blender/source/blender/blenkernel/intern/anim.c

/*
 * ***** BEGIN GPL LICENSE BLOCK *****
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version 2
 * of the License, or (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software Foundation,
 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
 *
 * The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
 * All rights reserved.
 *
 * The Original Code is: all of this file.
 *
 * Contributor(s): none yet.
 *
 * ***** END GPL LICENSE BLOCK *****
 */

/** \file blender/blenkernel/intern/anim.c
 *  \ingroup bke
 */

#include <stdio.h>
#include <math.h>
#include <string.h>

#include "MEM_guardedalloc.h"

#include "BLI_blenlib.h"
#include "BLI_math.h"
#include "BLI_rand.h"
#include "BLI_utildefines.h"

#include "DNA_anim_types.h"
#include "DNA_armature_types.h"
#include "DNA_group_types.h"
#include "DNA_key_types.h"
#include "DNA_mesh_types.h"
#include "DNA_meshdata_types.h"
#include "DNA_scene_types.h"
#include "DNA_vfont_types.h"

#include "BKE_animsys.h"
#include "BKE_curve.h"
#include "BKE_DerivedMesh.h"
#include "BKE_depsgraph.h"
#include "BKE_font.h"
#include "BKE_group.h"
#include "BKE_global.h"
#include "BKE_key.h"
#include "BKE_lattice.h"
#include "BKE_main.h"
#include "BKE_mesh.h"
#include "BKE_object.h"
#include "BKE_particle.h"
#include "BKE_scene.h"
#include "BKE_utildefines.h"
#include "BKE_tessmesh.h"
#include "BKE_depsgraph.h"
#include "BKE_anim.h"
#include "BKE_report.h"


// XXX bad level call...

/* --------------------- */
/* forward declarations */

static void object_duplilist_recursive(ID *id, Scene *scene, Object *ob, ListBase *duplilist, float par_space_mat[][4], int level, int animated);

/* ******************************************************************** */
/* Animation Visualisation */

/* Initialize the default settings for animation visualisation */
void animviz_settings_init(bAnimVizSettings *avs)
{
	/* sanity check */
	if (avs == NULL)
		return;
		
	/* ghosting settings */
	avs->ghost_bc= avs->ghost_ac= 10;
	
	avs->ghost_sf= 1; // xxx - take from scene instead?
	avs->ghost_ef= 250; // xxx - take from scene instead?
	
	avs->ghost_step= 1;
	
	
	/* path settings */
	avs->path_bc= avs->path_ac= 10;
	
	avs->path_sf= 1; // xxx - take from scene instead?
	avs->path_ef= 250; // xxx - take from scene instead?
	
	avs->path_viewflag= (MOTIONPATH_VIEW_KFRAS|MOTIONPATH_VIEW_KFNOS);
	
	avs->path_step= 1;
}

/* ------------------- */

/* Free the given motion path's cache */
void animviz_free_motionpath_cache(bMotionPath *mpath) 
{
	/* sanity check */
	if (mpath == NULL) 
		return;
		
	/* free the path if necessary */
	if (mpath->points)
		MEM_freeN(mpath->points);
	
	/* reset the relevant parameters */
	mpath->points= NULL;
	mpath->length= 0;
}

/* Free the given motion path instance and its data 
 * NOTE: this frees the motion path given!
 */
void animviz_free_motionpath(bMotionPath *mpath)
{
	/* sanity check */
	if (mpath == NULL)
		return;
	
	/* free the cache first */
	animviz_free_motionpath_cache(mpath);
	
	/* now the instance itself */
	MEM_freeN(mpath);
}

/* ------------------- */

/* Setup motion paths for the given data
 * - Only used when explicitly calculating paths on bones which may/may not be consider already
 *
 * < scene: current scene (for frame ranges, etc.)
 * < ob: object to add paths for (must be provided)
 * < pchan: posechannel to add paths for (optional; if not provided, object-paths are assumed)
 */
bMotionPath *animviz_verify_motionpaths(ReportList *reports, Scene *scene, Object *ob, bPoseChannel *pchan)
{
	bAnimVizSettings *avs;
	bMotionPath *mpath, **dst;
	
	/* sanity checks */
	if (ELEM(NULL, scene, ob))
		return NULL;
		
	/* get destination data */
	if (pchan) {
		/* paths for posechannel - assume that posechannel belongs to the object */
		avs= &ob->pose->avs;
		dst= &pchan->mpath;
	}
	else {
		/* paths for object */
		avs= &ob->avs;
		dst= &ob->mpath;
	}

	/* avoid 0 size allocs */
	if (avs->path_sf >= avs->path_ef) {
		BKE_reportf(reports, RPT_ERROR,
					"Motion Path frame extents invalid for %s (%d to %d).%s\n",
					(pchan)? pchan->name : ob->id.name,
					avs->path_sf, avs->path_ef,
					(avs->path_sf == avs->path_ef)? " Cannot have single-frame paths." : "");
		return NULL;
	}

	/* if there is already a motionpath, just return that,
	 * provided it's settings are ok (saves extra free+alloc)
	 */
	if (*dst != NULL) {
		int expected_length = avs->path_ef - avs->path_sf;
		
		mpath= *dst;
		
		/* path is "valid" if length is valid, but must also be of the same length as is being requested */
		if ((mpath->start_frame != mpath->end_frame) && (mpath->length > 0)) {
			/* outer check ensures that we have some curve data for this path */
			if (mpath->length == expected_length) {
				/* return/use this as it is already valid length */
				return mpath;
			}
			else {
				/* clear the existing path (as the range has changed), and reallocate below */
				animviz_free_motionpath_cache(mpath);
			}
		}
	}
	else {
		/* create a new motionpath, and assign it */
		mpath= MEM_callocN(sizeof(bMotionPath), "bMotionPath");
		*dst= mpath;
	}
	
	/* set settings from the viz settings */
	mpath->start_frame= avs->path_sf;
	mpath->end_frame= avs->path_ef;
	
	mpath->length= mpath->end_frame - mpath->start_frame;
	
	if (avs->path_bakeflag & MOTIONPATH_BAKE_HEADS)
		mpath->flag |= MOTIONPATH_FLAG_BHEAD;
	else
		mpath->flag &= ~MOTIONPATH_FLAG_BHEAD;
	
	/* allocate a cache */
	mpath->points= MEM_callocN(sizeof(bMotionPathVert)*mpath->length, "bMotionPathVerts");
	
	/* tag viz settings as currently having some path(s) which use it */
	avs->path_bakeflag |= MOTIONPATH_BAKE_HAS_PATHS;
	
	/* return it */
	return mpath;
}

/* ------------------- */

/* Motion path needing to be baked (mpt) */
typedef struct MPathTarget {
	struct MPathTarget *next, *prev;
	
	bMotionPath *mpath;			/* motion path in question */
	
	Object *ob;					/* source object */
	bPoseChannel *pchan;		/* source posechannel (if applicable) */
} MPathTarget;

/* ........ */

/* get list of motion paths to be baked for the given object
 * 	- assumes the given list is ready to be used
 */
// TODO: it would be nice in future to be able to update objects dependent on these bones too?
void animviz_get_object_motionpaths(Object *ob, ListBase *targets)
{
	MPathTarget *mpt;
	
	/* object itself first */
	if ((ob->avs.recalc & ANIMVIZ_RECALC_PATHS) && (ob->mpath)) {
		/* new target for object */
		mpt= MEM_callocN(sizeof(MPathTarget), "MPathTarget Ob");
		BLI_addtail(targets, mpt);
		
		mpt->mpath= ob->mpath;
		mpt->ob= ob;
	}
	
	/* bones */
	if ((ob->pose) && (ob->pose->avs.recalc & ANIMVIZ_RECALC_PATHS)) {
		bArmature *arm= ob->data;
		bPoseChannel *pchan;
		
		for (pchan= ob->pose->chanbase.first; pchan; pchan= pchan->next) {
			if ((pchan->bone) && (arm->layer & pchan->bone->layer) && (pchan->mpath)) {
				/* new target for bone */
				mpt= MEM_callocN(sizeof(MPathTarget), "MPathTarget PoseBone");
				BLI_addtail(targets, mpt);
				
				mpt->mpath= pchan->mpath;
				mpt->ob= ob;
				mpt->pchan= pchan;
			}
		}
	}
}

/* ........ */

/* Note on evaluation optimisations:
 * Optimisations currently used here play tricks with the depsgraph in order to try and 
 * evaluate as few objects as strictly necessary to get nicer performance under standard
 * production conditions. For those people who really need the accurate version, 
 * disable the ifdef (i.e. 1 -> 0) and comment out the call to motionpaths_calc_optimise_depsgraph()
 */

/* tweak the object ordering to trick depsgraph into making MotionPath calculations run faster */
static void motionpaths_calc_optimise_depsgraph(Scene *scene, ListBase *targets)
{
	Base *base, *baseNext;
	MPathTarget *mpt;
	
	/* make sure our temp-tag isn't already in use */
	for (base= scene->base.first; base; base= base->next)
		base->object->flag &= ~BA_TEMP_TAG;
	
	/* for each target, dump its object to the start of the list if it wasn't moved already */
	for (mpt= targets->first; mpt; mpt= mpt->next) {
		for (base=scene->base.first; base; base=baseNext) {
			baseNext = base->next;
			
			if ((base->object == mpt->ob) && !(mpt->ob->flag & BA_TEMP_TAG)) {
				BLI_remlink(&scene->base, base);
				BLI_addhead(&scene->base, base);
				
				mpt->ob->flag |= BA_TEMP_TAG;
				break; // we really don't need to continue anymore once this happens, but this line might really 'break'
			}
		}
	}
	
	/* "brew me a list that's sorted a bit faster now depsy" */
	DAG_scene_sort(G.main, scene);
}

/* update scene for current frame */
static void motionpaths_calc_update_scene(Scene *scene)
{
#if 1 // 'production' optimisations always on
	Base *base, *last=NULL;
	
	/* only stuff that moves or needs display still */
	DAG_scene_update_flags(G.main, scene, scene->lay, TRUE);
	
	/* find the last object with the tag 
	 *	- all those afterwards are assumed to not be relevant for our calculations
	 */
	// optimize further by moving out...
	for (base=scene->base.first; base; base=base->next) {
		if (base->object->flag & BA_TEMP_TAG)
			last = base;
	}
	
	/* perform updates for tagged objects */
	// XXX: this will break if rigs depend on scene or other data that 
	// is animated but not attached to/updatable from objects
	for (base=scene->base.first; base; base=base->next) {
		/* update this object */
		object_handle_update(scene, base->object);
		
		/* if this is the last one we need to update, let's stop to save some time */
		if (base == last)
			break;
	}
#else // original, 'always correct' version
	/* do all updates 
	 *	- if this is too slow, resort to using a more efficient way 
	 * 	  that doesn't force complete update, but for now, this is the
	 *	  most accurate way!
	 */
	scene_update_for_newframe(G.main, scene, scene->lay); // XXX this is the best way we can get anything moving
#endif
}

/* ........ */

/* perform baking for the targets on the current frame */
static void motionpaths_calc_bake_targets(Scene *scene, ListBase *targets)
{
	MPathTarget *mpt;
	
	/* for each target, check if it can be baked on the current frame */
	for (mpt= targets->first; mpt; mpt= mpt->next) {	
		bMotionPath *mpath= mpt->mpath;
		bMotionPathVert *mpv;
		
		/* current frame must be within the range the cache works for 
		 *	- is inclusive of the first frame, but not the last otherwise we get buffer overruns
		 */
		if ((CFRA < mpath->start_frame) || (CFRA >= mpath->end_frame))
			continue;
		
		/* get the relevant cache vert to write to */
		mpv= mpath->points + (CFRA - mpath->start_frame);
		
		/* pose-channel or object path baking? */
		if (mpt->pchan) {
			/* heads or tails */
			if (mpath->flag & MOTIONPATH_FLAG_BHEAD) {
				copy_v3_v3(mpv->co, mpt->pchan->pose_head);
			}
			else {
				copy_v3_v3(mpv->co, mpt->pchan->pose_tail);
			}
			
			/* result must be in worldspace */
			mul_m4_v3(mpt->ob->obmat, mpv->co);
		}
		else {
			/* worldspace object location */
			copy_v3_v3(mpv->co, mpt->ob->obmat[3]);
		}
	}
}

/* Perform baking of the given object's and/or its bones' transforms to motion paths 
 *	- scene: current scene
 *	- ob: object whose flagged motionpaths should get calculated
 *	- recalc: whether we need to 
 */
// TODO: include reports pointer?
void animviz_calc_motionpaths(Scene *scene, ListBase *targets)
{
	MPathTarget *mpt;
	int sfra, efra;
	int cfra;
	
	/* sanity check */
	if (ELEM(NULL, targets, targets->first))
		return;
	
	/* set frame values */
	cfra = CFRA;
	sfra = efra = cfra;
	
	// TODO: this method could be improved...
	//	1) max range for standard baking
	//	2) minimum range for recalc baking (i.e. between keyframes, but how?)
	for (mpt= targets->first; mpt; mpt= mpt->next) {
		/* try to increase area to do (only as much as needed) */
		sfra= MIN2(sfra, mpt->mpath->start_frame);
		efra= MAX2(efra, mpt->mpath->end_frame);
	}
	if (efra <= sfra) return;
	
	/* optimize the depsgraph for faster updates */
	// TODO: whether this is used should depend on some setting for the level of optimisations used
	motionpaths_calc_optimise_depsgraph(scene, targets);
	
	/* calculate path over requested range */
	for (CFRA=sfra; CFRA<=efra; CFRA++) {
		/* update relevant data for new frame */
		motionpaths_calc_update_scene(scene);
		
		/* perform baking for targets */
		motionpaths_calc_bake_targets(scene, targets);
	}
	
	/* reset original environment */
	CFRA= cfra;
	motionpaths_calc_update_scene(scene);
	
	/* clear recalc flags from targets */
	for (mpt= targets->first; mpt; mpt= mpt->next) {
		bAnimVizSettings *avs;
		
		/* get pointer to animviz settings for each target */
		if (mpt->pchan)
			avs= &mpt->ob->pose->avs;
		else	
			avs= &mpt->ob->avs;
		
		/* clear the flag requesting recalculation of targets */
		avs->recalc &= ~ANIMVIZ_RECALC_PATHS;
	}
}

/* ******************************************************************** */
/* Curve Paths - for curve deforms and/or curve following */

/* free curve path data 
 * NOTE: frees the path itself!
 * NOTE: this is increasingly inaccurate with non-uniform BevPoint subdivisions [#24633]
 */
void free_path(Path *path)
{
	if (path->data) MEM_freeN(path->data);
	MEM_freeN(path);
}

/* calculate a curve-deform path for a curve 
 * 	- only called from displist.c -> do_makeDispListCurveTypes
 */
void calc_curvepath(Object *ob)
{
	BevList *bl;
	BevPoint *bevp, *bevpn, *bevpfirst, *bevplast;
	PathPoint *pp;
	Curve *cu;
	Nurb *nu;
	Path *path;
	float *fp, *dist, *maxdist, xyz[3];
	float fac, d=0, fac1, fac2;
	int a, tot, cycl=0;
	ListBase *nurbs;
	
	/* in a path vertices are with equal differences: path->len = number of verts */
	/* NOW WITH BEVELCURVE!!! */
	
	if (ob==NULL || ob->type != OB_CURVE) return;
	cu= ob->data;

	nurbs= BKE_curve_nurbs_get(cu);
	nu= nurbs->first;

	if (cu->path) free_path(cu->path);
	cu->path= NULL;
	
	bl= cu->bev.first;
	if (bl==NULL || !bl->nr) return;

	cu->path=path= MEM_callocN(sizeof(Path), "calc_curvepath");
	
	/* if POLY: last vertice != first vertice */
	cycl= (bl->poly!= -1);
	
	if (cycl) tot= bl->nr;
	else tot= bl->nr-1;
	
	path->len= tot+1;
	/* exception: vector handle paths and polygon paths should be subdivided at least a factor resolu */
	if (path->len<nu->resolu*SEGMENTSU(nu)) path->len= nu->resolu*SEGMENTSU(nu);
	
	dist= (float *)MEM_mallocN((tot+1)*4, "calcpathdist");

		/* all lengths in *dist */
	bevp= bevpfirst= (BevPoint *)(bl+1);
	fp= dist;
	*fp= 0;
	for (a=0; a<tot; a++) {
		fp++;
		if (cycl && a==tot-1)
			sub_v3_v3v3(xyz, bevpfirst->vec, bevp->vec);
		else
			sub_v3_v3v3(xyz, (bevp+1)->vec, bevp->vec);
		
		*fp= *(fp-1)+len_v3(xyz);
		bevp++;
	}
	
	path->totdist= *fp;
	
		/* the path verts  in path->data */
		/* now also with TILT value */
	pp= path->data = (PathPoint *)MEM_callocN(sizeof(PathPoint)*path->len, "pathdata");
	
	bevp= bevpfirst;
	bevpn= bevp+1;
	bevplast= bevpfirst + (bl->nr-1);
	fp= dist+1;
	maxdist= dist+tot;
	fac= 1.0f/((float)path->len-1.0f);
		fac = fac * path->totdist;
	
	for (a=0; a<path->len; a++) {
		
		d= ((float)a)*fac;
		
		/* we're looking for location (distance) 'd' in the array */
		while ((d>= *fp) && fp<maxdist) {
			fp++;
			if (bevp<bevplast) bevp++;
			bevpn= bevp+1;
			if (bevpn>bevplast) {
				if (cycl) bevpn= bevpfirst;
				else bevpn= bevplast;
			}
		}
		
		fac1= *(fp)- *(fp-1);
		fac2= *(fp)-d;
		fac1= fac2/fac1;
		fac2= 1.0f-fac1;
		
		interp_v3_v3v3(pp->vec, bevp->vec, bevpn->vec, fac2);
		pp->vec[3]= fac1*bevp->alfa + fac2*bevpn->alfa;
		pp->radius= fac1*bevp->radius + fac2*bevpn->radius;
		pp->weight= fac1*bevp->weight + fac2*bevpn->weight;
		interp_qt_qtqt(pp->quat, bevp->quat, bevpn->quat, fac2);
		normalize_qt(pp->quat);
		
		pp++;
	}
	
	MEM_freeN(dist);
}


/* is this only used internally?*/
int interval_test(int min, int max, int p1, int cycl)
{
	if (cycl) {
		if (p1 < min) 
			p1=  ((p1 -min) % (max-min+1)) + max+1;
		else if (p1 > max)
			p1=  ((p1 -min) % (max-min+1)) + min;
	}
	else {
		if (p1 < min) p1= min;
		else if (p1 > max) p1= max;
	}
	return p1;
}


/* calculate the deformation implied by the curve path at a given parametric position,
 * and returns whether this operation succeeded.
 *
 * note: ctime is normalized range <0-1>
 *
 * returns OK: 1/0
 */
int where_on_path(Object *ob, float ctime, float vec[4], float dir[3], float quat[4], float *radius, float *weight)
{
	Curve *cu;
	Nurb *nu;
	BevList *bl;
	Path *path;
	PathPoint *pp, *p0, *p1, *p2, *p3;
	float fac;
	float data[4];
	int cycl=0, s0, s1, s2, s3;

	if (ob==NULL || ob->type != OB_CURVE) return 0;
	cu= ob->data;
	if (cu->path==NULL || cu->path->data==NULL) {
		printf("no path!\n");
		return 0;
	}
	path= cu->path;
	pp= path->data;
	
	/* test for cyclic */
	bl= cu->bev.first;
	if (!bl) return 0;
	if (!bl->nr) return 0;
	if (bl->poly> -1) cycl= 1;

	ctime *= (path->len-1);
	
	s1= (int)floor(ctime);
	fac= (float)(s1+1)-ctime;

	/* path->len is corected for cyclic */
	s0= interval_test(0, path->len-1-cycl, s1-1, cycl);
	s1= interval_test(0, path->len-1-cycl, s1, cycl);
	s2= interval_test(0, path->len-1-cycl, s1+1, cycl);
	s3= interval_test(0, path->len-1-cycl, s1+2, cycl);

	p0= pp + s0;
	p1= pp + s1;
	p2= pp + s2;
	p3= pp + s3;

	/* note, commented out for follow constraint */
	//if (cu->flag & CU_FOLLOW) {

		key_curve_tangent_weights(1.0f-fac, data, KEY_BSPLINE);

		interp_v3_v3v3v3v3(dir, p0->vec, p1->vec, p2->vec, p3->vec, data);

		/* make compatible with vectoquat */
		negate_v3(dir);
	//}
	
	nu= cu->nurb.first;

	/* make sure that first and last frame are included in the vectors here  */
	if (nu->type == CU_POLY) key_curve_position_weights(1.0f-fac, data, KEY_LINEAR);
	else if (nu->type == CU_BEZIER) key_curve_position_weights(1.0f-fac, data, KEY_LINEAR);
	else if (s0==s1 || p2==p3) key_curve_position_weights(1.0f-fac, data, KEY_CARDINAL);
	else key_curve_position_weights(1.0f-fac, data, KEY_BSPLINE);

	vec[0]= data[0]*p0->vec[0] + data[1]*p1->vec[0] + data[2]*p2->vec[0] + data[3]*p3->vec[0] ; /* X */
	vec[1]= data[0]*p0->vec[1] + data[1]*p1->vec[1] + data[2]*p2->vec[1] + data[3]*p3->vec[1] ; /* Y */
	vec[2]= data[0]*p0->vec[2] + data[1]*p1->vec[2] + data[2]*p2->vec[2] + data[3]*p3->vec[2] ; /* Z */
	vec[3]= data[0]*p0->vec[3] + data[1]*p1->vec[3] + data[2]*p2->vec[3] + data[3]*p3->vec[3] ; /* Tilt, should not be needed since we have quat still used */

	if (quat) {
		float totfac, q1[4], q2[4];

		totfac= data[0]+data[3];
		if (totfac>FLT_EPSILON)	interp_qt_qtqt(q1, p0->quat, p3->quat, data[3] / totfac);
		else					copy_qt_qt(q1, p1->quat);

		totfac= data[1]+data[2];
		if (totfac>FLT_EPSILON)	interp_qt_qtqt(q2, p1->quat, p2->quat, data[2] / totfac);
		else					copy_qt_qt(q2, p3->quat);

		totfac = data[0]+data[1]+data[2]+data[3];
		if (totfac>FLT_EPSILON)	interp_qt_qtqt(quat, q1, q2, (data[1]+data[2]) / totfac);
		else					copy_qt_qt(quat, q2);
	}

	if (radius)
		*radius= data[0]*p0->radius + data[1]*p1->radius + data[2]*p2->radius + data[3]*p3->radius;

	if (weight)
		*weight= data[0]*p0->weight + data[1]*p1->weight + data[2]*p2->weight + data[3]*p3->weight;

	return 1;
}

/* ******************************************************************** */
/* Dupli-Geometry */

static DupliObject *new_dupli_object(ListBase *lb, Object *ob, float mat[][4], int lay, int index, int type, int animated)
{
	DupliObject *dob= MEM_callocN(sizeof(DupliObject), "dupliobject");
	
	BLI_addtail(lb, dob);
	dob->ob= ob;
	copy_m4_m4(dob->mat, mat);
	copy_m4_m4(dob->omat, ob->obmat);
	dob->origlay= ob->lay;
	dob->index= index;
	dob->type= type;
	dob->animated= (type == OB_DUPLIGROUP) && animated;
	ob->lay= lay;
	
	return dob;
}

static void group_duplilist(ListBase *lb, Scene *scene, Object *ob, int level, int animated)
{
	DupliObject *dob;
	Group *group;
	GroupObject *go;
	float mat[4][4], tmat[4][4];
	
	if (ob->dup_group==NULL) return;
	group= ob->dup_group;
	
	/* simple preventing of too deep nested groups */
	if (level>MAX_DUPLI_RECUR) return;
	
	/* handles animated groups, and */
	/* we need to check update for objects that are not in scene... */
	group_handle_recalc_and_update(scene, ob, group);
	animated= animated || group_is_animated(ob, group);
	
	for (go= group->gobject.first; go; go= go->next) {
		/* note, if you check on layer here, render goes wrong... it still deforms verts and uses parent imat */
		if (go->ob!=ob) {
			
			/* group dupli offset, should apply after everything else */
			if (!is_zero_v3(group->dupli_ofs)) {
				copy_m4_m4(tmat, go->ob->obmat);
				sub_v3_v3v3(tmat[3], tmat[3], group->dupli_ofs);
				mult_m4_m4m4(mat, ob->obmat, tmat);
			}
			else {
				mult_m4_m4m4(mat, ob->obmat, go->ob->obmat);
			}
			
			dob= new_dupli_object(lb, go->ob, mat, ob->lay, 0, OB_DUPLIGROUP, animated);

			/* check the group instance and object layers match, also that the object visible flags are ok. */
			if (	(dob->origlay & group->layer)==0 ||
				(G.rendering==0 && dob->ob->restrictflag & OB_RESTRICT_VIEW) ||
				(G.rendering && dob->ob->restrictflag & OB_RESTRICT_RENDER)
			) {
				dob->no_draw= 1;
			}
			else {
				dob->no_draw= 0;
			}

			if (go->ob->transflag & OB_DUPLI) {
				copy_m4_m4(dob->ob->obmat, dob->mat);
				object_duplilist_recursive(&group->id, scene, go->ob, lb, ob->obmat, level+1, animated);
				copy_m4_m4(dob->ob->obmat, dob->omat);
			}
		}
	}
}

static void frames_duplilist(ListBase *lb, Scene *scene, Object *ob, int level, int animated)
{
	extern int enable_cu_speed;	/* object.c */
	Object copyob;
	int cfrao = scene->r.cfra;
	int dupend = ob->dupend;
	
	/* simple prevention of too deep nested groups */
	if (level > MAX_DUPLI_RECUR) return;
	
	/* if we don't have any data/settings which will lead to object movement,
	 * don't waste time trying, as it will all look the same...
	 */
	if (ob->parent==NULL && ob->constraints.first==NULL && ob->adt==NULL) 
		return;
	
	/* make a copy of the object's original data (before any dupli-data overwrites it) 
	 * as we'll need this to keep track of unkeyed data
	 *	- this doesn't take into account other data that can be reached from the object,
	 *	  for example it's shapekeys or bones, hence the need for an update flush at the end
	 */
	copyob = *ob;
	
	/* duplicate over the required range */
	if (ob->transflag & OB_DUPLINOSPEED) enable_cu_speed= 0;
	
	for (scene->r.cfra= ob->dupsta; scene->r.cfra<=dupend; scene->r.cfra++) {
		short ok= 1;
		
		/* - dupoff = how often a frames within the range shouldn't be made into duplis
		 * - dupon = the length of each "skipping" block in frames
		 */
		if (ob->dupoff) {
			ok= scene->r.cfra - ob->dupsta;
			ok= ok % (ob->dupon+ob->dupoff);
			ok= (ok < ob->dupon);
		}
		
		if (ok) {	
			DupliObject *dob;
			
			/* WARNING: doing animation updates in this way is not terribly accurate, as the dependencies
			 * and/or other objects which may affect this object's transforms are not updated either.
			 * However, this has always been the way that this worked (i.e. pre 2.5), so I guess that it'll be fine!
			 */
			BKE_animsys_evaluate_animdata(scene, &ob->id, ob->adt, (float)scene->r.cfra, ADT_RECALC_ANIM); /* ob-eval will do drivers, so we don't need to do them */
			where_is_object_time(scene, ob, (float)scene->r.cfra);
			
			dob= new_dupli_object(lb, ob, ob->obmat, ob->lay, scene->r.cfra, OB_DUPLIFRAMES, animated);
			copy_m4_m4(dob->omat, copyob.obmat);
		}
	}

	enable_cu_speed= 1;
	
	/* reset frame to original frame, then re-evaluate animation as above 
	 * as 2.5 animation data may have far-reaching consequences
	 */
	scene->r.cfra= cfrao;
	
	BKE_animsys_evaluate_animdata(scene, &ob->id, ob->adt, (float)scene->r.cfra, ADT_RECALC_ANIM); /* ob-eval will do drivers, so we don't need to do them */
	where_is_object_time(scene, ob, (float)scene->r.cfra);
	
	/* but, to make sure unkeyed object transforms are still sane, 
	 * let's copy object's original data back over
	 */
	*ob = copyob;
}

typedef struct vertexDupliData {
	ID *id; /* scene or group, for recursive loops */
	int level;
	int animated;
	ListBase *lb;
	float pmat[4][4];
	float obmat[4][4]; /* Only used for dupliverts inside dupligroups, where the ob->obmat is modified */
	Scene *scene;
	Object *ob, *par;
	float (*orco)[3];
} vertexDupliData;

/* ------------- */

static void vertex_dupli__mapFunc(void *userData, int index, const float co[3],
                                  const float no_f[3], const short no_s[3])
{
	DupliObject *dob;
	vertexDupliData *vdd= userData;
	float vec[3], q2[4], mat[3][3], tmat[4][4], obmat[4][4];
	int origlay;
	
	mul_v3_m4v3(vec, vdd->pmat, co);
	sub_v3_v3(vec, vdd->pmat[3]);
	add_v3_v3(vec, vdd->obmat[3]);
	
	copy_m4_m4(obmat, vdd->obmat);
	copy_v3_v3(obmat[3], vec);
	
	if (vdd->par->transflag & OB_DUPLIROT) {
		if (no_f) {
			vec[0]= -no_f[0]; vec[1]= -no_f[1]; vec[2]= -no_f[2];
		}
		else if (no_s) {
			vec[0]= -no_s[0]; vec[1]= -no_s[1]; vec[2]= -no_s[2];
		}
		
		vec_to_quat(q2, vec, vdd->ob->trackflag, vdd->ob->upflag);
		
		quat_to_mat3(mat, q2);
		copy_m4_m4(tmat, obmat);
		mul_m4_m4m3(obmat, tmat, mat);
	}

	origlay = vdd->ob->lay;
	
	dob= new_dupli_object(vdd->lb, vdd->ob, obmat, vdd->par->lay, index, OB_DUPLIVERTS, vdd->animated);

	/* restore the original layer so that each dupli will have proper dob->origlay */
	vdd->ob->lay = origlay;

	if (vdd->orco)
		copy_v3_v3(dob->orco, vdd->orco[index]);
	
	if (vdd->ob->transflag & OB_DUPLI) {
		float tmpmat[4][4];
		copy_m4_m4(tmpmat, vdd->ob->obmat);
		copy_m4_m4(vdd->ob->obmat, obmat); /* pretend we are really this mat */
		object_duplilist_recursive((ID *)vdd->id, vdd->scene, vdd->ob, vdd->lb, obmat, vdd->level+1, vdd->animated);
		copy_m4_m4(vdd->ob->obmat, tmpmat);
	}
}

static void vertex_duplilist(ListBase *lb, ID *id, Scene *scene, Object *par, float par_space_mat[][4], int level, int animated)
{
	Object *ob, *ob_iter;
	Mesh *me= par->data;
	Base *base = NULL;
	DerivedMesh *dm;
	vertexDupliData vdd;
	Scene *sce = NULL;
	Group *group = NULL;
	GroupObject * go = NULL;
	BMEditMesh *em;
	float vec[3], no[3], pmat[4][4];
	int totvert, a, oblay;
	unsigned int lay;
	
	copy_m4_m4(pmat, par->obmat);
	
	/* simple preventing of too deep nested groups */
	if (level>MAX_DUPLI_RECUR) return;
	
	em = me->edit_btmesh;
	
	if (em) {
		dm= editbmesh_get_derived_cage(scene, par, em, CD_MASK_BAREMESH);
	}
	else
		dm= mesh_get_derived_deform(scene, par, CD_MASK_BAREMESH);
	
	if (G.rendering) {
		vdd.orco= (float(*)[3])get_mesh_orco_verts(par);
		transform_mesh_orco_verts(me, vdd.orco, me->totvert, 0);
	}
	else
		vdd.orco= NULL;
	
	totvert = dm->getNumVerts(dm);

	/* having to loop on scene OR group objects is NOT FUN */
	if (GS(id->name) == ID_SCE) {
		sce = (Scene *)id;
		lay= sce->lay;
		base= sce->base.first;
	}
	else {
		group = (Group *)id;
		lay= group->layer;
		go = group->gobject.first;
	}
	
	/* Start looping on Scene OR Group objects */
	while (base || go) { 
		if (sce) {
			ob_iter= base->object;
			oblay = base->lay;
		}
		else {
			ob_iter= go->ob;
			oblay = ob_iter->lay;
		}
		
		if (lay & oblay && scene->obedit!=ob_iter) {
			ob=ob_iter->parent;
			while (ob) {
				if (ob==par) {
					ob = ob_iter;
	/* End Scene/Group object loop, below is generic */
					
					
					/* par_space_mat - only used for groups so we can modify the space dupli's are in
					 * when par_space_mat is NULL ob->obmat can be used instead of ob__obmat
					 */
					if (par_space_mat)
						mult_m4_m4m4(vdd.obmat, par_space_mat, ob->obmat);
					else
						copy_m4_m4(vdd.obmat, ob->obmat);

					vdd.id= id;
					vdd.level= level;
					vdd.animated= animated;
					vdd.lb= lb;
					vdd.ob= ob;
					vdd.scene= scene;
					vdd.par= par;
					copy_m4_m4(vdd.pmat, pmat);
					
					/* mballs have a different dupli handling */
					if (ob->type!=OB_MBALL) ob->flag |= OB_DONE;	/* doesnt render */

					if (me->edit_btmesh) {
						dm->foreachMappedVert(dm, vertex_dupli__mapFunc, (void*) &vdd);
					}
					else {
						for (a=0; a<totvert; a++) {
							dm->getVertCo(dm, a, vec);
							dm->getVertNo(dm, a, no);
							
							vertex_dupli__mapFunc(&vdd, a, vec, no, NULL);
						}
					}
					if (sce) {
						/* Set proper layer in case of scene looping,
						 * in case of groups the object layer will be
						 * changed when it's duplicated due to the
						 * group duplication.
						 */
						ob->lay = vdd.par->lay;
					}
					
					break;
				}
				ob= ob->parent;
			}
		}
		if (sce)	base= base->next;	/* scene loop */
		else		go= go->next;		/* group loop */
	}

	if (vdd.orco)
		MEM_freeN(vdd.orco);
	dm->release(dm);
}

static void face_duplilist(ListBase *lb, ID *id, Scene *scene, Object *par, float par_space_mat[][4], int level, int animated)
{
	Object *ob, *ob_iter;
	Base *base = NULL;
	DupliObject *dob;
	DerivedMesh *dm;
	Mesh *me= par->data;
	MLoopUV *mloopuv;
	MPoly *mpoly, *mp;
	MLoop *mloop;
	MVert *mvert;
	float pmat[4][4], imat[3][3], (*orco)[3] = NULL, w;
	int lay, oblay, totface, a;
	Scene *sce = NULL;
	Group *group = NULL;
	GroupObject *go = NULL;
	BMEditMesh *em;
	float ob__obmat[4][4]; /* needed for groups where the object matrix needs to be modified */
	
	/* simple preventing of too deep nested groups */
	if (level>MAX_DUPLI_RECUR) return;
	
	copy_m4_m4(pmat, par->obmat);
	em = me->edit_btmesh;

	if (em) {
		dm= editbmesh_get_derived_cage(scene, par, em, CD_MASK_BAREMESH);
	}
	else {
		dm = mesh_get_derived_deform(scene, par, CD_MASK_BAREMESH);
	}

	totface= dm->getNumPolys(dm);
	mpoly= dm->getPolyArray(dm);
	mloop= dm->getLoopArray(dm);
	mvert= dm->getVertArray(dm);

	if (G.rendering) {

		orco= (float(*)[3])get_mesh_orco_verts(par);
		transform_mesh_orco_verts(me, orco, me->totvert, 0);
		mloopuv= me->mloopuv;
	}
	else {
		orco= NULL;
		mloopuv= NULL;
	}
	
	/* having to loop on scene OR group objects is NOT FUN */
	if (GS(id->name) == ID_SCE) {
		sce = (Scene *)id;
		lay= sce->lay;
		base= sce->base.first;
	}
	else {
		group = (Group *)id;
		lay= group->layer;
		go = group->gobject.first;
	}
	
	/* Start looping on Scene OR Group objects */
	while (base || go) { 
		if (sce) {
			ob_iter= base->object;
			oblay = base->lay;
		}
		else {
			ob_iter= go->ob;
			oblay = ob_iter->lay;
		}
		
		if (lay & oblay && scene->obedit!=ob_iter) {
			ob=ob_iter->parent;
			while (ob) {
				if (ob==par) {
					ob = ob_iter;
	/* End Scene/Group object loop, below is generic */
					
					/* par_space_mat - only used for groups so we can modify the space dupli's are in
					 * when par_space_mat is NULL ob->obmat can be used instead of ob__obmat
					 */
					if (par_space_mat)
						mult_m4_m4m4(ob__obmat, par_space_mat, ob->obmat);
					else
						copy_m4_m4(ob__obmat, ob->obmat);
					
					copy_m3_m4(imat, ob->parentinv);
						
					/* mballs have a different dupli handling */
					if (ob->type!=OB_MBALL) ob->flag |= OB_DONE;	/* doesnt render */

					for (a=0, mp= mpoly; a<totface; a++, mp++) {
						int mv1;
						int mv2;
						int mv3;
						/* int mv4; */ /* UNUSED */
						float *v1;
						float *v2;
						float *v3;
						/* float *v4; */ /* UNUSED */
						float cent[3], quat[4], mat[3][3], mat3[3][3], tmat[4][4], obmat[4][4];
						MLoop *loopstart= mloop + mp->loopstart;

						if (mp->totloop < 3) {
							/* highly unlikely but to be safe */
							continue;
						}
						else {
							v1= mvert[(mv1= loopstart[0].v)].co;
							v2= mvert[(mv2= loopstart[1].v)].co;
							v3= mvert[(mv3= loopstart[2].v)].co;
#if 0
							if (mp->totloop > 3) {
								v4= mvert[(mv4= loopstart[3].v)].co;
							}
#endif
						}

						/* translation */
						mesh_calc_poly_center(mp, loopstart, mvert, cent);

						mul_m4_v3(pmat, cent);
						
						sub_v3_v3v3(cent, cent, pmat[3]);
						add_v3_v3(cent, ob__obmat[3]);
						
						copy_m4_m4(obmat, ob__obmat);
						
						copy_v3_v3(obmat[3], cent);
						
						/* rotation */
						tri_to_quat(quat, v1, v2, v3);
						quat_to_mat3(mat, quat);
						
						/* scale */
						if (par->transflag & OB_DUPLIFACES_SCALE) {
							float size= mesh_calc_poly_area(mp, loopstart, mvert, NULL);
							size= sqrtf(size) * par->dupfacesca;
							mul_m3_fl(mat, size);
						}
						
						copy_m3_m3(mat3, mat);
						mul_m3_m3m3(mat, imat, mat3);
						
						copy_m4_m4(tmat, obmat);
						mul_m4_m4m3(obmat, tmat, mat);
						
						dob= new_dupli_object(lb, ob, obmat, par->lay, a, OB_DUPLIFACES, animated);
						if (G.rendering) {
							w= 1.0f / (float)mp->totloop;

							if (orco) {
								int j;
								for (j = 0; j < mpoly->totloop; j++) {
									madd_v3_v3fl(dob->orco, orco[loopstart[j].v], w);
								}
							}

							if (mloopuv) {
								int j;
								for (j = 0; j < mpoly->totloop; j++) {
									madd_v2_v2fl(dob->orco, mloopuv[loopstart[j].v].uv, w);
								}
							}
						}
						
						if (ob->transflag & OB_DUPLI) {
							float tmpmat[4][4];
							copy_m4_m4(tmpmat, ob->obmat);
							copy_m4_m4(ob->obmat, obmat); /* pretend we are really this mat */
							object_duplilist_recursive((ID *)id, scene, ob, lb, ob->obmat, level+1, animated);
							copy_m4_m4(ob->obmat, tmpmat);
						}
					}
					
					break;
				}
				ob= ob->parent;
			}
		}
		if (sce)	base= base->next;	/* scene loop */
		else		go= go->next;		/* group loop */
	}

	if (orco)
		MEM_freeN(orco);
	
	dm->release(dm);
}

static void new_particle_duplilist(ListBase *lb, ID *id, Scene *scene, Object *par, float par_space_mat[][4], ParticleSystem *psys, int level, int animated)
{
	GroupObject *go;
	Object *ob=NULL, **oblist=NULL, obcopy, *obcopylist=NULL;
	DupliObject *dob;
	ParticleDupliWeight *dw;
	ParticleSettings *part;
	ParticleData *pa;
	ChildParticle *cpa=NULL;
	ParticleKey state;
	ParticleCacheKey *cache;
	float ctime, pa_time, scale = 1.0f;
	float tmat[4][4], mat[4][4], pamat[4][4], vec[3], size=0.0;
	float (*obmat)[4], (*oldobmat)[4];
	int a, b, counter, hair = 0;
	int totpart, totchild, totgroup=0 /*, pa_num */;

	int no_draw_flag = PARS_UNEXIST;

	if (psys==NULL) return;
	
	/* simple preventing of too deep nested groups */
	if (level>MAX_DUPLI_RECUR) return;
	
	part=psys->part;

	if (part==NULL)
		return;

	if (!psys_check_enabled(par, psys))
		return;

	if (G.rendering == 0)
		no_draw_flag |= PARS_NO_DISP;
	
	ctime = BKE_curframe(scene); /* NOTE: in old animsys, used parent object's timeoffset... */

	totpart = psys->totpart;
	totchild = psys->totchild;

	BLI_srandom(31415926 + psys->seed);

	if ((psys->renderdata || part->draw_as==PART_DRAW_REND) && ELEM(part->ren_as, PART_DRAW_OB, PART_DRAW_GR)) {
		ParticleSimulationData sim= {NULL};
		sim.scene= scene;
		sim.ob= par;
		sim.psys= psys;
		sim.psmd= psys_get_modifier(par, psys);
		/* make sure emitter imat is in global coordinates instead of render view coordinates */
		invert_m4_m4(par->imat, par->obmat);

		/* first check for loops (particle system object used as dupli object) */
		if (part->ren_as == PART_DRAW_OB) {
			if (ELEM(part->dup_ob, NULL, par))
				return;
		}
		else { /*PART_DRAW_GR */
			if (part->dup_group == NULL || part->dup_group->gobject.first == NULL)
				return;

			for (go=part->dup_group->gobject.first; go; go=go->next)
				if (go->ob == par)
					return;
		}

		/* if we have a hair particle system, use the path cache */
		if (part->type == PART_HAIR) {
			if (psys->flag & PSYS_HAIR_DONE)
				hair= (totchild == 0 || psys->childcache) && psys->pathcache;
			if (!hair)
				return;
			
			/* we use cache, update totchild according to cached data */
			totchild = psys->totchildcache;
			totpart = psys->totcached;
		}

		psys_check_group_weights(part);

		psys->lattice = psys_get_lattice(&sim);

		/* gather list of objects or single object */
		if (part->ren_as==PART_DRAW_GR) {
			group_handle_recalc_and_update(scene, par, part->dup_group);

			if (part->draw & PART_DRAW_COUNT_GR) {
				for (dw=part->dupliweights.first; dw; dw=dw->next)
					totgroup += dw->count;
			}
			else {
				for (go=part->dup_group->gobject.first; go; go=go->next)
					totgroup++;
			}

			/* we also copy the actual objects to restore afterwards, since
			 * where_is_object_time will change the object which breaks transform */
			oblist = MEM_callocN(totgroup*sizeof(Object *), "dupgroup object list");
			obcopylist = MEM_callocN(totgroup*sizeof(Object), "dupgroup copy list");

			
			if (part->draw & PART_DRAW_COUNT_GR && totgroup) {
				dw = part->dupliweights.first;

				for (a=0; a<totgroup; dw=dw->next) {
					for (b=0; b<dw->count; b++, a++) {
						oblist[a] = dw->ob;
						obcopylist[a] = *dw->ob;
					}
				}
			}
			else {
				go = part->dup_group->gobject.first;
				for (a=0; a<totgroup; a++, go=go->next) {
					oblist[a] = go->ob;
					obcopylist[a] = *go->ob;
				}
			}
		}
		else {
			ob = part->dup_ob;
			obcopy = *ob;
		}

		if (totchild==0 || part->draw & PART_DRAW_PARENT)
			a = 0;
		else
			a = totpart;

		for (pa=psys->particles, counter=0; a<totpart+totchild; a++, pa++, counter++) {
			if (a<totpart) {
				/* handle parent particle */
				if (pa->flag & no_draw_flag)
					continue;

				/* pa_num = pa->num; */ /* UNUSED */
				pa_time = pa->time;
				size = pa->size;
			}
			else {
				/* handle child particle */
				cpa = &psys->child[a - totpart];

				/* pa_num = a; */ /* UNUSED */
				pa_time = psys->particles[cpa->parent].time;
				size = psys_get_child_size(psys, cpa, ctime, NULL);
			}

			/* some hair paths might be non-existent so they can't be used for duplication */
			if (hair &&
			    ((a < totpart && psys->pathcache[a]->steps < 0) ||
			     (a >= totpart && psys->childcache[a-totpart]->steps < 0)))
			{
				continue;
			}

			if (part->ren_as==PART_DRAW_GR) {
				/* prevent divide by zero below [#28336] */
				if (totgroup == 0)
					continue;

				/* for groups, pick the object based on settings */
				if (part->draw&PART_DRAW_RAND_GR)
					b= BLI_rand() % totgroup;
				else
					b= a % totgroup;

				ob = oblist[b];
				obmat = oblist[b]->obmat;
				oldobmat = obcopylist[b].obmat;
			}
			else {
				obmat= ob->obmat;
				oldobmat= obcopy.obmat;
			}

			if (hair) {
				/* hair we handle separate and compute transform based on hair keys */
				if (a < totpart) {
					cache = psys->pathcache[a];
					psys_get_dupli_path_transform(&sim, pa, NULL, cache, pamat, &scale);
				}
				else {
					cache = psys->childcache[a-totpart];
					psys_get_dupli_path_transform(&sim, NULL, cpa, cache, pamat, &scale);
				}

				copy_v3_v3(pamat[3], cache->co);
				pamat[3][3]= 1.0f;
				
			}
			else {
				/* first key */
				state.time = ctime;
				if (psys_get_particle_state(&sim, a, &state, 0) == 0) {
					continue;
				}
				else {
					float tquat[4];
					normalize_qt_qt(tquat, state.rot);
					quat_to_mat4(pamat, tquat);
					copy_v3_v3(pamat[3], state.co);
					pamat[3][3]= 1.0f;
				}
			}

			if (part->ren_as==PART_DRAW_GR && psys->part->draw & PART_DRAW_WHOLE_GR) {
				for (go= part->dup_group->gobject.first, b=0; go; go= go->next, b++) {

					copy_m4_m4(tmat, oblist[b]->obmat);
					/* apply particle scale */
					mul_mat3_m4_fl(tmat, size*scale);
					mul_v3_fl(tmat[3], size*scale);
					/* group dupli offset, should apply after everything else */
					if (!is_zero_v3(part->dup_group->dupli_ofs))
						sub_v3_v3v3(tmat[3], tmat[3], part->dup_group->dupli_ofs);
					/* individual particle transform */
					mult_m4_m4m4(tmat, pamat, tmat);

					if (par_space_mat)
						mult_m4_m4m4(mat, par_space_mat, tmat);
					else
						copy_m4_m4(mat, tmat);

					dob= new_dupli_object(lb, go->ob, mat, par->lay, counter, OB_DUPLIPARTS, animated);
					copy_m4_m4(dob->omat, obcopylist[b].obmat);
					if (G.rendering)
						psys_get_dupli_texture(psys, part, sim.psmd, pa, cpa, dob->uv, dob->orco);
				}
			}
			else {
				/* to give ipos in object correct offset */
				where_is_object_time(scene, ob, ctime-pa_time);

				copy_v3_v3(vec, obmat[3]);
				obmat[3][0] = obmat[3][1] = obmat[3][2] = 0.0f;

				/* particle rotation uses x-axis as the aligned axis, so pre-rotate the object accordingly */
				if ((part->draw & PART_DRAW_ROTATE_OB) == 0) {
					float xvec[3], q[4];
					xvec[0] = -1.f;
					xvec[1] = xvec[2] = 0;
					vec_to_quat(q, xvec, ob->trackflag, ob->upflag);
					quat_to_mat4(obmat, q);
					obmat[3][3]= 1.0f;
				}
				
				/* Normal particles and cached hair live in global space so we need to
				 * remove the real emitter's transformation before 2nd order duplication.
				 */
				if (par_space_mat && GS(id->name) != ID_GR)
					mult_m4_m4m4(mat, psys->imat, pamat);
				else
					copy_m4_m4(mat, pamat);

				mult_m4_m4m4(tmat, mat, obmat);
				mul_mat3_m4_fl(tmat, size*scale);

				if (par_space_mat)
					mult_m4_m4m4(mat, par_space_mat, tmat);
				else
					copy_m4_m4(mat, tmat);

				if (part->draw & PART_DRAW_GLOBAL_OB)
					add_v3_v3v3(mat[3], mat[3], vec);

				dob= new_dupli_object(lb, ob, mat, ob->lay, counter, GS(id->name) == ID_GR ? OB_DUPLIGROUP : OB_DUPLIPARTS, animated);
				copy_m4_m4(dob->omat, oldobmat);
				if (G.rendering)
					psys_get_dupli_texture(psys, part, sim.psmd, pa, cpa, dob->uv, dob->orco);
			}
		}

		/* restore objects since they were changed in where_is_object_time */
		if (part->ren_as==PART_DRAW_GR) {
			for (a=0; a<totgroup; a++)
				*(oblist[a])= obcopylist[a];
		}
		else
			*ob= obcopy;
	}

	/* clean up */
	if (oblist)
		MEM_freeN(oblist);
	if (obcopylist)
		MEM_freeN(obcopylist);

	if (psys->lattice) {
		end_latt_deform(psys->lattice);
		psys->lattice = NULL;
	}
}

static Object *find_family_object(Object **obar, char *family, char ch)
{
	Object *ob;
	int flen;
	
	if ( obar[(int)ch] ) return obar[(int)ch];
	
	flen= strlen(family);
	
	ob= G.main->object.first;
	while (ob) {
		if (ob->id.name[flen + 2] == ch) {
			if ( strncmp(ob->id.name+2, family, flen)==0 ) break;
		}
		ob= ob->id.next;
	}
	
	obar[(int)ch]= ob;
	
	return ob;
}


static void font_duplilist(ListBase *lb, Scene *scene, Object *par, int level, int animated)
{
	Object *ob, *obar[256]= {NULL};
	Curve *cu;
	struct chartrans *ct, *chartransdata;
	float vec[3], obmat[4][4], pmat[4][4], fsize, xof, yof;
	int slen, a;
	
	/* simple preventing of too deep nested groups */
	if (level>MAX_DUPLI_RECUR) return;
	
	copy_m4_m4(pmat, par->obmat);
	
	/* in par the family name is stored, use this to find the other objects */
	
	chartransdata= BKE_text_to_curve(G.main, scene, par, FO_DUPLI);
	if (chartransdata==NULL) return;

	cu= par->data;
	slen= strlen(cu->str);
	fsize= cu->fsize;
	xof= cu->xof;
	yof= cu->yof;
	
	ct= chartransdata;
	
	for (a=0; a<slen; a++, ct++) {
		
		ob= find_family_object(obar, cu->family, cu->str[a]);
		if (ob) {
			vec[0]= fsize*(ct->xof - xof);
			vec[1]= fsize*(ct->yof - yof);
			vec[2]= 0.0;
			
			mul_m4_v3(pmat, vec);
			
			copy_m4_m4(obmat, par->obmat);
			copy_v3_v3(obmat[3], vec);
			
			new_dupli_object(lb, ob, obmat, par->lay, a, OB_DUPLIVERTS, animated);
		}
	}
	
	MEM_freeN(chartransdata);
}

/* ------------- */

static void object_duplilist_recursive(ID *id, Scene *scene, Object *ob, ListBase *duplilist, float par_space_mat[][4], int level, int animated)
{	
	if ((ob->transflag & OB_DUPLI)==0)
		return;
	
	/* Should the dupli's be generated for this object? - Respect restrict flags */
	if (G.rendering) {
		if (ob->restrictflag & OB_RESTRICT_RENDER) {
			return;
		}
	}
	else {
		if (ob->restrictflag & OB_RESTRICT_VIEW) {
			return;
		}
	}

	if (ob->transflag & OB_DUPLIPARTS) {
		ParticleSystem *psys = ob->particlesystem.first;
		for (; psys; psys=psys->next)
			new_particle_duplilist(duplilist, id, scene, ob, par_space_mat, psys, level+1, animated);
	}
	else if (ob->transflag & OB_DUPLIVERTS) {
		if (ob->type==OB_MESH) {
			vertex_duplilist(duplilist, id, scene, ob, par_space_mat, level+1, animated);
		}
		else if (ob->type==OB_FONT) {
			if (GS(id->name)==ID_SCE) { /* TODO - support dupligroups */
				font_duplilist(duplilist, scene, ob, level+1, animated);
			}
		}
	}
	else if (ob->transflag & OB_DUPLIFACES) {
		if (ob->type==OB_MESH)
			face_duplilist(duplilist, id, scene, ob, par_space_mat, level+1, animated);
	}
	else if (ob->transflag & OB_DUPLIFRAMES) {
		if (GS(id->name)==ID_SCE) { /* TODO - support dupligroups */
			frames_duplilist(duplilist, scene, ob, level+1, animated);
		}
	}
	else if (ob->transflag & OB_DUPLIGROUP) {
		DupliObject *dob;
		
		group_duplilist(duplilist, scene, ob, level+1, animated); /* now recursive */

		if (level==0) {
			for (dob= duplilist->first; dob; dob= dob->next)
				if (dob->type == OB_DUPLIGROUP)
					copy_m4_m4(dob->ob->obmat, dob->mat);
		}
	}
}

/* Returns a list of DupliObject
 * note; group dupli's already set transform matrix. see note in group_duplilist() */
ListBase *object_duplilist(Scene *sce, Object *ob)
{
	ListBase *duplilist= MEM_mallocN(sizeof(ListBase), "duplilist");
	duplilist->first= duplilist->last= NULL;
	object_duplilist_recursive((ID *)sce, sce, ob, duplilist, NULL, 0, 0);
	return duplilist;
}

void free_object_duplilist(ListBase *lb)
{
	DupliObject *dob;
	
	/* loop in reverse order, if object is instanced multiple times
	 * the original layer may not really be original otherwise, proper
	 * solution is more complicated */
	for (dob= lb->last; dob; dob= dob->prev) {
		dob->ob->lay= dob->origlay;
		copy_m4_m4(dob->ob->obmat, dob->omat);
	}
	
	BLI_freelistN(lb);
	MEM_freeN(lb);
}

int count_duplilist(Object *ob)
{
	if (ob->transflag & OB_DUPLI) {
		if (ob->transflag & OB_DUPLIVERTS) {
			if (ob->type==OB_MESH) {
				if (ob->transflag & OB_DUPLIVERTS) {
					ParticleSystem *psys = ob->particlesystem.first;
					int pdup=0;

					for (; psys; psys=psys->next)
						pdup += psys->totpart;

					if (pdup==0) {
						Mesh *me= ob->data;
						return me->totvert;
					}
					else
						return pdup;
				}
			}
		}
		else if (ob->transflag & OB_DUPLIFRAMES) {
			int tot= ob->dupend - ob->dupsta; 
			tot/= (ob->dupon+ob->dupoff);
			return tot*ob->dupon;
		}
	}
	return 1;
}