--- ray/src/rt/aniso.c	1992/03/03 16:20:00	2.11
+++ ray/src/rt/aniso.c	1994/12/21 09:51:44	2.28
@@ -19,9 +19,13 @@ static char SCCSid[] = "$SunId$ LBL";
 extern double  specthresh;		/* specular sampling threshold */
 extern double  specjitter;		/* specular sampling jitter */
 
+extern int  backvis;			/* back faces visible? */
+
+static  agaussamp(), getacoords();
+
 /*
- *	This anisotropic reflection model uses a variant on the
- *  exponential Gaussian used in normal.c.
+ *	This routine implements the anisotropic Gaussian
+ *  model described by Ward in Siggraph `92 article.
  *	We orient the surface towards the incoming ray, so a single
  *  surface can be used to represent an infinitely thin object.
  *
@@ -34,8 +38,6 @@ extern double  specjitter;		/* specular sampling jitte
  *  8  red 	grn	blu	rspec	u-rough	v-rough	trans	tspec
  */
 
-#define  BSPEC(m)	(6.0)		/* specularity parameter b */
-
 				/* specularity flags */
 #define  SP_REFL	01		/* has reflected specular component */
 #define  SP_TRAN	02		/* has transmitted specular */
@@ -70,7 +72,7 @@ FVECT  ldir;			/* light source direction */
 double  omega;			/* light source size */
 {
 	double  ldot;
-	double  dtmp, dtmp2;
+	double  dtmp, dtmp1, dtmp2;
 	FVECT  h;
 	double  au2, av2;
 	COLOR  ctmp;
@@ -103,25 +105,26 @@ double  omega;			/* light source size */
 			au2 = av2 = omega/(4.0*PI);
 		else
 			au2 = av2 = 0.0;
-		au2 += np->u_alpha * np->u_alpha;
-		av2 += np->v_alpha * np->v_alpha;
+		au2 += np->u_alpha*np->u_alpha;
+		av2 += np->v_alpha*np->v_alpha;
 						/* half vector */
 		h[0] = ldir[0] - np->rp->rdir[0];
 		h[1] = ldir[1] - np->rp->rdir[1];
 		h[2] = ldir[2] - np->rp->rdir[2];
-		normalize(h);
 						/* ellipse */
-		dtmp = DOT(np->u, h);
-		dtmp *= dtmp / au2;
+		dtmp1 = DOT(np->u, h);
+		dtmp1 *= dtmp1 / au2;
 		dtmp2 = DOT(np->v, h);
 		dtmp2 *= dtmp2 / av2;
 						/* gaussian */
-		dtmp = (dtmp + dtmp2) / (1.0 + DOT(np->pnorm, h));
-		dtmp = exp(-2.0*dtmp) / (4.0*PI * sqrt(au2*av2));
+		dtmp = DOT(np->pnorm, h);
+		dtmp = (dtmp1 + dtmp2) / (dtmp*dtmp);
+		dtmp = exp(-dtmp) * (0.25/PI)
+				* sqrt(ldot/(np->pdot*au2*av2));
 						/* worth using? */
 		if (dtmp > FTINY) {
 			copycolor(ctmp, np->scolor);
-			dtmp *= omega / np->pdot;
+			dtmp *= omega;
 			scalecolor(ctmp, dtmp);
 			addcolor(cval, ctmp);
 		}
@@ -141,12 +144,33 @@ double  omega;			/* light source size */
 		 *  is always modified by material color.
 		 */
 						/* roughness + source */
+		au2 = av2 = omega / PI;
+		au2 += np->u_alpha*np->u_alpha;
+		av2 += np->v_alpha*np->v_alpha;
+						/* "half vector" */
+		h[0] = ldir[0] - np->prdir[0];
+		h[1] = ldir[1] - np->prdir[1];
+		h[2] = ldir[2] - np->prdir[2];
+		dtmp = DOT(h,h);
+		if (dtmp > FTINY*FTINY) {
+			dtmp1 = DOT(h,np->pnorm);
+			dtmp = 1.0 - dtmp1*dtmp1/dtmp;
+			if (dtmp > FTINY*FTINY) {
+				dtmp1 = DOT(h,np->u);
+				dtmp1 *= dtmp1 / au2;
+				dtmp2 = DOT(h,np->v);
+				dtmp2 *= dtmp2 / av2;
+				dtmp = (dtmp1 + dtmp2) / dtmp;
+			}
+		} else
+			dtmp = 0.0;
 						/* gaussian */
-		dtmp = 0.0;
+		dtmp = exp(-dtmp) * (1.0/PI)
+				* sqrt(-ldot/(np->pdot*au2*av2));
 						/* worth using? */
 		if (dtmp > FTINY) {
 			copycolor(ctmp, np->mcolor);
-			dtmp *= np->tspec * omega / np->pdot;
+			dtmp *= np->tspec * omega;
 			scalecolor(ctmp, dtmp);
 			addcolor(cval, ctmp);
 		}
@@ -159,12 +183,11 @@ register OBJREC  *m;
 register RAY  *r;
 {
 	ANISODAT  nd;
-	double  dtmp;
 	COLOR  ctmp;
 	register int  i;
 						/* easy shadow test */
 	if (r->crtype & SHADOW)
-		return;
+		return(1);
 
 	if (m->oargs.nfargs != (m->otype == MAT_TRANS2 ? 8 : 6))
 		objerror(m, USER, "bad number of real arguments");
@@ -178,11 +201,16 @@ register RAY  *r;
 	nd.specfl = 0;
 	nd.u_alpha = m->oargs.farg[4];
 	nd.v_alpha = m->oargs.farg[5];
-	if (nd.u_alpha < 1e-6 || nd.v_alpha <= 1e-6)
+	if (nd.u_alpha < FTINY || nd.v_alpha <= FTINY)
 		objerror(m, USER, "roughness too small");
-						/* reorient if necessary */
-	if (r->rod < 0.0)
-		flipsurface(r);
+						/* check for back side */
+	if (r->rod < 0.0) {
+		if (!backvis && m->otype != MAT_TRANS2) {
+			raytrans(r);
+			return(1);
+		}
+		flipsurface(r);			/* reorient if backvis */
+	}
 						/* get modifiers */
 	raytexture(r, m->omod);
 	nd.pdot = raynormal(nd.pnorm, r);	/* perturb normal */
@@ -198,15 +226,8 @@ register RAY  *r;
 		else
 			setcolor(nd.scolor, 1.0, 1.0, 1.0);
 		scalecolor(nd.scolor, nd.rspec);
-						/* improved model */
-		dtmp = exp(-BSPEC(m)*nd.pdot);
-		for (i = 0; i < 3; i++)
-			colval(nd.scolor,i) += (1.0-colval(nd.scolor,i))*dtmp;
-		nd.rspec += (1.0-nd.rspec)*dtmp;
 						/* check threshold */
-		if (specthresh > FTINY &&
-				((specthresh >= 1.-FTINY ||
-				specthresh + (.05 - .1*frandom()) > nd.rspec)))
+		if (specthresh >= nd.rspec-FTINY)
 			nd.specfl |= SP_RBLT;
 						/* compute refl. direction */
 		for (i = 0; i < 3; i++)
@@ -216,24 +237,20 @@ register RAY  *r;
 				nd.vrefl[i] = r->rdir[i] + 2.*r->rod*r->ron[i];
 	}
 						/* compute transmission */
-	if (m->otype == MAT_TRANS) {
+	if (m->otype == MAT_TRANS2) {
 		nd.trans = m->oargs.farg[6]*(1.0 - nd.rspec);
 		nd.tspec = nd.trans * m->oargs.farg[7];
 		nd.tdiff = nd.trans - nd.tspec;
 		if (nd.tspec > FTINY) {
 			nd.specfl |= SP_TRAN;
 							/* check threshold */
-			if (specthresh > FTINY &&
-					((specthresh >= 1.-FTINY ||
-					specthresh +
-					    (.05 - .1*frandom()) > nd.tspec)))
+			if (specthresh >= nd.tspec-FTINY)
 				nd.specfl |= SP_TBLT;
 			if (DOT(r->pert,r->pert) <= FTINY*FTINY) {
 				VCOPY(nd.prdir, r->rdir);
 			} else {
 				for (i = 0; i < 3; i++)		/* perturb */
-					nd.prdir[i] = r->rdir[i] -
-							0.5*r->pert[i];
+					nd.prdir[i] = r->rdir[i] - r->pert[i];
 				if (DOT(nd.prdir, r->ron) < -FTINY)
 					normalize(nd.prdir);	/* OK */
 				else
@@ -277,6 +294,8 @@ register RAY  *r;
 	}
 					/* add direct component */
 	direct(r, diraniso, &nd);
+
+	return(1);
 }
 
 
@@ -298,7 +317,8 @@ register ANISODAT  *np;
 		np->specfl |= SP_BADU;
 		return;
 	}
-	multv3(np->u, np->u, mf->f->xfm);
+	if (mf->f != &unitxf)
+		multv3(np->u, np->u, mf->f->xfm);
 	fcross(np->v, np->pnorm, np->u);
 	if (normalize(np->v) == 0.0) {
 		objerror(np->mp, WARNING, "illegal orientation vector");
@@ -326,8 +346,8 @@ register ANISODAT  *np;
 		d = urand(ilhash(dimlist,ndims)+samplendx);
 		multisamp(rv, 2, d);
 		d = 2.0*PI * rv[0];
-		cosp = np->u_alpha * cos(d);
-		sinp = np->v_alpha * sin(d);
+		cosp = cos(d) * np->u_alpha;
+		sinp = sin(d) * np->v_alpha;
 		d = sqrt(cosp*cosp + sinp*sinp);
 		cosp /= d;
 		sinp /= d;
@@ -358,15 +378,18 @@ register ANISODAT  *np;
 		d = urand(ilhash(dimlist,ndims)+1823+samplendx);
 		multisamp(rv, 2, d);
 		d = 2.0*PI * rv[0];
-		cosp = cos(d);
-		sinp = sin(d);
+		cosp = cos(d) * np->u_alpha;
+		sinp = sin(d) * np->v_alpha;
+		d = sqrt(cosp*cosp + sinp*sinp);
+		cosp /= d;
+		sinp /= d;
 		rv[1] = 1.0 - specjitter*rv[1];
 		if (rv[1] <= FTINY)
 			d = 1.0;
 		else
 			d = sqrt(-log(rv[1]) /
-				(cosp*cosp*4./(np->u_alpha*np->u_alpha) +
-				 sinp*sinp*4./(np->v_alpha*np->v_alpha)));
+				(cosp*cosp/(np->u_alpha*np->u_alpha) +
+				 sinp*sinp/(np->v_alpha*np->u_alpha)));
 		for (i = 0; i < 3; i++)
 			sr.rdir[i] = np->prdir[i] +
 					d*(cosp*np->u[i] + sinp*np->v[i]);