[Bf-blender-cvs] SVN commit: /data/svn/bf-blender [50198] trunk/blender/source/blender/ blenlib/intern/math_rotation.c: don' t use a while loop when in compatible_eul().

[Campbell Barton]

Revision: 50198
          http://projects.blender.org/scm/viewvc.php?view=rev&root=bf-blender&revision=50198
Author:   campbellbarton
Date:     2012-08-25 12:32:22 +0000 (Sat, 25 Aug 2012)
Log Message:
-----------
don't use a while loop when in compatible_eul(). this is a low level rotation function called in many places so better to avoid iterations.

checked this function against the previous method using random rotation inputs and compared results, while this isnt exactly the same the results are very close and acceptable in both cases, also checked baking actions that the resulting FCurves are good and give matching rotations.

Modified Paths:
--------------
    trunk/blender/source/blender/blenlib/intern/math_rotation.c

Modified: trunk/blender/source/blender/blenlib/intern/math_rotation.c
===================================================================
--- trunk/blender/source/blender/blenlib/intern/math_rotation.c	2012-08-25 11:54:58 UTC (rev 50197)
+++ trunk/blender/source/blender/blenlib/intern/math_rotation.c	2012-08-25 12:32:22 UTC (rev 50198)
@@ -1040,84 +1040,46 @@
 
 }
 
-/* exported to transform.c */
-
 /* order independent! */
 void compatible_eul(float eul[3], const float oldrot[3])
 {
-	float dx, dy, dz;
+	/* we could use M_PI as pi_thresh: which is correct but 5.1 gives better results.
+	 * Checked with baking actions to fcurves - campbell */
+	const float pi_thresh = (5.1f);
+	const float pi_x2     = (2.0f * (float)M_PI);
 
-	/* correct differences of about 360 degrees first */
-	dx = eul[0] - oldrot[0];
-	dy = eul[1] - oldrot[1];
-	dz = eul[2] - oldrot[2];
+	float deul[3];
+	unsigned int i;
 
-	while (fabsf(dx) > 5.1f) {
-		if (dx > 0.0f) eul[0] -= 2.0f * (float)M_PI;
-		else eul[0] += 2.0f * (float)M_PI;
-		dx = eul[0] - oldrot[0];
+	/* correct differences of about 360 degrees first */
+	for (i = 0; i < 3; i++) {
+		deul[i] = eul[i] - oldrot[i];
+		if (deul[i] > pi_thresh) {
+			eul[i] -= floorf(( deul[i] / pi_x2) + 0.5) * pi_x2;
+			deul[i] = eul[i] - oldrot[i];
+		}
+		else if (deul[i] < -pi_thresh) {
+			eul[i] += floorf((-deul[i] / pi_x2) + 0.5) * pi_x2;
+			deul[i] = eul[i] - oldrot[i];
+		}
 	}
-	while (fabsf(dy) > 5.1f) {
-		if (dy > 0.0f) eul[1] -= 2.0f * (float)M_PI;
-		else eul[1] += 2.0f * (float)M_PI;
-		dy = eul[1] - oldrot[1];
-	}
-	while (fabsf(dz) > 5.1f) {
-		if (dz > 0.0f) eul[2] -= 2.0f * (float)M_PI;
-		else eul[2] += 2.0f * (float)M_PI;
-		dz = eul[2] - oldrot[2];
-	}
 
 	/* is 1 of the axis rotations larger than 180 degrees and the other small? NO ELSE IF!! */
-	if (fabsf(dx) > 3.2f && fabsf(dy) < 1.6f && fabsf(dz) < 1.6f) {
-		if (dx > 0.0f) eul[0] -= 2.0f * (float)M_PI;
-		else eul[0] += 2.0f * (float)M_PI;
+	if (fabsf(deul[0]) > 3.2f && fabsf(deul[1]) < 1.6f && fabsf(deul[2]) < 1.6f) {
+		if (deul[0] > 0.0f) eul[0] -= pi_x2;
+		else                eul[0] += pi_x2;
 	}
-	if (fabsf(dy) > 3.2f && fabsf(dz) < 1.6f && fabsf(dx) < 1.6f) {
-		if (dy > 0.0f) eul[1] -= 2.0f * (float)M_PI;
-		else eul[1] += 2.0f * (float)M_PI;
+	if (fabsf(deul[1]) > 3.2f && fabsf(deul[2]) < 1.6f && fabsf(deul[0]) < 1.6f) {
+		if (deul[1] > 0.0f) eul[1] -= pi_x2;
+		else                eul[1] += pi_x2;
 	}
-	if (fabsf(dz) > 3.2f && fabsf(dx) < 1.6f && fabsf(dy) < 1.6f) {
-		if (dz > 0.0f) eul[2] -= 2.0f * (float)M_PI;
-		else eul[2] += 2.0f * (float)M_PI;
+	if (fabsf(deul[2]) > 3.2f && fabsf(deul[0]) < 1.6f && fabsf(deul[1]) < 1.6f) {
+		if (deul[2] > 0.0f) eul[2] -= pi_x2;
+		else                eul[2] += pi_x2;
 	}
 
-	/* the method below was there from ancient days... but why! probably because the code sucks :)
-	 */
-#if 0
-	/* calc again */
-	dx = eul[0] - oldrot[0];
-	dy = eul[1] - oldrot[1];
-	dz = eul[2] - oldrot[2];
-
-	/* special case, tested for x-z  */
-
-	if ((fabsf(dx) > 3.1f && fabsf(dz) > 1.5f) || (fabsf(dx) > 1.5f && fabsf(dz) > 3.1f)) {
-		if (dx > 0.0f) eul[0] -= M_PI;
-		else eul[0] += M_PI;
-		if (eul[1] > 0.0) eul[1] = M_PI - eul[1];
-		else eul[1] = -M_PI - eul[1];
-		if (dz > 0.0f) eul[2] -= M_PI;
-		else eul[2] += M_PI;
-
-	}
-	else if ((fabsf(dx) > 3.1f && fabsf(dy) > 1.5f) || (fabsf(dx) > 1.5f && fabsf(dy) > 3.1f)) {
-		if (dx > 0.0f) eul[0] -= M_PI;
-		else eul[0] += M_PI;
-		if (dy > 0.0f) eul[1] -= M_PI;
-		else eul[1] += M_PI;
-		if (eul[2] > 0.0f) eul[2] = M_PI - eul[2];
-		else eul[2] = -M_PI - eul[2];
-	}
-	else if ((fabsf(dy) > 3.1f && fabsf(dz) > 1.5f) || (fabsf(dy) > 1.5f && fabsf(dz) > 3.f1)) {
-		if (eul[0] > 0.0f) eul[0] = M_PI - eul[0];
-		else eul[0] = -M_PI - eul[0];
-		if (dy > 0.0f) eul[1] -= M_PI;
-		else eul[1] += M_PI;
-		if (dz > 0.0f) eul[2] -= M_PI;
-		else eul[2] += M_PI;
-	}
-#endif
+#undef PI_THRESH
+#undef PI_2F
 }
 
 /* uses 2 methods to retrieve eulers, and picks the closest */