	+	newrbf->next = rbf;
179	+	rbf_last->next = newrbf;
180	+	}
181	+
182	+	/* Count up filled nodes and build RBF representation from current grid */
183	+	static RBFNODE *
184	+	make_rbfrep(void)
185	+	{
186	+	int     niter = 16;
187	+	int     minrad = ANG2R(pow(2., 1.-samp_order));
188	+	double  lastVar, thisVar = 100.;
189	+	int     nn;
190	+	RBFNODE *newnode;
191	+	int     i, j;
192	+
193	+	nn = 0;                 /* count selected bins */
194	+	for (i = 0; i < GRIDRES; i++)
195	+	for (j = 0; j < GRIDRES; j++)
196	+	nn += dsf_grid[i][j].nval;
197	+	/* allocate RBF array */
198	+	newnode = (RBFNODE *)malloc(sizeof(RBFNODE) + sizeof(RBFVAL)*(nn-1));
199	+	if (newnode == NULL) {
200	+	fputs("Out of memory in make_rbfrep()\n", stderr);
201	+	exit(1);
202	+	}
203	+	newnode->next = NULL;
204	+	newnode->ejl = NULL;
205	+	newnode->invec[2] = sin(M_PI/180.*theta_in_deg);
206	+	newnode->invec[0] = cos(M_PI/180.*phi_in_deg)*newnode->invec[2];
207	+	newnode->invec[1] = sin(M_PI/180.*phi_in_deg)*newnode->invec[2];
208	+	newnode->invec[2] = input_orient*sqrt(1. - newnode->invec[2]*newnode->invec[2]);
209	+	newnode->vtotal = 0;
210	+	newnode->nrbf = nn;
211	+	nn = 0;                 /* fill RBF array */
212	+	for (i = 0; i < GRIDRES; i++)
213	+	for (j = 0; j < GRIDRES; j++)
214	+	if (dsf_grid[i][j].nval) {
215	+	newnode->rbfa[nn].peak = dsf_grid[i][j].vsum;
216	+	newnode->rbfa[nn].crad = RSCA*dsf_grid[i][j].crad + .5;
217	+	newnode->rbfa[nn].gx = i;
218	+	newnode->rbfa[nn].gy = j;
219	+	if (newnode->rbfa[nn].crad < minrad)
220	+	minrad = newnode->rbfa[nn].crad;
221	+	++nn;
222	+	}
223	+	/* iterate to improve interpolation accuracy */
224	+	do {
225	+	double  dsum = 0, dsum2 = 0;
226	+	nn = 0;
227	+	for (i = 0; i < GRIDRES; i++)
228	+	for (j = 0; j < GRIDRES; j++)
229	+	if (dsf_grid[i][j].nval) {
230	+	FVECT   odir;
231	+	double  corr;
232	+	ovec_from_pos(odir, i, j);
233	+	newnode->rbfa[nn++].peak *= corr =
234	+	dsf_grid[i][j].vsum /
235	+	eval_rbfrep(newnode, odir);
236	+	dsum += corr - 1.;
237	+	dsum2 += (corr-1.)*(corr-1.);
238	+	}
239	+	lastVar = thisVar;
240	+	thisVar = dsum2/(double)nn;
241	+	#ifdef DEBUG
242	+	fprintf(stderr, "Avg., RMS error: %.1f%%  %.1f%%\n",
243	+	100.*dsum/(double)nn,
244	+	100.*sqrt(thisVar));
245	+	#endif
246	+	} while (--niter > 0 && lastVar-thisVar > 0.02*lastVar);
247	+
248	+	nn = 0;                 /* compute sum for normalization */
249	+	while (nn < newnode->nrbf)
250	+	newnode->vtotal += rbf_volume(&newnode->rbfa[nn++]);
251	+
252	+	insert_dsf(newnode);
253	+	/* adjust sampling resolution */
254	+	samp_order = log(2./R2ANG(minrad))/M_LN2 + .5;
255	+
256	+	return(newnode);
257	+	}
258	+
259		/* Load a set of measurements corresponding to a particular incident angle */
260		static int
261	<	load_bsdf_meas(const char *fname)
261	>	load_pabopto_meas(const char *fname)
262		{
263		FILE    *fp = fopen(fname, "r");
264		int     inp_is_DSF = -1;
265	<	double  theta_out, phi_out, val;
265	>	double  new_phi, theta_out, phi_out, val;
266		char    buf[2048];
267		int     n, c;
268
#	Line 176 | Line 271 | load_bsdf_meas(const char *fname)
271		fputs(": cannot open\n", stderr);
272		return(0);
273		}
274	<	memset(bsdf_grid, 0, sizeof(bsdf_grid));
274	>	memset(dsf_grid, 0, sizeof(dsf_grid));
275	>	#ifdef DEBUG
276	>	fprintf(stderr, "Loading measurement file '%s'...\n", fname);
277	>	#endif
278		/* read header information */
279		while ((c = getc(fp)) == '#' || c == EOF) {
280		if (fgets(buf, sizeof(buf), fp) == NULL) {
#	Line 195 | Line 293 | load_bsdf_meas(const char *fname)
293		}
294		if (sscanf(buf, "intheta %lf", &theta_in_deg) == 1)
295		continue;
296	<	if (sscanf(buf, "inphi %lf", &phi_in_deg) == 1)
296	>	if (sscanf(buf, "inphi %lf", &new_phi) == 1)
297		continue;
298		if (sscanf(buf, "incident_angle %lf %lf",
299	<	&theta_in_deg, &phi_in_deg) == 2)
299	>	&theta_in_deg, &new_phi) == 2)
300		continue;
301		}
302		if (inp_is_DSF < 0) {
#	Line 207 | Line 305 | load_bsdf_meas(const char *fname)
305		fclose(fp);
306		return(0);
307		}
308	<	ungetc(c, fp);          /* read actual data */
308	>	if (!input_orient)              /* check input orientation */
309	>	input_orient = 1 - 2*(theta_in_deg > 90.);
310	>	else if (input_orient > 0 ^ theta_in_deg < 90.) {
311	>	fputs("Cannot handle input angles on both sides of surface\n",
312	>	stderr);
313	>	exit(1);
314	>	}
315	>	if (single_plane_incident > 0)  /* check if still in plane */
316	>	single_plane_incident = (round(new_phi) == round(phi_in_deg));
317	>	else if (single_plane_incident < 0)
318	>	single_plane_incident = 1;
319	>	phi_in_deg = new_phi;
320	>	ungetc(c, fp);                  /* read actual data */
321		while (fscanf(fp, "%lf %lf %lf\n", &theta_out, &phi_out, &val) == 3) {
322		FVECT   ovec;
323		int     pos[2];
324
325	+	if (!output_orient)     /* check output orientation */
326	+	output_orient = 1 - 2*(theta_out > 90.);
327	+	else if (output_orient > 0 ^ theta_out < 90.) {
328	+	fputs("Cannot handle output angles on both sides of surface\n",
329	+	stderr);
330	+	exit(1);
331	+	}
332		ovec[2] = sin(M_PI/180.*theta_out);
333		ovec[0] = cos(M_PI/180.*phi_out) * ovec[2];
334		ovec[1] = sin(M_PI/180.*phi_out) * ovec[2];
335		ovec[2] = sqrt(1. - ovec[2]*ovec[2]);
336
337	<	if (inp_is_DSF)
338	<	val /= ovec[2]; /* convert from DSF to BSDF */
337	>	if (!inp_is_DSF)
338	>	val *= ovec[2]; /* convert from BSDF to DSF */
339
340		pos_from_vec(pos, ovec);
341
342	<	bsdf_grid[pos[0]][pos[1]].vsum += val;
343	<	bsdf_grid[pos[0]][pos[1]].nval++;
342	>	dsf_grid[pos[0]][pos[1]].vsum += val;
343	>	dsf_grid[pos[0]][pos[1]].nval++;
344		}
345		n = 0;
346		while ((c = getc(fp)) != EOF)
#	Line 241 | Line 358 | load_bsdf_meas(const char *fname)
358		static void
359		compute_radii(void)
360		{
361	<	unsigned char   fill_grid[GRIDRES][GRIDRES];
362	<	int             r, r2, lastr2;
363	<	int             i, j, jn, ii, jj, inear, jnear;
364	<	/* proceed in zig-zag */
365	<	lastr2 = GRIDRES*GRIDRES;
361	>	unsigned int    fill_grid[GRIDRES][GRIDRES];
362	>	unsigned short  fill_cnt[GRIDRES][GRIDRES];
363	>	FVECT           ovec0, ovec1;
364	>	double          ang2, lastang2;
365	>	int             r, i, j, jn, ii, jj, inear, jnear;
366	>
367	>	r = GRIDRES/2;                          /* proceed in zig-zag */
368		for (i = 0; i < GRIDRES; i++)
369		for (jn = 0; jn < GRIDRES; jn++) {
370		j = (i&1) ? jn : GRIDRES-1-jn;
371	<	if (bsdf_grid[i][j].nval)       /* find empty grid pos. */
371	>	if (dsf_grid[i][j].nval)        /* find empty grid pos. */
372		continue;
373	<	r = (int)sqrt((double)lastr2) + 2;
373	>	ovec_from_pos(ovec0, i, j);
374		inear = jnear = -1;             /* find nearest non-empty */
375	<	lastr2 = 2*GRIDRES*GRIDRES;
375	>	lastang2 = M_PI*M_PI;
376		for (ii = i-r; ii <= i+r; ii++) {
377		if (ii < 0) continue;
378		if (ii >= GRIDRES) break;
379		for (jj = j-r; jj <= j+r; jj++) {
380		if (jj < 0) continue;
381		if (jj >= GRIDRES) break;
382	<	if (!bsdf_grid[ii][jj].nval)
382	>	if (!dsf_grid[ii][jj].nval)
383		continue;
384	<	r2 = (ii-i)*(ii-i) + (jj-j)*(jj-j);
385	<	if (r2 >= lastr2)
384	>	ovec_from_pos(ovec1, ii, jj);
385	>	ang2 = 2. - 2.*DOT(ovec0,ovec1);
386	>	if (ang2 >= lastang2)
387		continue;
388	<	lastr2 = r2;
388	>	lastang2 = ang2;
389		inear = ii; jnear = jj;
390		}
391		}
392	<	/* record if > previous */
393	<	if (bsdf_grid[inear][jnear].hrad2 < lastr2)
394	<	bsdf_grid[inear][jnear].hrad2 = lastr2;
392	>	if (inear < 0) {
393	>	fputs("Could not find non-empty neighbor!\n", stderr);
394	>	exit(1);
395	>	}
396	>	ang2 = sqrt(lastang2);
397	>	r = ANG2R(ang2);                /* record if > previous */
398	>	if (r > dsf_grid[inear][jnear].crad)
399	>	dsf_grid[inear][jnear].crad = r;
400	>	/* next search radius */
401	>	r = ang2*(2.*GRIDRES/M_PI) + 3;
402		}
403	<	/* fill in others */
403	>	/* blur radii over hemisphere */
404		memset(fill_grid, 0, sizeof(fill_grid));
405	+	memset(fill_cnt, 0, sizeof(fill_cnt));
406		for (i = 0; i < GRIDRES; i++)
407		for (j = 0; j < GRIDRES; j++) {
408	<	if (!bsdf_grid[i][j].nval)
409	<	continue;
410	<	if (bsdf_grid[i][j].hrad2)
283	<	continue;
284	<	r = GRIDRES/20;
285	<	lastr2 = 2*r*r;
408	>	if (!dsf_grid[i][j].crad)
409	>	continue;               /* missing distance */
410	>	r = R2ANG(dsf_grid[i][j].crad)*(2.*RSCA*GRIDRES/M_PI);
411		for (ii = i-r; ii <= i+r; ii++) {
412		if (ii < 0) continue;
413		if (ii >= GRIDRES) break;
414		for (jj = j-r; jj <= j+r; jj++) {
415		if (jj < 0) continue;
416		if (jj >= GRIDRES) break;
417	<	if (!bsdf_grid[ii][jj].hrad2)
417	>	if ((ii-i)*(ii-i) + (jj-j)*(jj-j) > r*r)
418		continue;
419	<	r2 = (ii-i)*(ii-i) + (jj-j)*(jj-j);
420	<	if (r2 >= lastr2)
296	<	continue;
297	<	fill_grid[i][j] = bsdf_grid[ii][jj].hrad2;
298	<	lastr2 = r2;
419	>	fill_grid[ii][jj] += dsf_grid[i][j].crad;
420	>	fill_cnt[ii][jj]++;
421		}
422		}
423		}
424	+	/* copy back blurred radii */
425		for (i = 0; i < GRIDRES; i++)
426		for (j = 0; j < GRIDRES; j++)
427	<	if (fill_grid[i][j])
428	<	bsdf_grid[i][j].hrad2 = fill_grid[i][j];
427	>	if (fill_cnt[i][j])
428	>	dsf_grid[i][j].crad = fill_grid[i][j]/fill_cnt[i][j];
429		}
430
431	<	/* Cull points for more uniform distribution */
431	>	/* Cull points for more uniform distribution, leave all nval 0 or 1 */
432		static void
433		cull_values(void)
434		{
435	<	int     i, j, ii, jj, r, r2;
435	>	FVECT   ovec0, ovec1;
436	>	double  maxang, maxang2;
437	>	int     i, j, ii, jj, r;
438		/* simple greedy algorithm */
439		for (i = 0; i < GRIDRES; i++)
440		for (j = 0; j < GRIDRES; j++) {
441	<	if (!bsdf_grid[i][j].nval)
441	>	if (!dsf_grid[i][j].nval)
442		continue;
443	<	if (!bsdf_grid[i][j].hrad2)
444	<	continue;
445	<	r = (int)(2.*sqrt((double)bsdf_grid[i][j].hrad2) + .9999);
443	>	if (!dsf_grid[i][j].crad)
444	>	continue;               /* shouldn't happen */
445	>	ovec_from_pos(ovec0, i, j);
446	>	maxang = 2.*R2ANG(dsf_grid[i][j].crad);
447	>	if (maxang > ovec0[2])          /* clamp near horizon */
448	>	maxang = ovec0[2];
449	>	r = maxang*(2.*GRIDRES/M_PI) + 1;
450	>	maxang2 = maxang*maxang;
451		for (ii = i-r; ii <= i+r; ii++) {
452		if (ii < 0) continue;
453		if (ii >= GRIDRES) break;
454		for (jj = j-r; jj <= j+r; jj++) {
455		if (jj < 0) continue;
456		if (jj >= GRIDRES) break;
457	<	if (!bsdf_grid[ii][jj].nval)
457	>	if (!dsf_grid[ii][jj].nval)
458		continue;
459	<	r2 = (ii-i)*(ii-i) + (jj-j)*(jj-j);
460	<	if (!r2 | (r2 > r*r))
459	>	if ((ii == i) & (jj == j))
460	>	continue;       /* don't get self-absorbed */
461	>	ovec_from_pos(ovec1, ii, jj);
462	>	if (2. - 2.*DOT(ovec0,ovec1) >= maxang2)
463		continue;
464	<	/* absorb victim's value */
465	<	bsdf_grid[i][j].vsum += bsdf_grid[ii][jj].vsum;
466	<	bsdf_grid[i][j].nval += bsdf_grid[ii][jj].nval;
467	<	memset(&bsdf_grid[ii][jj], 0, sizeof(GRIDVAL));
464	>	/* absorb sum */
465	>	dsf_grid[i][j].vsum += dsf_grid[ii][jj].vsum;
466	>	dsf_grid[i][j].nval += dsf_grid[ii][jj].nval;
467	>	/* keep value, though */
468	>	dsf_grid[ii][jj].vsum /= (float)dsf_grid[ii][jj].nval;
469	>	dsf_grid[ii][jj].nval = 0;
470		}
471		}
472		}
473	+	/* final averaging pass */
474	+	for (i = 0; i < GRIDRES; i++)
475	+	for (j = 0; j < GRIDRES; j++)
476	+	if (dsf_grid[i][j].nval > 1) {
477	+	dsf_grid[i][j].vsum /= (float)dsf_grid[i][j].nval;
478	+	dsf_grid[i][j].nval = 1;
479	+	}
480		}
481
482	+	/* Compute (and allocate) migration price matrix for optimization */
483	+	static float *
484	+	price_routes(const RBFNODE *from_rbf, const RBFNODE *to_rbf)
485	+	{
486	+	float   *pmtx = (float *)malloc(sizeof(float) *
487	+	from_rbf->nrbf * to_rbf->nrbf);
488	+	FVECT   *vto = (FVECT *)malloc(sizeof(FVECT) * to_rbf->nrbf);
489	+	int     i, j;
490
491	+	if ((pmtx == NULL) | (vto == NULL)) {
492	+	fputs("Out of memory in migration_costs()\n", stderr);
493	+	exit(1);
494	+	}
495	+	for (j = to_rbf->nrbf; j--; )           /* save repetitive ops. */
496	+	ovec_from_pos(vto[j], to_rbf->rbfa[j].gx, to_rbf->rbfa[j].gy);
497	+
498	+	for (i = from_rbf->nrbf; i--; ) {
499	+	const double        from_ang = R2ANG(from_rbf->rbfa[i].crad);
500	+	FVECT               vfrom;
501	+	ovec_from_pos(vfrom, from_rbf->rbfa[i].gx, from_rbf->rbfa[i].gy);
502	+	for (j = to_rbf->nrbf; j--; )
503	+	pmtx[i*to_rbf->nrbf + j] = acos(DOT(vfrom, vto[j])) +
504	+	fabs(R2ANG(to_rbf->rbfa[j].crad) - from_ang);
505	+	}
506	+	free(vto);
507	+	return(pmtx);
508	+	}
509	+
510	+	/* Comparison routine needed for sorting price row */
511	+	static const float      *price_arr;
512	+	static int
513	+	msrt_cmp(const void *p1, const void *p2)
514	+	{
515	+	float   c1 = price_arr[*(const int *)p1];
516	+	float   c2 = price_arr[*(const int *)p2];
517	+
518	+	if (c1 > c2) return(1);
519	+	if (c1 < c2) return(-1);
520	+	return(0);
521	+	}
522	+
523	+	/* Compute minimum (optimistic) cost for moving the given source material */
524	+	static double
525	+	min_cost(double amt2move, const double *avail, const float *price, int n)
526	+	{
527	+	static int      *price_sort = NULL;
528	+	static int      n_alloc = 0;
529	+	double          total_cost = 0;
530	+	int             i;
531	+
532	+	if (amt2move <= FTINY)                  /* pre-emptive check */
533	+	return(0.);
534	+	if (n > n_alloc) {                      /* (re)allocate sort array */
535	+	if (n_alloc) free(price_sort);
536	+	price_sort = (int *)malloc(sizeof(int)*n);
537	+	if (price_sort == NULL) {
538	+	fputs("Out of memory in min_cost()\n", stderr);
539	+	exit(1);
540	+	}
541	+	n_alloc = n;
542	+	}
543	+	for (i = n; i--; )
544	+	price_sort[i] = i;
545	+	price_arr = price;
546	+	qsort(price_sort, n, sizeof(int), &msrt_cmp);
547	+	/* move cheapest first */
548	+	for (i = 0; i < n && amt2move > FTINY; i++) {
549	+	int     d = price_sort[i];
550	+	double  amt = (amt2move < avail[d]) ? amt2move : avail[d];
551	+
552	+	total_cost += amt * price[d];
553	+	amt2move -= amt;
554	+	}
555	+	return(total_cost);
556	+	}
557	+
558	+	/* Take a step in migration by choosing optimal bucket to transfer */
559	+	static double
560	+	migration_step(MIGRATION *mig, double *src_rem, double *dst_rem, const float *pmtx)
561	+	{
562	+	static double   *src_cost = NULL;
563	+	int             n_alloc = 0;
564	+	const double    maxamt = 2./(mtx_nrows(mig)*mtx_ncols(mig));
565	+	double          amt = 0;
566	+	struct {
567	+	int     s, d;   /* source and destination */
568	+	double  price;  /* price estimate per amount moved */
569	+	double  amt;    /* amount we can move */
570	+	} cur, best;
571	+	int             i;
572	+
573	+	if (mtx_nrows(mig) > n_alloc) {         /* allocate cost array */
574	+	if (n_alloc)
575	+	free(src_cost);
576	+	src_cost = (double *)malloc(sizeof(double)*mtx_nrows(mig));
577	+	if (src_cost == NULL) {
578	+	fputs("Out of memory in migration_step()\n", stderr);
579	+	exit(1);
580	+	}
581	+	n_alloc = mtx_nrows(mig);
582	+	}
583	+	for (i = mtx_nrows(mig); i--; )         /* starting costs for diff. */
584	+	src_cost[i] = min_cost(src_rem[i], dst_rem,
585	+	pmtx+i*mtx_ncols(mig), mtx_ncols(mig));
586	+
587	+	/* find best source & dest. */
588	+	best.s = best.d = -1; best.price = FHUGE; best.amt = 0;
589	+	for (cur.s = mtx_nrows(mig); cur.s--; ) {
590	+	const float *price = pmtx + cur.s*mtx_ncols(mig);
591	+	double      cost_others = 0;
592	+	if (src_rem[cur.s] <= FTINY)
593	+	continue;
594	+	cur.d = -1;                         /* examine cheapest dest. */
595	+	for (i = mtx_ncols(mig); i--; )
596	+	if (dst_rem[i] > FTINY &&
597	+	(cur.d < 0 || price[i] < price[cur.d]))
598	+	cur.d = i;
599	+	if (cur.d < 0)
600	+	return(.0);
601	+	if ((cur.price = price[cur.d]) >= best.price)
602	+	continue;                   /* no point checking further */
603	+	cur.amt = (src_rem[cur.s] < dst_rem[cur.d]) ?
604	+	src_rem[cur.s] : dst_rem[cur.d];
605	+	if (cur.amt > maxamt) cur.amt = maxamt;
606	+	dst_rem[cur.d] -= cur.amt;          /* add up differential costs */
607	+	for (i = mtx_nrows(mig); i--; ) {
608	+	if (i == cur.s) continue;
609	+	cost_others += min_cost(src_rem[i], dst_rem, price, mtx_ncols(mig))
610	+	- src_cost[i];
611	+	}
612	+	dst_rem[cur.d] += cur.amt;          /* undo trial move */
613	+	cur.price += cost_others/cur.amt;   /* adjust effective price */
614	+	if (cur.price < best.price)         /* are we better than best? */
615	+	best = cur;
616	+	}
617	+	if ((best.s < 0) | (best.d < 0))
618	+	return(.0);
619	+	/* make the actual move */
620	+	mig->mtx[mtx_ndx(mig,best.s,best.d)] += best.amt;
621	+	src_rem[best.s] -= best.amt;
622	+	dst_rem[best.d] -= best.amt;
623	+	return(best.amt);
624	+	}
625	+
626	+	/* Compute (and insert) migration along directed edge */
627	+	static MIGRATION *
628	+	make_migration(RBFNODE *from_rbf, RBFNODE *to_rbf)
629	+	{
630	+	const double    end_thresh = 0.02/(from_rbf->nrbf*to_rbf->nrbf);
631	+	float           *pmtx = price_routes(from_rbf, to_rbf);
632	+	MIGRATION       *newmig = (MIGRATION *)malloc(sizeof(MIGRATION) +
633	+	sizeof(float) *
634	+	(from_rbf->nrbf*to_rbf->nrbf - 1));
635	+	double          *src_rem = (double *)malloc(sizeof(double)*from_rbf->nrbf);
636	+	double          *dst_rem = (double *)malloc(sizeof(double)*to_rbf->nrbf);
637	+	double          total_rem = 1.;
638	+	int             i;
639	+
640	+	if ((newmig == NULL) | (src_rem == NULL) | (dst_rem == NULL)) {
641	+	fputs("Out of memory in make_migration()\n", stderr);
642	+	exit(1);
643	+	}
644	+	#ifdef DEBUG
645	+	{
646	+	double  theta, phi;
647	+	theta = acos(from_rbf->invec[2])*(180./M_PI);
648	+	phi = atan2(from_rbf->invec[1],from_rbf->invec[0])*(180./M_PI);
649	+	fprintf(stderr, "Building path from (theta,phi) (%d,%d) to ",
650	+	round(theta), round(phi));
651	+	theta = acos(to_rbf->invec[2])*(180./M_PI);
652	+	phi = atan2(to_rbf->invec[1],to_rbf->invec[0])*(180./M_PI);
653	+	fprintf(stderr, "(%d,%d)", round(theta), round(phi));
654	+	}
655	+	#endif
656	+	newmig->next = NULL;
657	+	newmig->rbfv[0] = from_rbf;
658	+	newmig->rbfv[1] = to_rbf;
659	+	newmig->enxt[0] = newmig->enxt[1] = NULL;
660	+	memset(newmig->mtx, 0, sizeof(float)*from_rbf->nrbf*to_rbf->nrbf);
661	+	/* starting quantities */
662	+	for (i = from_rbf->nrbf; i--; )
663	+	src_rem[i] = rbf_volume(&from_rbf->rbfa[i]) / from_rbf->vtotal;
664	+	for (i = to_rbf->nrbf; i--; )
665	+	dst_rem[i] = rbf_volume(&to_rbf->rbfa[i]) / to_rbf->vtotal;
666	+	/* move a bit at a time */
667	+	while (total_rem > end_thresh) {
668	+	total_rem -= migration_step(newmig, src_rem, dst_rem, pmtx);
669	+	#ifdef DEBUG
670	+	fputc('.', stderr);
671	+	/*XXX*/break;
672	+	#endif
673	+	}
674	+	#ifdef DEBUG
675	+	fputs("done.\n", stderr);
676	+	#endif
677	+
678	+	free(pmtx);                             /* free working arrays */
679	+	free(src_rem);
680	+	free(dst_rem);
681	+	for (i = from_rbf->nrbf; i--; ) {       /* normalize final matrix */
682	+	float       nf = rbf_volume(&from_rbf->rbfa[i]);
683	+	int         j;
684	+	if (nf <= FTINY) continue;
685	+	nf = from_rbf->vtotal / nf;
686	+	for (j = to_rbf->nrbf; j--; )
687	+	newmig->mtx[mtx_ndx(newmig,i,j)] *= nf;
688	+	}
689	+	/* insert in edge lists */
690	+	newmig->enxt[0] = from_rbf->ejl;
691	+	from_rbf->ejl = newmig;
692	+	newmig->enxt[1] = to_rbf->ejl;
693	+	to_rbf->ejl = newmig;
694	+	newmig->next = mig_list;
695	+	return(mig_list = newmig);
696	+	}
697	+
698	+	/* Get triangle surface orientation (unnormalized) */
699	+	static void
700	+	tri_orient(FVECT vres, const FVECT v1, const FVECT v2, const FVECT v3)
701	+	{
702	+	FVECT   v2minus1, v3minus2;
703	+
704	+	VSUB(v2minus1, v2, v1);
705	+	VSUB(v3minus2, v3, v2);
706	+	VCROSS(vres, v2minus1, v3minus2);
707	+	}
708	+
709	+	/* Determine if vertex order is reversed (inward normal) */
710	+	static int
711	+	is_rev_tri(const FVECT v1, const FVECT v2, const FVECT v3)
712	+	{
713	+	FVECT   tor;
714	+
715	+	tri_orient(tor, v1, v2, v3);
716	+
717	+	return(DOT(tor, v2) < 0.);
718	+	}
719	+
720	+	/* Find vertices completing triangles on either side of the given edge */
721	+	static int
722	+	get_triangles(RBFNODE *rbfv[2], const MIGRATION *mig)
723	+	{
724	+	const MIGRATION *ej, *ej2;
725	+	RBFNODE         *tv;
726	+
727	+	rbfv[0] = rbfv[1] = NULL;
728	+	if (mig == NULL)
729	+	return(0);
730	+	for (ej = mig->rbfv[0]->ejl; ej != NULL;
731	+	ej = nextedge(mig->rbfv[0],ej)) {
732	+	if (ej == mig)
733	+	continue;
734	+	tv = opp_rbf(mig->rbfv[0],ej);
735	+	for (ej2 = tv->ejl; ej2 != NULL; ej2 = nextedge(tv,ej2))
736	+	if (opp_rbf(tv,ej2) == mig->rbfv[1]) {
737	+	rbfv[is_rev_tri(mig->rbfv[0]->invec,
738	+	mig->rbfv[1]->invec,
739	+	tv->invec)] = tv;
740	+	break;
741	+	}
742	+	}
743	+	return((rbfv[0] != NULL) + (rbfv[1] != NULL));
744	+	}
745	+
746	+	/* Check if prospective vertex would create overlapping triangle */
747	+	static int
748	+	overlaps_tri(const RBFNODE *bv0, const RBFNODE *bv1, const RBFNODE *pv)
749	+	{
750	+	const MIGRATION *ej;
751	+	RBFNODE         *vother[2];
752	+	int             im_rev;
753	+	/* find shared edge in mesh */
754	+	for (ej = pv->ejl; ej != NULL; ej = nextedge(pv,ej)) {
755	+	const RBFNODE   *tv = opp_rbf(pv,ej);
756	+	if (tv == bv0) {
757	+	im_rev = is_rev_tri(ej->rbfv[0]->invec,
758	+	ej->rbfv[1]->invec, bv1->invec);
759	+	break;
760	+	}
761	+	if (tv == bv1) {
762	+	im_rev = is_rev_tri(ej->rbfv[0]->invec,
763	+	ej->rbfv[1]->invec, bv0->invec);
764	+	break;
765	+	}
766	+	}
767	+	if (!get_triangles(vother, ej))
768	+	return(0);
769	+	return(vother[im_rev] != NULL);
770	+	}
771	+
772	+	/* Find context hull vertex to complete triangle (oriented call) */
773	+	static RBFNODE *
774	+	find_chull_vert(const RBFNODE *rbf0, const RBFNODE *rbf1)
775	+	{
776	+	FVECT   vmid, vor;
777	+	RBFNODE *rbf, *rbfbest = NULL;
778	+	double  dprod2, area2, bestarea2 = FHUGE, bestdprod2 = 0.5;
779	+
780	+	VADD(vmid, rbf0->invec, rbf1->invec);
781	+	if (normalize(vmid) == 0)
782	+	return(NULL);
783	+	/* XXX exhaustive search */
784	+	for (rbf = dsf_list; rbf != NULL; rbf = rbf->next) {
785	+	if ((rbf == rbf0) | (rbf == rbf1))
786	+	continue;
787	+	tri_orient(vor, rbf0->invec, rbf1->invec, rbf->invec);
788	+	dprod2 = DOT(vor, vmid);
789	+	if (dprod2 <= FTINY)
790	+	continue;               /* wrong orientation */
791	+	area2 = DOT(vor, vor);
792	+	dprod2 *= dprod2 / area2;
793	+	if (dprod2 > bestdprod2 +
794	+	FTINY*(1 - 2*(area2 < bestarea2)) &&
795	+	!overlaps_tri(rbf0, rbf1, rbf)) {
796	+	rbfbest = rbf;
797	+	bestdprod2 = dprod2;
798	+	bestarea2 = area2;
799	+	}
800	+	}
801	+	return(rbfbest);
802	+	}
803	+
804	+	/* Create new migration edge and grow mesh recursively around it */
805	+	static void
806	+	mesh_from_edge(MIGRATION *edge)
807	+	{
808	+	MIGRATION       *ej0, *ej1;
809	+	RBFNODE         *tvert[2];
810	+	/* triangle on either side? */
811	+	get_triangles(tvert, edge);
812	+	if (tvert[0] == NULL) {                 /* grow mesh on right */
813	+	tvert[0] = find_chull_vert(edge->rbfv[0], edge->rbfv[1]);
814	+	if (tvert[0] != NULL) {
815	+	if (tvert[0] > edge->rbfv[0])
816	+	ej0 = make_migration(edge->rbfv[0], tvert[0]);
817	+	else
818	+	ej0 = make_migration(tvert[0], edge->rbfv[0]);
819	+	if (tvert[0] > edge->rbfv[1])
820	+	ej1 = make_migration(edge->rbfv[1], tvert[0]);
821	+	else
822	+	ej1 = make_migration(tvert[0], edge->rbfv[1]);
823	+	mesh_from_edge(ej0);
824	+	mesh_from_edge(ej1);
825	+	}
826	+	}
827	+	if (tvert[1] == NULL) {                 /* grow mesh on left */
828	+	tvert[1] = find_chull_vert(edge->rbfv[1], edge->rbfv[0]);
829	+	if (tvert[1] != NULL) {
830	+	if (tvert[1] > edge->rbfv[0])
831	+	ej0 = make_migration(edge->rbfv[0], tvert[1]);
832	+	else
833	+	ej0 = make_migration(tvert[1], edge->rbfv[0]);
834	+	if (tvert[1] > edge->rbfv[1])
835	+	ej1 = make_migration(edge->rbfv[1], tvert[1]);
836	+	else
837	+	ej1 = make_migration(tvert[1], edge->rbfv[1]);
838	+	mesh_from_edge(ej0);
839	+	mesh_from_edge(ej1);
840	+	}
841	+	}
842	+	}
843	+
844	+	#ifdef DEBUG
845	+	#include "random.h"
846	+	#include "bmpfile.h"
847	+	/* Hash pointer to byte value */
848	+	static int
849	+	byte_hash(const void *p)
850	+	{
851	+	size_t  h = (size_t)p;
852	+	h ^= (size_t)p >> 8;
853	+	h ^= (size_t)p >> 16;
854	+	h ^= (size_t)p >> 24;
855	+	return(h & 0xff);
856	+	}
857	+	/* Write out BMP image showing edges */
858	+	static void
859	+	write_edge_image(const char *fname)
860	+	{
861	+	BMPHeader       *hdr = BMPmappedHeader(GRIDRES, GRIDRES, 0, 256);
862	+	BMPWriter       *wtr;
863	+	int             i, j;
864	+
865	+	fprintf(stderr, "Writing incident mesh drawing to '%s'\n", fname);
866	+	hdr->compr = BI_RLE8;
867	+	for (i = 256; --i; ) {                  /* assign random color map */
868	+	hdr->palette[i].r = random() & 0xff;
869	+	hdr->palette[i].r = random() & 0xff;
870	+	hdr->palette[i].r = random() & 0xff;
871	+	}
872	+	hdr->palette[0].r = hdr->palette[0].g = hdr->palette[0].b = 0;
873	+	/* open output */
874	+	wtr = BMPopenOutputFile(fname, hdr);
875	+	if (wtr == NULL) {
876	+	free(hdr);
877	+	return;
878	+	}
879	+	for (i = 0; i < GRIDRES; i++) {         /* write scanlines */
880	+	for (j = 0; j < GRIDRES; j++)
881	+	wtr->scanline[j] = byte_hash(mig_grid[i][j]);
882	+	if (BMPwriteScanline(wtr) != BIR_OK)
883	+	break;
884	+	}
885	+	BMPcloseOutput(wtr);                    /* close & clean up */
886	+	}
887	+	#endif
888	+
889	+	/* Draw edge list into mig_grid array */
890	+	static void
891	+	draw_edges()
892	+	{
893	+	int             nnull = 0, ntot = 0;
894	+	MIGRATION       *ej;
895	+	int             p0[2], p1[2];
896	+
897	+	/* memset(mig_grid, 0, sizeof(mig_grid)); */
898	+	for (ej = mig_list; ej != NULL; ej = ej->next) {
899	+	++ntot;
900	+	pos_from_vec(p0, ej->rbfv[0]->invec);
901	+	pos_from_vec(p1, ej->rbfv[1]->invec);
902	+	if ((p0[0] == p1[0]) & (p0[1] == p1[1])) {
903	+	++nnull;
904	+	mig_grid[p0[0]][p0[1]] = ej;
905	+	continue;
906	+	}
907	+	if (abs(p1[0]-p0[0]) > abs(p1[1]-p0[1])) {
908	+	const int       xstep = 2*(p1[0] > p0[0]) - 1;
909	+	const double    ystep = (double)((p1[1]-p0[1])*xstep) /
910	+	(double)(p1[0]-p0[0]);
911	+	int             x;
912	+	double          y;
913	+	for (x = p0[0], y = p0[1]+.5; x != p1[0];
914	+	x += xstep, y += ystep)
915	+	mig_grid[x][(int)y] = ej;
916	+	mig_grid[x][(int)y] = ej;
917	+	} else {
918	+	const int       ystep = 2*(p1[1] > p0[1]) - 1;
919	+	const double    xstep = (double)((p1[0]-p0[0])*ystep) /
920	+	(double)(p1[1]-p0[1]);
921	+	int             y;
922	+	double          x;
923	+	for (y = p0[1], x = p0[0]+.5; y != p1[1];
924	+	y += ystep, x += xstep)
925	+	mig_grid[(int)x][y] = ej;
926	+	mig_grid[(int)x][y] = ej;
927	+	}
928	+	}
929	+	if (nnull)
930	+	fprintf(stderr, "Warning: %d of %d edges are null\n",
931	+	nnull, ntot);
932	+	#ifdef DEBUG
933	+	write_edge_image("bsdf_edges.bmp");
934	+	#endif
935	+	}
936	+
937	+	/* Build our triangle mesh from recorded RBFs */
938	+	static void
939	+	build_mesh()
940	+	{
941	+	double          best2 = M_PI*M_PI;
942	+	RBFNODE         *shrt_edj[2];
943	+	RBFNODE         *rbf0, *rbf1;
944	+	/* check if isotropic */
945	+	if (single_plane_incident) {
946	+	for (rbf0 = dsf_list; rbf0 != NULL; rbf0 = rbf0->next)
947	+	if (rbf0->next != NULL)
948	+	make_migration(rbf0, rbf0->next);
949	+	return;
950	+	}
951	+	/* start w/ shortest edge */
952	+	shrt_edj[0] = shrt_edj[1] = NULL;
953	+	for (rbf0 = dsf_list; rbf0 != NULL; rbf0 = rbf0->next)
954	+	for (rbf1 = rbf0->next; rbf1 != NULL; rbf1 = rbf1->next) {
955	+	double  dist2 = 2. - 2.*DOT(rbf0->invec,rbf1->invec);
956	+	if (dist2 < best2) {
957	+	shrt_edj[0] = rbf0;
958	+	shrt_edj[1] = rbf1;
959	+	best2 = dist2;
960	+	}
961	+	}
962	+	if (shrt_edj[0] == NULL) {
963	+	fputs("Cannot find shortest edge\n", stderr);
964	+	exit(1);
965	+	}
966	+	/* build mesh from this edge */
967	+	if (shrt_edj[0] < shrt_edj[1])
968	+	mesh_from_edge(make_migration(shrt_edj[0], shrt_edj[1]));
969	+	else
970	+	mesh_from_edge(make_migration(shrt_edj[1], shrt_edj[0]));
971	+	/* draw edge list into grid */
972	+	draw_edges();
973	+	}
974	+
975	+	/* Identify enclosing triangle for this position (flood fill raster check) */
976	+	static int
977	+	identify_tri(MIGRATION *miga[3], unsigned char vmap[GRIDRES][(GRIDRES+7)/8],
978	+	int px, int py)
979	+	{
980	+	const int       btest = 1<<(py&07);
981	+
982	+	if (vmap[px][py>>3] & btest)            /* already visited here? */
983	+	return(1);
984	+	/* else mark it */
985	+	vmap[px][py>>3] |= btest;
986	+
987	+	if (mig_grid[px][py] != NULL) {         /* are we on an edge? */
988	+	int     i;
989	+	for (i = 0; i < 3; i++) {
990	+	if (miga[i] == mig_grid[px][py])
991	+	return(1);
992	+	if (miga[i] != NULL)
993	+	continue;
994	+	miga[i] = mig_grid[px][py];
995	+	return(1);
996	+	}
997	+	return(0);                      /* outside triangle! */
998	+	}
999	+	/* check neighbors (flood) */
1000	+	if (px > 0 && !identify_tri(miga, vmap, px-1, py))
1001	+	return(0);
1002	+	if (px < GRIDRES-1 && !identify_tri(miga, vmap, px+1, py))
1003	+	return(0);
1004	+	if (py > 0 && !identify_tri(miga, vmap, px, py-1))
1005	+	return(0);
1006	+	if (py < GRIDRES-1 && !identify_tri(miga, vmap, px, py+1))
1007	+	return(0);
1008	+	return(1);                              /* this neighborhood done */
1009	+	}
1010	+
1011	+	/* Find edge(s) for interpolating the given incident vector */
1012	+	static int
1013	+	get_interp(MIGRATION *miga[3], const FVECT invec)
1014	+	{
1015	+	miga[0] = miga[1] = miga[2] = NULL;
1016	+	if (single_plane_incident) {            /* isotropic BSDF? */
1017	+	RBFNODE *rbf;                   /* find edge we're on */
1018	+	for (rbf = dsf_list; rbf != NULL; rbf = rbf->next) {
1019	+	if (input_orient*rbf->invec[2] < input_orient*invec[2])
1020	+	break;
1021	+	if (rbf->next != NULL &&
1022	+	input_orient*rbf->next->invec[2] <
1023	+	input_orient*invec[2]) {
1024	+	for (miga[0] = rbf->ejl; miga[0] != NULL;
1025	+	miga[0] = nextedge(rbf,miga[0]))
1026	+	if (opp_rbf(rbf,miga[0]) == rbf->next)
1027	+	return(1);
1028	+	break;
1029	+	}
1030	+	}
1031	+	return(0);                      /* outside range! */
1032	+	}
1033	+	{                                       /* else use triangle mesh */
1034	+	unsigned char   floodmap[GRIDRES][(GRIDRES+7)/8];
1035	+	int             pstart[2];
1036	+
1037	+	pos_from_vec(pstart, invec);
1038	+	memset(floodmap, 0, sizeof(floodmap));
1039	+	/* call flooding function */
1040	+	if (!identify_tri(miga, floodmap, pstart[0], pstart[1]))
1041	+	return(0);              /* outside mesh */
1042	+	if ((miga[0] == NULL) | (miga[2] == NULL))
1043	+	return(0);              /* should never happen */
1044	+	if (miga[1] == NULL)
1045	+	return(1);              /* on edge */
1046	+	return(3);                      /* else in triangle */
1047	+	}
1048	+	}
1049	+
1050	+	/* Advect and allocate new RBF along edge */
1051	+	static RBFNODE *
1052	+	e_advect_rbf(const MIGRATION *mig, const FVECT invec)
1053	+	{
1054	+	RBFNODE         *rbf;
1055	+	int             n, i, j;
1056	+	double          t, full_dist;
1057	+	/* get relative position */
1058	+	t = acos(DOT(invec, mig->rbfv[0]->invec));
1059	+	if (t < M_PI/GRIDRES) {                 /* near first DSF */
1060	+	n = sizeof(RBFNODE) + sizeof(RBFVAL)*(mig->rbfv[0]->nrbf-1);
1061	+	rbf = (RBFNODE *)malloc(n);
1062	+	if (rbf == NULL)
1063	+	goto memerr;
1064	+	memcpy(rbf, mig->rbfv[0], n);   /* just duplicate */
1065	+	return(rbf);
1066	+	}
1067	+	full_dist = acos(DOT(mig->rbfv[0]->invec, mig->rbfv[1]->invec));
1068	+	if (t > full_dist-M_PI/GRIDRES) {       /* near second DSF */
1069	+	n = sizeof(RBFNODE) + sizeof(RBFVAL)*(mig->rbfv[1]->nrbf-1);
1070	+	rbf = (RBFNODE *)malloc(n);
1071	+	if (rbf == NULL)
1072	+	goto memerr;
1073	+	memcpy(rbf, mig->rbfv[1], n);   /* just duplicate */
1074	+	return(rbf);
1075	+	}
1076	+	t /= full_dist;
1077	+	n = 0;                                  /* count migrating particles */
1078	+	for (i = 0; i < mtx_nrows(mig); i++)
1079	+	for (j = 0; j < mtx_ncols(mig); j++)
1080	+	n += (mig->mtx[mtx_ndx(mig,i,j)] > FTINY);
1081	+	#ifdef DEBUG
1082	+	fprintf(stderr, "Input RBFs have %d, %d nodes -> output has %d\n",
1083	+	mig->rbfv[0]->nrbf, mig->rbfv[1]->nrbf, n);
1084	+	#endif
1085	+	rbf = (RBFNODE *)malloc(sizeof(RBFNODE) + sizeof(RBFVAL)*(n-1));
1086	+	if (rbf == NULL)
1087	+	goto memerr;
1088	+	rbf->next = NULL; rbf->ejl = NULL;
1089	+	VCOPY(rbf->invec, invec);
1090	+	rbf->nrbf = n;
1091	+	rbf->vtotal = 1.-t + t*mig->rbfv[1]->vtotal/mig->rbfv[0]->vtotal;
1092	+	n = 0;                                  /* advect RBF lobes */
1093	+	for (i = 0; i < mtx_nrows(mig); i++) {
1094	+	const RBFVAL        *rbf0i = &mig->rbfv[0]->rbfa[i];
1095	+	const float         peak0 = rbf0i->peak;
1096	+	const double        rad0 = R2ANG(rbf0i->crad);
1097	+	FVECT               v0;
1098	+	float               mv;
1099	+	ovec_from_pos(v0, rbf0i->gx, rbf0i->gy);
1100	+	for (j = 0; j < mtx_ncols(mig); j++)
1101	+	if ((mv = mig->mtx[mtx_ndx(mig,i,j)]) > FTINY) {
1102	+	const RBFVAL    *rbf1j = &mig->rbfv[1]->rbfa[j];
1103	+	double          rad1 = R2ANG(rbf1j->crad);
1104	+	FVECT           v;
1105	+	int             pos[2];
1106	+	rbf->rbfa[n].peak = peak0 * mv * rbf->vtotal;
1107	+	rbf->rbfa[n].crad = ANG2R(sqrt(rad0*rad0*(1.-t) +
1108	+	rad1*rad1*t));
1109	+	ovec_from_pos(v, rbf1j->gx, rbf1j->gy);
1110	+	geodesic(v, v0, v, t, GEOD_REL);
1111	+	pos_from_vec(pos, v);
1112	+	rbf->rbfa[n].gx = pos[0];
1113	+	rbf->rbfa[n].gy = pos[1];
1114	+	++n;
1115	+	}
1116	+	}
1117	+	rbf->vtotal *= mig->rbfv[0]->vtotal;    /* turn ratio into actual */
1118	+	return(rbf);
1119	+	memerr:
1120	+	fputs("Out of memory in e_advect_rbf()\n", stderr);
1121	+	exit(1);
1122	+	return(NULL);   /* pro forma return */
1123	+	}
1124	+
1125	+	/* Insert vertex in ordered list */
1126	+	static void
1127	+	insert_vert(RBFNODE **vlist, RBFNODE *v)
1128	+	{
1129	+	int     i, j;
1130	+
1131	+	for (i = 0; vlist[i] != NULL; i++) {
1132	+	if (v == vlist[i])
1133	+	return;
1134	+	if (v < vlist[i])
1135	+	break;
1136	+	}
1137	+	for (j = i; vlist[j] != NULL; j++)
1138	+	;
1139	+	while (j > i) {
1140	+	vlist[j] = vlist[j-1];
1141	+	--j;
1142	+	}
1143	+	vlist[i] = v;
1144	+	}
1145	+
1146	+	/* Sort triangle edges in standard order */
1147	+	static void
1148	+	order_triangle(MIGRATION *miga[3])
1149	+	{
1150	+	RBFNODE         *vert[4];
1151	+	MIGRATION       *ord[3];
1152	+	int             i;
1153	+	/* order vertices, first */
1154	+	memset(vert, 0, sizeof(vert));
1155	+	for (i = 0; i < 3; i++) {
1156	+	insert_vert(vert, miga[i]->rbfv[0]);
1157	+	insert_vert(vert, miga[i]->rbfv[1]);
1158	+	}
1159	+	/* identify edge 0 */
1160	+	for (i = 0; i < 3; i++)
1161	+	if (miga[i]->rbfv[0] == vert[0] &&
1162	+	miga[i]->rbfv[1] == vert[1]) {
1163	+	ord[0] = miga[i];
1164	+	break;
1165	+	}
1166	+	/* identify edge 1 */
1167	+	for (i = 0; i < 3; i++)
1168	+	if (miga[i]->rbfv[0] == vert[1] &&
1169	+	miga[i]->rbfv[1] == vert[2]) {
1170	+	ord[1] = miga[i];
1171	+	break;
1172	+	}
1173	+	/* identify edge 2 */
1174	+	for (i = 0; i < 3; i++)
1175	+	if (miga[i]->rbfv[0] == vert[0] &&
1176	+	miga[i]->rbfv[1] == vert[2]) {
1177	+	ord[2] = miga[i];
1178	+	break;
1179	+	}
1180	+	miga[0] = ord[0]; miga[1] = ord[1]; miga[2] = ord[2];
1181	+	}
1182	+
1183	+	/* Partially advect between recorded incident angles and allocate new RBF */
1184	+	static RBFNODE *
1185	+	advect_rbf(const FVECT invec)
1186	+	{
1187	+	MIGRATION       *miga[3];
1188	+	RBFNODE         *rbf;
1189	+	float           mbfact, mcfact;
1190	+	int             n, i, j, k;
1191	+	FVECT           v0, v1, v2;
1192	+	double          s, t;
1193	+
1194	+	if (!get_interp(miga, invec))           /* can't interpolate? */
1195	+	return(NULL);
1196	+	if (miga[1] == NULL)                    /* advect along edge? */
1197	+	return(e_advect_rbf(miga[0], invec));
1198	+	/* put in standard order */
1199	+	order_triangle(miga);
1200	+	#ifdef DEBUG
1201	+	if (miga[0]->rbfv[0] != miga[2]->rbfv[0] |
1202	+	miga[0]->rbfv[1] != miga[1]->rbfv[0] |
1203	+	miga[1]->rbfv[1] != miga[2]->rbfv[1]) {
1204	+	fputs("Triangle vertex screw-up!\n", stderr);
1205	+	exit(1);
1206	+	}
1207	+	#endif
1208	+	/* figure out position */
1209	+	fcross(v0, miga[2]->rbfv[0]->invec, miga[2]->rbfv[1]->invec);
1210	+	normalize(v0);
1211	+	fcross(v2, miga[1]->rbfv[0]->invec, miga[1]->rbfv[1]->invec);
1212	+	normalize(v2);
1213	+	fcross(v1, invec, miga[1]->rbfv[1]->invec);
1214	+	normalize(v1);
1215	+	s = acos(DOT(v0,v1)) / acos(DOT(v0,v2));
1216	+	geodesic(v1, miga[0]->rbfv[0]->invec, miga[0]->rbfv[1]->invec,
1217	+	s, GEOD_REL);
1218	+	t = acos(DOT(v1,invec)) / acos(DOT(v1,miga[1]->rbfv[1]->invec));
1219	+	n = 0;                                  /* count migrating particles */
1220	+	for (i = 0; i < mtx_nrows(miga[0]); i++)
1221	+	for (j = 0; j < mtx_ncols(miga[0]); j++)
1222	+	for (k = (miga[0]->mtx[mtx_ndx(miga[0],i,j)] > FTINY) *
1223	+	mtx_ncols(miga[2]); k--; )
1224	+	n += (miga[2]->mtx[mtx_ndx(miga[2],i,k)] > FTINY &&
1225	+	miga[1]->mtx[mtx_ndx(miga[1],j,k)] > FTINY);
1226	+	#ifdef DEBUG
1227	+	fprintf(stderr, "Input RBFs have %d, %d, %d nodes -> output has %d\n",
1228	+	miga[0]->rbfv[0]->nrbf, miga[0]->rbfv[1]->nrbf,
1229	+	miga[2]->rbfv[1]->nrbf, n);
1230	+	#endif
1231	+	rbf = (RBFNODE *)malloc(sizeof(RBFNODE) + sizeof(RBFVAL)*(n-1));
1232	+	if (rbf == NULL) {
1233	+	fputs("Out of memory in advect_rbf()\n", stderr);
1234	+	exit(1);
1235	+	}
1236	+	rbf->next = NULL; rbf->ejl = NULL;
1237	+	VCOPY(rbf->invec, invec);
1238	+	rbf->nrbf = n;
1239	+	n = 0;                                  /* compute RBF lobes */
1240	+	mbfact = s * miga[0]->rbfv[1]->vtotal/miga[0]->rbfv[0]->vtotal *
1241	+	(1.-t + t*miga[1]->rbfv[1]->vtotal/miga[1]->rbfv[0]->vtotal);
1242	+	mcfact = (1.-s) *
1243	+	(1.-t + t*miga[2]->rbfv[1]->vtotal/miga[2]->rbfv[0]->vtotal);
1244	+	for (i = 0; i < mtx_nrows(miga[0]); i++) {
1245	+	const RBFVAL        *rbf0i = &miga[0]->rbfv[0]->rbfa[i];
1246	+	const float         w0i = rbf0i->peak;
1247	+	const double        rad0i = R2ANG(rbf0i->crad);
1248	+	ovec_from_pos(v0, rbf0i->gx, rbf0i->gy);
1249	+	for (j = 0; j < mtx_ncols(miga[0]); j++) {
1250	+	const float     ma = miga[0]->mtx[mtx_ndx(miga[0],i,j)];
1251	+	const RBFVAL    *rbf1j;
1252	+	double          rad1j, srad2;
1253	+	if (ma <= FTINY)
1254	+	continue;
1255	+	rbf1j = &miga[0]->rbfv[1]->rbfa[j];
1256	+	rad1j = R2ANG(rbf1j->crad);
1257	+	srad2 = (1.-s)*(1.-t)*rad0i*rad0i + s*(1.-t)*rad1j*rad1j;
1258	+	ovec_from_pos(v1, rbf1j->gx, rbf1j->gy);
1259	+	geodesic(v1, v0, v1, s, GEOD_REL);
1260	+	for (k = 0; k < mtx_ncols(miga[2]); k++) {
1261	+	float               mb = miga[1]->mtx[mtx_ndx(miga[1],j,k)];
1262	+	float               mc = miga[2]->mtx[mtx_ndx(miga[2],i,k)];
1263	+	const RBFVAL        *rbf2k;
1264	+	double              rad2k;
1265	+	FVECT               vout;
1266	+	int                 pos[2];
1267	+	if ((mb <= FTINY) | (mc <= FTINY))
1268	+	continue;
1269	+	rbf2k = &miga[2]->rbfv[1]->rbfa[k];
1270	+	rbf->rbfa[n].peak = w0i * ma * (mb*mbfact + mc*mcfact);
1271	+	rad2k = R2ANG(rbf2k->crad);
1272	+	rbf->rbfa[n].crad = ANG2R(sqrt(srad2 + t*rad2k*rad2k));
1273	+	ovec_from_pos(v2, rbf2k->gx, rbf2k->gy);
1274	+	geodesic(vout, v1, v2, t, GEOD_REL);
1275	+	pos_from_vec(pos, vout);
1276	+	rbf->rbfa[n].gx = pos[0];
1277	+	rbf->rbfa[n].gy = pos[1];
1278	+	++n;
1279	+	}
1280	+	}
1281	+	}
1282	+	rbf->vtotal = miga[0]->rbfv[0]->vtotal * (mbfact + mcfact);
1283	+	return(rbf);
1284	+	}
1285	+
1286	+	/* Interpolate and output isotropic BSDF data */
1287	+	static void
1288	+	interp_isotropic()
1289	+	{
1290	+	const int       sqres = 1<<samp_order;
1291	+	FILE            *ofp = NULL;
1292	+	char            cmd[128];
1293	+	int             ix, ox, oy;
1294	+	FVECT           ivec, ovec;
1295	+	double          bsdf;
1296	+	#if DEBUG
1297	+	fprintf(stderr, "Writing isotropic order %d ", samp_order);
1298	+	if (pctcull >= 0) fprintf(stderr, "data with %d%% culling\n", pctcull);
1299	+	else fputs("raw data\n", stderr);
1300	+	#endif
1301	+	if (pctcull >= 0) {                     /* begin output */
1302	+	sprintf(cmd, "rttree_reduce -h -a -fd -r 3 -t %d -g %d",
1303	+	pctcull, samp_order);
1304	+	fflush(stdout);
1305	+	ofp = popen(cmd, "w");
1306	+	if (ofp == NULL) {
1307	+	fputs("Cannot create pipe for rttree_reduce\n", stderr);
1308	+	exit(1);
1309	+	}
1310	+	} else
1311	+	fputs("{\n", stdout);
1312	+	/* run through directions */
1313	+	for (ix = 0; ix < sqres/2; ix++) {
1314	+	RBFNODE *rbf;
1315	+	SDsquare2disk(ivec, (ix+.5)/sqres, .5);
1316	+	ivec[2] = input_orient *
1317	+	sqrt(1. - ivec[0]*ivec[0] - ivec[1]*ivec[1]);
1318	+	rbf = advect_rbf(ivec);
1319	+	for (ox = 0; ox < sqres; ox++)
1320	+	for (oy = 0; oy < sqres; oy++) {
1321	+	SDsquare2disk(ovec, (ox+.5)/sqres, (oy+.5)/sqres);
1322	+	ovec[2] = output_orient *
1323	+	sqrt(1. - ovec[0]*ovec[0] - ovec[1]*ovec[1]);
1324	+	bsdf = eval_rbfrep(rbf, ovec) / fabs(ovec[2]);
1325	+	if (pctcull >= 0)
1326	+	fwrite(&bsdf, sizeof(bsdf), 1, ofp);
1327	+	else
1328	+	printf("\t%.3e\n", bsdf);
1329	+	}
1330	+	free(rbf);
1331	+	}
1332	+	if (pctcull >= 0) {                     /* finish output */
1333	+	if (pclose(ofp)) {
1334	+	fprintf(stderr, "Error running '%s'\n", cmd);
1335	+	exit(1);
1336	+	}
1337	+	} else {
1338	+	for (ix = sqres*sqres*sqres/2; ix--; )
1339	+	fputs("\t0\n", stdout);
1340	+	fputs("}\n", stdout);
1341	+	}
1342	+	}
1343	+
1344	+	/* Interpolate and output anisotropic BSDF data */
1345	+	static void
1346	+	interp_anisotropic()
1347	+	{
1348	+	const int       sqres = 1<<samp_order;
1349	+	FILE            *ofp = NULL;
1350	+	char            cmd[128];
1351	+	int             ix, iy, ox, oy;
1352	+	FVECT           ivec, ovec;
1353	+	double          bsdf;
1354	+	#if DEBUG
1355	+	fprintf(stderr, "Writing anisotropic order %d ", samp_order);
1356	+	if (pctcull >= 0) fprintf(stderr, "data with %d%% culling\n", pctcull);
1357	+	else fputs("raw data\n", stderr);
1358	+	#endif
1359	+	if (pctcull >= 0) {                     /* begin output */
1360	+	sprintf(cmd, "rttree_reduce -h -a -fd -r 4 -t %d -g %d",
1361	+	pctcull, samp_order);
1362	+	fflush(stdout);
1363	+	ofp = popen(cmd, "w");
1364	+	if (ofp == NULL) {
1365	+	fputs("Cannot create pipe for rttree_reduce\n", stderr);
1366	+	exit(1);
1367	+	}
1368	+	} else
1369	+	fputs("{\n", stdout);
1370	+	/* run through directions */
1371	+	for (ix = 0; ix < sqres; ix++)
1372	+	for (iy = 0; iy < sqres; iy++) {
1373	+	RBFNODE *rbf;
1374	+	SDsquare2disk(ivec, (ix+.5)/sqres, (iy+.5)/sqres);
1375	+	ivec[2] = input_orient *
1376	+	sqrt(1. - ivec[0]*ivec[0] - ivec[1]*ivec[1]);
1377	+	rbf = advect_rbf(ivec);
1378	+	for (ox = 0; ox < sqres; ox++)
1379	+	for (oy = 0; oy < sqres; oy++) {
1380	+	SDsquare2disk(ovec, (ox+.5)/sqres, (oy+.5)/sqres);
1381	+	ovec[2] = output_orient *
1382	+	sqrt(1. - ovec[0]*ovec[0] - ovec[1]*ovec[1]);
1383	+	bsdf = eval_rbfrep(rbf, ovec) / fabs(ovec[2]);
1384	+	if (pctcull >= 0)
1385	+	fwrite(&bsdf, sizeof(bsdf), 1, ofp);
1386	+	else
1387	+	printf("\t%.3e\n", bsdf);
1388	+	}
1389	+	free(rbf);
1390	+	}
1391	+	if (pctcull >= 0) {                     /* finish output */
1392	+	if (pclose(ofp)) {
1393	+	fprintf(stderr, "Error running '%s'\n", cmd);
1394	+	exit(1);
1395	+	}
1396	+	} else
1397	+	fputs("}\n", stdout);
1398	+	}
1399	+
1400		#if 1
1401	+	/* Read in BSDF files and interpolate as tensor tree representation */
1402	+	int
1403	+	main(int argc, char *argv[])
1404	+	{
1405	+	RBFNODE         *rbf;
1406	+	double          bsdf;
1407	+	int             i;
1408	+
1409	+	progname = argv[0];
1410	+	if (argc > 2 && !strcmp(argv[1], "-t")) {
1411	+	pctcull = atoi(argv[2]);
1412	+	argv += 2; argc -= 2;
1413	+	}
1414	+	if (argc < 3) {
1415	+	fprintf(stderr,
1416	+	"Usage: %s [-t pctcull] meas1.dat meas2.dat .. > bsdf.xml\n",
1417	+	progname);
1418	+	return(1);
1419	+	}
1420	+	for (i = 1; i < argc; i++) {            /* compile measurements */
1421	+	if (!load_pabopto_meas(argv[i]))
1422	+	return(1);
1423	+	compute_radii();
1424	+	cull_values();
1425	+	make_rbfrep();
1426	+	}
1427	+	build_mesh();                           /* create interpolation */
1428	+	/* xml_prologue();                              /* start XML output */
1429	+	if (single_plane_incident)              /* resample dist. */
1430	+	interp_isotropic();
1431	+	else
1432	+	interp_anisotropic();
1433	+	/* xml_epilogue();                              /* finish XML output */
1434	+	return(0);
1435	+	}
1436	+	#else
1437		/* Test main produces a Radiance model from the given input file */
1438		int
1439		main(int argc, char *argv[])
#	Line 354 | Line 1448 | main(int argc, char *argv[])
1448		fprintf(stderr, "Usage: %s input.dat > output.rad\n", argv[0]);
1449		return(1);
1450		}
1451	<	if (!load_bsdf_meas(argv[1]))
1451	>	if (!load_pabopto_meas(argv[1]))
1452		return(1);
359	–	/* produce spheres at meas. */
360	–	puts("void plastic orange\n0\n0\n5 .6 .4 .01 .04 .08\n");
361	–	n = 0;
362	–	for (i = 0; i < GRIDRES; i++)
363	–	for (j = 0; j < GRIDRES; j++)
364	–	if (bsdf_grid[i][j].nval) {
365	–	double  bsdf = bsdf_grid[i][j].vsum /
366	–	(double)bsdf_grid[i][j].nval;
367	–	FVECT   dir;
1453
369	–	vec_from_pos(dir, i, j);
370	–	printf("orange sphere s%04d\n0\n0\n", ++n);
371	–	printf("4 %.6g %.6g %.6g .0015\n\n",
372	–	dir[0]*bsdf, dir[1]*bsdf, dir[2]*bsdf);
373	–	}
1454		compute_radii();
1455		cull_values();
1456	<	/* highlight chosen values */
1456	>	make_rbfrep();
1457	>	/* produce spheres at meas. */
1458	>	puts("void plastic yellow\n0\n0\n5 .6 .4 .01 .04 .08\n");
1459		puts("void plastic pink\n0\n0\n5 .5 .05 .9 .04 .08\n");
1460		n = 0;
1461		for (i = 0; i < GRIDRES; i++)
1462		for (j = 0; j < GRIDRES; j++)
1463	<	if (bsdf_grid[i][j].nval) {
1464	<	bsdf = bsdf_grid[i][j].vsum /
1465	<	(double)bsdf_grid[i][j].nval;
1466	<	vec_from_pos(dir, i, j);
1467	<	printf("pink cone c%04d\n0\n0\n8\n", ++n);
1468	<	printf("\t%.6g %.6g %.6g\n",
1463	>	if (dsf_grid[i][j].vsum > .0f) {
1464	>	ovec_from_pos(dir, i, j);
1465	>	bsdf = dsf_grid[i][j].vsum / dir[2];
1466	>	if (dsf_grid[i][j].nval) {
1467	>	printf("pink cone c%04d\n0\n0\n8\n", ++n);
1468	>	printf("\t%.6g %.6g %.6g\n",
1469		dir[0]*bsdf, dir[1]*bsdf, dir[2]*bsdf);
1470	<	printf("\t%.6g %.6g %.6g\n",
1470	>	printf("\t%.6g %.6g %.6g\n",
1471		dir[0]*(bsdf+.005), dir[1]*(bsdf+.005),
1472		dir[2]*(bsdf+.005));
1473	<	puts("\t.003\t0\n");
1473	>	puts("\t.003\t0\n");
1474	>	} else {
1475	>	ovec_from_pos(dir, i, j);
1476	>	printf("yellow sphere s%04d\n0\n0\n", ++n);
1477	>	printf("4 %.6g %.6g %.6g .0015\n\n",
1478	>	dir[0]*bsdf, dir[1]*bsdf, dir[2]*bsdf);
1479	>	}
1480		}
1481		/* output continuous surface */
394	–	make_rbfrep();
1482		puts("void trans tgreen\n0\n0\n7 .7 1 .7 .04 .04 .9 .9\n");
1483		fflush(stdout);
1484	<	sprintf(buf, "gensurf tgreen bsdf - - - %d %d", GRIDRES, GRIDRES);
1484	>	sprintf(buf, "gensurf tgreen bsdf - - - %d %d", GRIDRES-1, GRIDRES-1);
1485		pfp = popen(buf, "w");
1486		if (pfp == NULL) {
1487		fputs(buf, stderr);
#	Line 403 | Line 1490 | main(int argc, char *argv[])
1490		}
1491		for (i = 0; i < GRIDRES; i++)
1492		for (j = 0; j < GRIDRES; j++) {
1493	<	vec_from_pos(dir, i, j);
1494	<	bsdf = eval_rbfrep(bsdf_list, dir);
1493	>	ovec_from_pos(dir, i, j);
1494	>	bsdf = eval_rbfrep(dsf_list, dir) / dir[2];
1495		fprintf(pfp, "%.8e %.8e %.8e\n",
1496		dir[0]*bsdf, dir[1]*bsdf, dir[2]*bsdf);
1497		}

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines