This repository has been archived on 2023-10-09. You can view files and clone it, but cannot push or open issues or pull requests.
Files
blender-archive/source/blender/blenkernel/intern/constraint.c

4831 lines
136 KiB
C
Raw Normal View History

/*
* ***** BEGIN GPL LICENSE BLOCK *****
2002-10-12 11:37:38 +00:00
*
* 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.
2002-10-12 11:37:38 +00:00
*
* 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,
2010-02-12 13:34:04 +00:00
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
2002-10-12 11:37:38 +00:00
*
* The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
* All rights reserved.
*
* The Original Code is: all of this file.
*
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
* Contributor(s): 2007, Joshua Leung, major recode
2002-10-12 11:37:38 +00:00
*
* ***** END GPL LICENSE BLOCK *****
2002-10-12 11:37:38 +00:00
*/
2011-02-27 20:40:57 +00:00
/** \file blender/blenkernel/intern/constraint.c
* \ingroup bke
*/
2002-10-12 11:37:38 +00:00
#include <stdio.h>
#include <stddef.h>
2002-10-12 11:37:38 +00:00
#include <string.h>
#include <math.h>
Merge of trunk into blender 2.5: svn merge https://svn.blender.org/svnroot/bf-blender/trunk/blender -r12987:17416 Issues: * GHOST/X11 had conflicting changes. Some code was added in 2.5, which was later added in trunk also, but reverted partially, specifically revision 16683. I have left out this reversion in the 2.5 branch since I think it is needed there. http://projects.blender.org/plugins/scmsvn/viewcvs.php?view=rev&root=bf-blender&revision=16683 * Scons had various conflicting changes, I decided to go with trunk version for everything except priorities and some library renaming. * In creator.c, there were various fixes and fixes for fixes related to the -w -W and -p options. In 2.5 -w and -W is not coded yet, and -p is done differently. Since this is changed so much, and I don't think those fixes would be needed in 2.5, I've left them out. * Also in creator.c: there was code for a python bugfix where the screen was not initialized when running with -P. The code that initializes the screen there I had to disable, that can't work in 2.5 anymore but left it commented as a reminder. Further I had to disable some new function calls. using src/ and python/, as was done already in this branch, disabled function calls: * bpath.c: error reporting * BME_conversions.c: editmesh conversion functions. * SHD_dynamic: disabled almost completely, there is no python/. * KX_PythonInit.cpp and Ketsji/ build files: Mathutils is not there, disabled. * text.c: clipboard copy call. * object.c: OB_SUPPORT_MATERIAL. * DerivedMesh.c and subsurf_ccg, stipple_quarttone. Still to be done: * Go over files and functions that were moved to a different location but could still use changes that were done in trunk.
2008-11-12 21:16:53 +00:00
#include <float.h>
2002-10-12 11:37:38 +00:00
#include "MEM_guardedalloc.h"
#include "BLI_blenlib.h"
#include "BLI_math.h"
#include "BLI_kdopbvh.h"
#include "BLI_utildefines.h"
2002-10-12 11:37:38 +00:00
#include "DNA_armature_types.h"
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
#include "DNA_camera_types.h"
2002-10-12 11:37:38 +00:00
#include "DNA_constraint_types.h"
#include "DNA_modifier_types.h"
2002-10-12 11:37:38 +00:00
#include "DNA_object_types.h"
#include "DNA_action_types.h"
#include "DNA_curve_types.h"
#include "DNA_mesh_types.h"
#include "DNA_meshdata_types.h"
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
#include "DNA_lattice_types.h"
#include "DNA_scene_types.h"
#include "DNA_text_types.h"
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
#include "DNA_tracking_types.h"
#include "DNA_movieclip_types.h"
2002-10-12 11:37:38 +00:00
2002-10-12 11:37:38 +00:00
#include "BKE_action.h"
#include "BKE_anim.h" /* for the curve calculation part */
2002-10-12 11:37:38 +00:00
#include "BKE_armature.h"
#include "BKE_blender.h"
#include "BKE_bvhutils.h"
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
#include "BKE_camera.h"
2002-10-12 11:37:38 +00:00
#include "BKE_constraint.h"
#include "BKE_displist.h"
#include "BKE_deform.h"
#include "BKE_DerivedMesh.h" /* for geometry targets */
#include "BKE_cdderivedmesh.h" /* for geometry targets */
2002-10-12 11:37:38 +00:00
#include "BKE_object.h"
#include "BKE_ipo.h"
#include "BKE_global.h"
#include "BKE_library.h"
#include "BKE_idprop.h"
#include "BKE_mesh.h"
#include "BKE_shrinkwrap.h"
#include "BKE_mesh.h"
#include "BKE_tessmesh.h"
#include "BKE_tracking.h"
#include "BKE_movieclip.h"
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
#include "BKE_tracking.h"
#include "BKE_movieclip.h"
#ifdef WITH_PYTHON
#include "BPY_extern.h"
#endif
2002-10-12 11:37:38 +00:00
#ifndef M_PI
#define M_PI 3.14159265358979323846
#endif
2002-10-12 11:37:38 +00:00
Result of 2 weeks of quiet coding work in Greece :) Aim was to get a total refresh of the animation system. This is needed because; - we need to upgrade it with 21st century features - current code is spaghetti/hack combo, and hides good design - it should become lag-free with using dependency graphs A full log, with complete code API/structure/design explanation will follow, that's a load of work... so here below the list with hot changes; - The entire object update system (matrices, geometry) is now centralized. Calls to where_is_object and makeDispList are forbidden, instead we tag objects 'changed' and let the depgraph code sort it out - Removed all old "Ika" code - Depgraph is aware of all relationships, including meta balls, constraints, bevelcurve, and so on. - Made depgraph aware of relation types and layers, to do smart flushing of 'changed' events. Nothing gets calculated too often! - Transform uses depgraph to detect changes - On frame-advance, depgraph flushes animated changes Armatures; Almost all armature related code has been fully built from scratch. It now reveils the original design much better, with a very clean implementation, lag free without even calculating each Bone more than once. Result is quite a speedup yes! Important to note is; 1) Armature is data containing the 'rest position' 2) Pose is the changes of rest position, and always on object level. That way more Objects can use same Pose. Also constraints are in Pose 3) Actions only contain the Ipos to change values in Poses. - Bones draw unrotated now - Drawing bones speedup enormously (10-20 times) - Bone selecting in EditMode, selection state is saved for PoseMode, and vice-versa - Undo in editmode - Bone renaming does vertexgroups, constraints, posechannels, actions, for all users of Armature in entire file - Added Bone renaming in NKey panel - Nkey PoseMode shows eulers now - EditMode and PoseMode now have 'active' bone too (last clicked) - Parenting in EditMode' CTRL+P, ALT+P, with nice options! - Pose is added in Outliner now, with showing that constraints are in the Pose, not Armature - Disconnected IK solving from constraints. It's a separate phase now, on top of the full Pose calculations - Pose itself has a dependency graph too, so evaluation order is lag free. TODO NOW; - Rotating in Posemode has incorrect inverse transform (Martin will fix) - Python Bone/Armature/Pose API disabled... needs full recode too (wait for my doc!) - Game engine will need upgrade too - Depgraph code needs revision, cleanup, can be much faster! (But, compliments for Jean-Luc, it works like a charm!) - IK changed, it now doesnt use previous position to advance to next position anymore. That system looks nice (no flips) but is not well suited for NLA and background render. TODO LATER; We now can do loadsa new nifty features as well; like: - Kill PoseMode (can be option for armatures itself) - Make B-Bones (Bezier, Bspline, like for spines) - Move all silly button level edit to 3d window (like CTRL+I = add IK) - Much better & informative drawing - Fix action/nla editors - Put all ipos in Actions (object, mesh key, lamp color) - Add hooks - Null bones - Much more advanced constraints... Bugfixes; - OGL render (view3d header) had wrong first frame on anim render - Ipo 'recording' mode had wrong playback speed - Vertex-key mode now sticks to show 'active key', until frame change -Ton-
2005-07-03 17:35:38 +00:00
/* ************************ Constraints - General Utilities *************************** */
/* These functions here don't act on any specific constraints, and are therefore should/will
* not require any of the special function-pointers afforded by the relevant constraint
* type-info structs.
*/
/* -------------- Naming -------------- */
/* Find the first available, non-duplicate name for a given constraint */
void unique_constraint_name(bConstraint *con, ListBase *list)
Result of 2 weeks of quiet coding work in Greece :) Aim was to get a total refresh of the animation system. This is needed because; - we need to upgrade it with 21st century features - current code is spaghetti/hack combo, and hides good design - it should become lag-free with using dependency graphs A full log, with complete code API/structure/design explanation will follow, that's a load of work... so here below the list with hot changes; - The entire object update system (matrices, geometry) is now centralized. Calls to where_is_object and makeDispList are forbidden, instead we tag objects 'changed' and let the depgraph code sort it out - Removed all old "Ika" code - Depgraph is aware of all relationships, including meta balls, constraints, bevelcurve, and so on. - Made depgraph aware of relation types and layers, to do smart flushing of 'changed' events. Nothing gets calculated too often! - Transform uses depgraph to detect changes - On frame-advance, depgraph flushes animated changes Armatures; Almost all armature related code has been fully built from scratch. It now reveils the original design much better, with a very clean implementation, lag free without even calculating each Bone more than once. Result is quite a speedup yes! Important to note is; 1) Armature is data containing the 'rest position' 2) Pose is the changes of rest position, and always on object level. That way more Objects can use same Pose. Also constraints are in Pose 3) Actions only contain the Ipos to change values in Poses. - Bones draw unrotated now - Drawing bones speedup enormously (10-20 times) - Bone selecting in EditMode, selection state is saved for PoseMode, and vice-versa - Undo in editmode - Bone renaming does vertexgroups, constraints, posechannels, actions, for all users of Armature in entire file - Added Bone renaming in NKey panel - Nkey PoseMode shows eulers now - EditMode and PoseMode now have 'active' bone too (last clicked) - Parenting in EditMode' CTRL+P, ALT+P, with nice options! - Pose is added in Outliner now, with showing that constraints are in the Pose, not Armature - Disconnected IK solving from constraints. It's a separate phase now, on top of the full Pose calculations - Pose itself has a dependency graph too, so evaluation order is lag free. TODO NOW; - Rotating in Posemode has incorrect inverse transform (Martin will fix) - Python Bone/Armature/Pose API disabled... needs full recode too (wait for my doc!) - Game engine will need upgrade too - Depgraph code needs revision, cleanup, can be much faster! (But, compliments for Jean-Luc, it works like a charm!) - IK changed, it now doesnt use previous position to advance to next position anymore. That system looks nice (no flips) but is not well suited for NLA and background render. TODO LATER; We now can do loadsa new nifty features as well; like: - Kill PoseMode (can be option for armatures itself) - Make B-Bones (Bezier, Bspline, like for spines) - Move all silly button level edit to 3d window (like CTRL+I = add IK) - Much better & informative drawing - Fix action/nla editors - Put all ipos in Actions (object, mesh key, lamp color) - Add hooks - Null bones - Much more advanced constraints... Bugfixes; - OGL render (view3d header) had wrong first frame on anim render - Ipo 'recording' mode had wrong playback speed - Vertex-key mode now sticks to show 'active key', until frame change -Ton-
2005-07-03 17:35:38 +00:00
{
BLI_uniquename(list, con, "Const", '.', offsetof(bConstraint, name), sizeof(con->name));
}
/* ----------------- Evaluation Loop Preparation --------------- */
/* package an object/bone for use in constraint evaluation */
/* This function MEM_calloc's a bConstraintOb struct, that will need to be freed after evaluation */
bConstraintOb *constraints_make_evalob (Scene *scene, Object *ob, void *subdata, short datatype)
{
bConstraintOb *cob;
/* create regardless of whether we have any data! */
cob= MEM_callocN(sizeof(bConstraintOb), "bConstraintOb");
/* for system time, part of deglobalization, code nicer later with local time (ton) */
cob->scene= scene;
/* based on type of available data */
switch (datatype) {
case CONSTRAINT_OBTYPE_OBJECT:
{
/* disregard subdata... calloc should set other values right */
if (ob) {
cob->ob = ob;
cob->type = datatype;
2009-09-16 17:43:09 +00:00
cob->rotOrder = EULER_ORDER_DEFAULT; // TODO: when objects have rotation order too, use that
copy_m4_m4(cob->matrix, ob->obmat);
}
else
unit_m4(cob->matrix);
copy_m4_m4(cob->startmat, cob->matrix);
}
break;
case CONSTRAINT_OBTYPE_BONE:
{
/* only set if we have valid bone, otherwise default */
if (ob && subdata) {
cob->ob = ob;
cob->pchan = (bPoseChannel *)subdata;
cob->type = datatype;
2009-09-16 17:43:09 +00:00
if (cob->pchan->rotmode > 0) {
/* should be some type of Euler order */
cob->rotOrder= cob->pchan->rotmode;
}
else {
/* Quats, so eulers should just use default order */
cob->rotOrder= EULER_ORDER_DEFAULT;
}
/* matrix in world-space */
mult_m4_m4m4(cob->matrix, ob->obmat, cob->pchan->pose_mat);
}
else
unit_m4(cob->matrix);
copy_m4_m4(cob->startmat, cob->matrix);
}
break;
default: /* other types not yet handled */
unit_m4(cob->matrix);
unit_m4(cob->startmat);
break;
}
return cob;
}
/* cleanup after constraint evaluation */
void constraints_clear_evalob(bConstraintOb *cob)
{
float delta[4][4], imat[4][4];
/* prevent crashes */
if (cob == NULL)
return;
/* calculate delta of constraints evaluation */
invert_m4_m4(imat, cob->startmat);
mult_m4_m4m4(delta, cob->matrix, imat);
/* copy matrices back to source */
switch (cob->type) {
case CONSTRAINT_OBTYPE_OBJECT:
{
/* cob->ob might not exist! */
if (cob->ob) {
/* copy new ob-matrix back to owner */
copy_m4_m4(cob->ob->obmat, cob->matrix);
/* copy inverse of delta back to owner */
invert_m4_m4(cob->ob->constinv, delta);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
}
break;
case CONSTRAINT_OBTYPE_BONE:
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
{
/* cob->ob or cob->pchan might not exist */
if (cob->ob && cob->pchan) {
/* copy new pose-matrix back to owner */
mult_m4_m4m4(cob->pchan->pose_mat, cob->ob->imat, cob->matrix);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* copy inverse of delta back to owner */
invert_m4_m4(cob->pchan->constinv, delta);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
}
break;
}
/* free tempolary struct */
MEM_freeN(cob);
}
/* -------------- Space-Conversion API -------------- */
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
#if 0 /* XXX Old code, does the same as one in armature.c, will remove it later. */
static void constraint_pchan_diff_mat(bPoseChannel *pchan, float diff_mat[4][4])
{
if (pchan->parent) {
float offs_bone[4][4];
/* construct offs_bone the same way it is done in armature.c */
copy_m4_m3(offs_bone, pchan->bone->bone_mat);
copy_v3_v3(offs_bone[3], pchan->bone->head);
offs_bone[3][1] += pchan->bone->parent->length;
if (pchan->bone->flag & BONE_HINGE) {
/* pose_mat = par_pose-space_location * chan_mat */
float tmat[4][4];
/* the rotation of the parent restposition */
copy_m4_m4(tmat, pchan->bone->parent->arm_mat);
/* the location of actual parent transform */
copy_v3_v3(tmat[3], offs_bone[3]);
zero_v3(offs_bone[3]);
mul_m4_v3(pchan->parent->pose_mat, tmat[3]);
mult_m4_m4m4(diff_mat, tmat, offs_bone);
}
else {
/* pose_mat = par_pose_mat * bone_mat * chan_mat */
if (pchan->bone->flag & BONE_NO_SCALE) {
float tmat[4][4];
copy_m4_m4(tmat, pchan->parent->pose_mat);
normalize_m4(tmat);
mult_m4_m4m4(diff_mat, tmat, offs_bone);
}
else {
mult_m4_m4m4(diff_mat, pchan->parent->pose_mat, offs_bone);
}
}
}
else {
/* pose_mat = chan_mat * arm_mat */
copy_m4_m4(diff_mat, pchan->bone->arm_mat);
}
}
#endif
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* This function is responsible for the correct transformations/conversions
* of a matrix from one space to another for constraint evaluation.
* For now, this is only implemented for Objects and PoseChannels.
*/
void constraint_mat_convertspace(Object *ob, bPoseChannel *pchan, float mat[][4], short from, short to)
2002-10-12 11:37:38 +00:00
{
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
float diff_mat[4][4];
float imat[4][4];
/* prevent crashes in these unlikely events */
if (ob==NULL || mat==NULL) return;
/* optimize trick - check if need to do anything */
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
if (from == to) return;
/* are we dealing with pose-channels or objects */
if (pchan) {
/* pose channels */
switch (from) {
case CONSTRAINT_SPACE_WORLD: /* ---------- FROM WORLDSPACE ---------- */
{
/* world to pose */
invert_m4_m4(imat, ob->obmat);
mult_m4_m4m4(mat, imat, mat);
/* use pose-space as stepping stone for other spaces... */
if (ELEM(to, CONSTRAINT_SPACE_LOCAL, CONSTRAINT_SPACE_PARLOCAL)) {
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* call self with slightly different values */
constraint_mat_convertspace(ob, pchan, mat, CONSTRAINT_SPACE_POSE, to);
}
}
break;
case CONSTRAINT_SPACE_POSE: /* ---------- FROM POSESPACE ---------- */
{
/* pose to world */
if (to == CONSTRAINT_SPACE_WORLD) {
mult_m4_m4m4(mat, ob->obmat, mat);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
/* pose to local */
else if (to == CONSTRAINT_SPACE_LOCAL) {
if (pchan->bone) {
armature_mat_pose_to_bone(pchan, mat, mat);
#if 0 /* XXX Old code, will remove it later. */
constraint_pchan_diff_mat(pchan, diff_mat);
invert_m4_m4(imat, diff_mat);
mult_m4_m4m4(mat, imat, mat);
/* override with local location */
if ((pchan->parent) && (pchan->bone->flag & BONE_NO_LOCAL_LOCATION)) {
armature_mat_pose_to_bone_ex(ob, pchan, pchan->pose_mat, tempmat);
copy_v3_v3(mat[3], tempmat[3]);
}
#endif
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
}
/* pose to local with parent */
else if (to == CONSTRAINT_SPACE_PARLOCAL) {
if (pchan->bone) {
invert_m4_m4(imat, pchan->bone->arm_mat);
mult_m4_m4m4(mat, imat, mat);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
}
}
break;
case CONSTRAINT_SPACE_LOCAL: /* ------------ FROM LOCALSPACE --------- */
{
/* local to pose - do inverse procedure that was done for pose to local */
if (pchan->bone) {
/* we need the posespace_matrix = local_matrix + (parent_posespace_matrix + restpos) */
armature_mat_bone_to_pose(pchan, mat, mat);
#if 0
constraint_pchan_diff_mat(pchan, diff_mat);
mult_m4_m4m4(mat, diff_mat, mat);
#endif
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
/* use pose-space as stepping stone for other spaces */
if (ELEM(to, CONSTRAINT_SPACE_WORLD, CONSTRAINT_SPACE_PARLOCAL)) {
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* call self with slightly different values */
constraint_mat_convertspace(ob, pchan, mat, CONSTRAINT_SPACE_POSE, to);
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
break;
case CONSTRAINT_SPACE_PARLOCAL: /* -------------- FROM LOCAL WITH PARENT ---------- */
{
/* local + parent to pose */
if (pchan->bone) {
copy_m4_m4(diff_mat, pchan->bone->arm_mat);
mult_m4_m4m4(mat, mat, diff_mat);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
/* use pose-space as stepping stone for other spaces */
if (ELEM(to, CONSTRAINT_SPACE_WORLD, CONSTRAINT_SPACE_LOCAL)) {
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* call self with slightly different values */
constraint_mat_convertspace(ob, pchan, mat, CONSTRAINT_SPACE_POSE, to);
}
}
break;
}
}
else {
/* objects */
if (from==CONSTRAINT_SPACE_WORLD && to==CONSTRAINT_SPACE_LOCAL) {
/* check if object has a parent */
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
if (ob->parent) {
/* 'subtract' parent's effects from owner */
mult_m4_m4m4(diff_mat, ob->parent->obmat, ob->parentinv);
invert_m4_m4(imat, diff_mat);
mult_m4_m4m4(mat, imat, mat);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
else {
/* Local space in this case will have to be defined as local to the owner's
* transform-property-rotated axes. So subtract this rotation component.
*/
BKE_object_to_mat4(ob, diff_mat);
normalize_m4(diff_mat);
zero_v3(diff_mat[3]);
invert_m4_m4(imat, diff_mat);
mult_m4_m4m4(mat, imat, mat);
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
else if (from==CONSTRAINT_SPACE_LOCAL && to==CONSTRAINT_SPACE_WORLD) {
/* check that object has a parent - otherwise this won't work */
if (ob->parent) {
/* 'add' parent's effect back to owner */
mult_m4_m4m4(diff_mat, ob->parent->obmat, ob->parentinv);
mult_m4_m4m4(mat, diff_mat, mat);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
else {
/* Local space in this case will have to be defined as local to the owner's
* transform-property-rotated axes. So add back this rotation component.
*/
BKE_object_to_mat4(ob, diff_mat);
normalize_m4(diff_mat);
zero_v3(diff_mat[3]);
mult_m4_m4m4(mat, diff_mat, mat);
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
}
}
/* ------------ General Target Matrix Tools ---------- */
2002-10-12 11:37:38 +00:00
/* function that sets the given matrix based on given vertex group in mesh */
2011-06-15 14:06:25 +00:00
static void contarget_get_mesh_mat (Object *ob, const char *substring, float mat[][4])
{
DerivedMesh *dm = NULL;
Mesh *me= ob->data;
BMEditMesh *em = me->edit_btmesh;
2010-08-15 15:14:08 +00:00
float vec[3] = {0.0f, 0.0f, 0.0f};
float normal[3] = {0.0f, 0.0f, 0.0f}, plane[3];
float imat[3][3], tmat[3][3];
const int defgroup= defgroup_name_index(ob, substring);
short freeDM = 0;
/* initialize target matrix using target matrix */
copy_m4_m4(mat, ob->obmat);
/* get index of vertex group */
if (defgroup == -1) return;
/* get DerivedMesh */
if (em) {
/* target is in editmode, so get a special derived mesh */
dm = CDDM_from_BMEditMesh(em, ob->data, FALSE, FALSE);
freeDM= 1;
}
else {
/* when not in EditMode, use the 'final' derived mesh, depsgraph
* ensures we build with CD_MDEFORMVERT layer
*/
dm = (DerivedMesh *)ob->derivedFinal;
}
/* only continue if there's a valid DerivedMesh */
if (dm) {
MDeformVert *dvert = dm->getVertDataArray(dm, CD_MDEFORMVERT);
int numVerts = dm->getNumVerts(dm);
int i, count = 0;
float co[3], nor[3];
/* check that dvert is a valid pointers (just in case) */
if (dvert) {
MDeformVert *dv= dvert;
/* get the average of all verts with that are in the vertex-group */
for (i = 0; i < numVerts; i++, dv++) {
MDeformWeight *dw= defvert_find_index(dv, defgroup);
if (dw && dw->weight != 0.0f) {
dm->getVertCo(dm, i, co);
dm->getVertNo(dm, i, nor);
add_v3_v3(vec, co);
add_v3_v3(normal, nor);
count++;
}
}
/* calculate averages of normal and coordinates */
if (count > 0) {
mul_v3_fl(vec, 1.0f / count);
mul_v3_fl(normal, 1.0f / count);
}
/* derive the rotation from the average normal:
* - code taken from transform_manipulator.c,
* calc_manipulator_stats, V3D_MANIP_NORMAL case
*/
/* we need the transpose of the inverse for a normal... */
copy_m3_m4(imat, ob->obmat);
invert_m3_m3(tmat, imat);
transpose_m3(tmat);
mul_m3_v3(tmat, normal);
normalize_v3(normal);
2010-08-15 15:14:08 +00:00
copy_v3_v3(plane, tmat[1]);
cross_v3_v3v3(mat[0], normal, plane);
if (len_v3(mat[0]) < 1e-3f) {
copy_v3_v3(plane, tmat[0]);
cross_v3_v3v3(mat[0], normal, plane);
}
copy_v3_v3(mat[2], normal);
cross_v3_v3v3(mat[1], mat[2], mat[0]);
normalize_m4(mat);
/* apply the average coordinate as the new location */
2010-08-15 15:14:08 +00:00
mul_v3_m4v3(mat[3], ob->obmat, vec);
}
}
/* free temporary DerivedMesh created (in EditMode case) */
if (dm && freeDM)
dm->release(dm);
}
/* function that sets the given matrix based on given vertex group in lattice */
static void contarget_get_lattice_mat (Object *ob, const char *substring, float mat[][4])
{
Lattice *lt= (Lattice *)ob->data;
DispList *dl = find_displist(&ob->disp, DL_VERTS);
float *co = dl?dl->verts:NULL;
BPoint *bp = lt->def;
MDeformVert *dv = lt->dvert;
int tot_verts= lt->pntsu*lt->pntsv*lt->pntsw;
float vec[3]= {0.0f, 0.0f, 0.0f}, tvec[3];
int grouped=0;
int i, n;
const int defgroup= defgroup_name_index(ob, substring);
/* initialize target matrix using target matrix */
copy_m4_m4(mat, ob->obmat);
/* get index of vertex group */
if (defgroup == -1) return;
if (dv == NULL) return;
/* 1. Loop through control-points checking if in nominated vertex-group.
* 2. If it is, add it to vec to find the average point.
*/
for (i=0; i < tot_verts; i++, dv++) {
for (n= 0; n < dv->totweight; n++) {
MDeformWeight *dw= defvert_find_index(dv, defgroup);
if (dw && dw->weight > 0.0f) {
/* copy coordinates of point to temporary vector, then add to find average */
memcpy(tvec, co ? co : bp->vec, 3 * sizeof(float));
add_v3_v3(vec, tvec);
grouped++;
}
}
/* advance pointer to coordinate data */
if (co) co += 3;
else bp++;
}
/* find average location, then multiply by ob->obmat to find world-space location */
if (grouped)
mul_v3_fl(vec, 1.0f / grouped);
mul_v3_m4v3(tvec, ob->obmat, vec);
/* copy new location to matrix */
2010-08-15 15:14:08 +00:00
copy_v3_v3(mat[3], tvec);
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* generic function to get the appropriate matrix for most target cases */
/* The cases where the target can be object data have not been implemented */
2011-06-15 14:06:25 +00:00
static void constraint_target_to_mat4 (Object *ob, const char *substring, float mat[][4], short from, short to, float headtail)
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
{
2002-10-12 11:37:38 +00:00
/* Case OBJECT */
Result of 2 weeks of quiet coding work in Greece :) Aim was to get a total refresh of the animation system. This is needed because; - we need to upgrade it with 21st century features - current code is spaghetti/hack combo, and hides good design - it should become lag-free with using dependency graphs A full log, with complete code API/structure/design explanation will follow, that's a load of work... so here below the list with hot changes; - The entire object update system (matrices, geometry) is now centralized. Calls to where_is_object and makeDispList are forbidden, instead we tag objects 'changed' and let the depgraph code sort it out - Removed all old "Ika" code - Depgraph is aware of all relationships, including meta balls, constraints, bevelcurve, and so on. - Made depgraph aware of relation types and layers, to do smart flushing of 'changed' events. Nothing gets calculated too often! - Transform uses depgraph to detect changes - On frame-advance, depgraph flushes animated changes Armatures; Almost all armature related code has been fully built from scratch. It now reveils the original design much better, with a very clean implementation, lag free without even calculating each Bone more than once. Result is quite a speedup yes! Important to note is; 1) Armature is data containing the 'rest position' 2) Pose is the changes of rest position, and always on object level. That way more Objects can use same Pose. Also constraints are in Pose 3) Actions only contain the Ipos to change values in Poses. - Bones draw unrotated now - Drawing bones speedup enormously (10-20 times) - Bone selecting in EditMode, selection state is saved for PoseMode, and vice-versa - Undo in editmode - Bone renaming does vertexgroups, constraints, posechannels, actions, for all users of Armature in entire file - Added Bone renaming in NKey panel - Nkey PoseMode shows eulers now - EditMode and PoseMode now have 'active' bone too (last clicked) - Parenting in EditMode' CTRL+P, ALT+P, with nice options! - Pose is added in Outliner now, with showing that constraints are in the Pose, not Armature - Disconnected IK solving from constraints. It's a separate phase now, on top of the full Pose calculations - Pose itself has a dependency graph too, so evaluation order is lag free. TODO NOW; - Rotating in Posemode has incorrect inverse transform (Martin will fix) - Python Bone/Armature/Pose API disabled... needs full recode too (wait for my doc!) - Game engine will need upgrade too - Depgraph code needs revision, cleanup, can be much faster! (But, compliments for Jean-Luc, it works like a charm!) - IK changed, it now doesnt use previous position to advance to next position anymore. That system looks nice (no flips) but is not well suited for NLA and background render. TODO LATER; We now can do loadsa new nifty features as well; like: - Kill PoseMode (can be option for armatures itself) - Make B-Bones (Bezier, Bspline, like for spines) - Move all silly button level edit to 3d window (like CTRL+I = add IK) - Much better & informative drawing - Fix action/nla editors - Put all ipos in Actions (object, mesh key, lamp color) - Add hooks - Null bones - Much more advanced constraints... Bugfixes; - OGL render (view3d header) had wrong first frame on anim render - Ipo 'recording' mode had wrong playback speed - Vertex-key mode now sticks to show 'active key', until frame change -Ton-
2005-07-03 17:35:38 +00:00
if (!strlen(substring)) {
copy_m4_m4(mat, ob->obmat);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
constraint_mat_convertspace(ob, NULL, mat, from, to);
}
/* Case VERTEXGROUP */
/* Current method just takes the average location of all the points in the
* VertexGroup, and uses that as the location value of the targets. Where
* possible, the orientation will also be calculated, by calculating an
* 'average' vertex normal, and deriving the rotation from that.
*
* NOTE: EditMode is not currently supported, and will most likely remain that
* way as constraints can only really affect things on object/bone level.
*/
else if (ob->type == OB_MESH) {
2011-06-15 14:06:25 +00:00
contarget_get_mesh_mat(ob, substring, mat);
constraint_mat_convertspace(ob, NULL, mat, from, to);
}
else if (ob->type == OB_LATTICE) {
contarget_get_lattice_mat(ob, substring, mat);
constraint_mat_convertspace(ob, NULL, mat, from, to);
2002-10-12 11:37:38 +00:00
}
/* Case BONE */
else {
Result of 2 weeks of quiet coding work in Greece :) Aim was to get a total refresh of the animation system. This is needed because; - we need to upgrade it with 21st century features - current code is spaghetti/hack combo, and hides good design - it should become lag-free with using dependency graphs A full log, with complete code API/structure/design explanation will follow, that's a load of work... so here below the list with hot changes; - The entire object update system (matrices, geometry) is now centralized. Calls to where_is_object and makeDispList are forbidden, instead we tag objects 'changed' and let the depgraph code sort it out - Removed all old "Ika" code - Depgraph is aware of all relationships, including meta balls, constraints, bevelcurve, and so on. - Made depgraph aware of relation types and layers, to do smart flushing of 'changed' events. Nothing gets calculated too often! - Transform uses depgraph to detect changes - On frame-advance, depgraph flushes animated changes Armatures; Almost all armature related code has been fully built from scratch. It now reveils the original design much better, with a very clean implementation, lag free without even calculating each Bone more than once. Result is quite a speedup yes! Important to note is; 1) Armature is data containing the 'rest position' 2) Pose is the changes of rest position, and always on object level. That way more Objects can use same Pose. Also constraints are in Pose 3) Actions only contain the Ipos to change values in Poses. - Bones draw unrotated now - Drawing bones speedup enormously (10-20 times) - Bone selecting in EditMode, selection state is saved for PoseMode, and vice-versa - Undo in editmode - Bone renaming does vertexgroups, constraints, posechannels, actions, for all users of Armature in entire file - Added Bone renaming in NKey panel - Nkey PoseMode shows eulers now - EditMode and PoseMode now have 'active' bone too (last clicked) - Parenting in EditMode' CTRL+P, ALT+P, with nice options! - Pose is added in Outliner now, with showing that constraints are in the Pose, not Armature - Disconnected IK solving from constraints. It's a separate phase now, on top of the full Pose calculations - Pose itself has a dependency graph too, so evaluation order is lag free. TODO NOW; - Rotating in Posemode has incorrect inverse transform (Martin will fix) - Python Bone/Armature/Pose API disabled... needs full recode too (wait for my doc!) - Game engine will need upgrade too - Depgraph code needs revision, cleanup, can be much faster! (But, compliments for Jean-Luc, it works like a charm!) - IK changed, it now doesnt use previous position to advance to next position anymore. That system looks nice (no flips) but is not well suited for NLA and background render. TODO LATER; We now can do loadsa new nifty features as well; like: - Kill PoseMode (can be option for armatures itself) - Make B-Bones (Bezier, Bspline, like for spines) - Move all silly button level edit to 3d window (like CTRL+I = add IK) - Much better & informative drawing - Fix action/nla editors - Put all ipos in Actions (object, mesh key, lamp color) - Add hooks - Null bones - Much more advanced constraints... Bugfixes; - OGL render (view3d header) had wrong first frame on anim render - Ipo 'recording' mode had wrong playback speed - Vertex-key mode now sticks to show 'active key', until frame change -Ton-
2005-07-03 17:35:38 +00:00
bPoseChannel *pchan;
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
Result of 2 weeks of quiet coding work in Greece :) Aim was to get a total refresh of the animation system. This is needed because; - we need to upgrade it with 21st century features - current code is spaghetti/hack combo, and hides good design - it should become lag-free with using dependency graphs A full log, with complete code API/structure/design explanation will follow, that's a load of work... so here below the list with hot changes; - The entire object update system (matrices, geometry) is now centralized. Calls to where_is_object and makeDispList are forbidden, instead we tag objects 'changed' and let the depgraph code sort it out - Removed all old "Ika" code - Depgraph is aware of all relationships, including meta balls, constraints, bevelcurve, and so on. - Made depgraph aware of relation types and layers, to do smart flushing of 'changed' events. Nothing gets calculated too often! - Transform uses depgraph to detect changes - On frame-advance, depgraph flushes animated changes Armatures; Almost all armature related code has been fully built from scratch. It now reveils the original design much better, with a very clean implementation, lag free without even calculating each Bone more than once. Result is quite a speedup yes! Important to note is; 1) Armature is data containing the 'rest position' 2) Pose is the changes of rest position, and always on object level. That way more Objects can use same Pose. Also constraints are in Pose 3) Actions only contain the Ipos to change values in Poses. - Bones draw unrotated now - Drawing bones speedup enormously (10-20 times) - Bone selecting in EditMode, selection state is saved for PoseMode, and vice-versa - Undo in editmode - Bone renaming does vertexgroups, constraints, posechannels, actions, for all users of Armature in entire file - Added Bone renaming in NKey panel - Nkey PoseMode shows eulers now - EditMode and PoseMode now have 'active' bone too (last clicked) - Parenting in EditMode' CTRL+P, ALT+P, with nice options! - Pose is added in Outliner now, with showing that constraints are in the Pose, not Armature - Disconnected IK solving from constraints. It's a separate phase now, on top of the full Pose calculations - Pose itself has a dependency graph too, so evaluation order is lag free. TODO NOW; - Rotating in Posemode has incorrect inverse transform (Martin will fix) - Python Bone/Armature/Pose API disabled... needs full recode too (wait for my doc!) - Game engine will need upgrade too - Depgraph code needs revision, cleanup, can be much faster! (But, compliments for Jean-Luc, it works like a charm!) - IK changed, it now doesnt use previous position to advance to next position anymore. That system looks nice (no flips) but is not well suited for NLA and background render. TODO LATER; We now can do loadsa new nifty features as well; like: - Kill PoseMode (can be option for armatures itself) - Make B-Bones (Bezier, Bspline, like for spines) - Move all silly button level edit to 3d window (like CTRL+I = add IK) - Much better & informative drawing - Fix action/nla editors - Put all ipos in Actions (object, mesh key, lamp color) - Add hooks - Null bones - Much more advanced constraints... Bugfixes; - OGL render (view3d header) had wrong first frame on anim render - Ipo 'recording' mode had wrong playback speed - Vertex-key mode now sticks to show 'active key', until frame change -Ton-
2005-07-03 17:35:38 +00:00
pchan = get_pose_channel(ob->pose, substring);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
if (pchan) {
/* Multiply the PoseSpace accumulation/final matrix for this
* PoseChannel by the Armature Object's Matrix to get a worldspace
* matrix.
Result of 2 weeks of quiet coding work in Greece :) Aim was to get a total refresh of the animation system. This is needed because; - we need to upgrade it with 21st century features - current code is spaghetti/hack combo, and hides good design - it should become lag-free with using dependency graphs A full log, with complete code API/structure/design explanation will follow, that's a load of work... so here below the list with hot changes; - The entire object update system (matrices, geometry) is now centralized. Calls to where_is_object and makeDispList are forbidden, instead we tag objects 'changed' and let the depgraph code sort it out - Removed all old "Ika" code - Depgraph is aware of all relationships, including meta balls, constraints, bevelcurve, and so on. - Made depgraph aware of relation types and layers, to do smart flushing of 'changed' events. Nothing gets calculated too often! - Transform uses depgraph to detect changes - On frame-advance, depgraph flushes animated changes Armatures; Almost all armature related code has been fully built from scratch. It now reveils the original design much better, with a very clean implementation, lag free without even calculating each Bone more than once. Result is quite a speedup yes! Important to note is; 1) Armature is data containing the 'rest position' 2) Pose is the changes of rest position, and always on object level. That way more Objects can use same Pose. Also constraints are in Pose 3) Actions only contain the Ipos to change values in Poses. - Bones draw unrotated now - Drawing bones speedup enormously (10-20 times) - Bone selecting in EditMode, selection state is saved for PoseMode, and vice-versa - Undo in editmode - Bone renaming does vertexgroups, constraints, posechannels, actions, for all users of Armature in entire file - Added Bone renaming in NKey panel - Nkey PoseMode shows eulers now - EditMode and PoseMode now have 'active' bone too (last clicked) - Parenting in EditMode' CTRL+P, ALT+P, with nice options! - Pose is added in Outliner now, with showing that constraints are in the Pose, not Armature - Disconnected IK solving from constraints. It's a separate phase now, on top of the full Pose calculations - Pose itself has a dependency graph too, so evaluation order is lag free. TODO NOW; - Rotating in Posemode has incorrect inverse transform (Martin will fix) - Python Bone/Armature/Pose API disabled... needs full recode too (wait for my doc!) - Game engine will need upgrade too - Depgraph code needs revision, cleanup, can be much faster! (But, compliments for Jean-Luc, it works like a charm!) - IK changed, it now doesnt use previous position to advance to next position anymore. That system looks nice (no flips) but is not well suited for NLA and background render. TODO LATER; We now can do loadsa new nifty features as well; like: - Kill PoseMode (can be option for armatures itself) - Make B-Bones (Bezier, Bspline, like for spines) - Move all silly button level edit to 3d window (like CTRL+I = add IK) - Much better & informative drawing - Fix action/nla editors - Put all ipos in Actions (object, mesh key, lamp color) - Add hooks - Null bones - Much more advanced constraints... Bugfixes; - OGL render (view3d header) had wrong first frame on anim render - Ipo 'recording' mode had wrong playback speed - Vertex-key mode now sticks to show 'active key', until frame change -Ton-
2005-07-03 17:35:38 +00:00
*/
if (headtail < 0.000001f) {
/* skip length interpolation if set to head */
mult_m4_m4m4(mat, ob->obmat, pchan->pose_mat);
}
else {
float tempmat[4][4], loc[3];
/* interpolate along length of bone */
interp_v3_v3v3(loc, pchan->pose_head, pchan->pose_tail, headtail);
/* use interpolated distance for subtarget */
copy_m4_m4(tempmat, pchan->pose_mat);
copy_v3_v3(tempmat[3], loc);
mult_m4_m4m4(mat, ob->obmat, tempmat);
}
2002-10-12 11:37:38 +00:00
}
else
copy_m4_m4(mat, ob->obmat);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* convert matrix space as required */
constraint_mat_convertspace(ob, pchan, mat, from, to);
}
}
/* ************************* Specific Constraints ***************************** */
/* Each constraint defines a set of functions, which will be called at the appropriate
* times. In addition to this, each constraint should have a type-info struct, where
* its functions are attached for use.
*/
/* Template for type-info data:
* - make a copy of this when creating new constraints, and just change the functions
* pointed to as necessary
* - although the naming of functions doesn't matter, it would help for code
* readability, to follow the same naming convention as is presented here
* - any functions that a constraint doesn't need to define, don't define
* for such cases, just use NULL
* - these should be defined after all the functions have been defined, so that
* forward-definitions/prototypes don't need to be used!
* - keep this copy #if-def'd so that future constraints can get based off this
*/
#if 0
static bConstraintTypeInfo CTI_CONSTRNAME = {
CONSTRAINT_TYPE_CONSTRNAME, /* type */
sizeof(bConstrNameConstraint), /* size */
"ConstrName", /* name */
"bConstrNameConstraint", /* struct name */
constrname_free, /* free data */
constrname_relink, /* relink data */
constrname_id_looper, /* id looper */
constrname_copy, /* copy data */
constrname_new_data, /* new data */
constrname_get_tars, /* get constraint targets */
constrname_flush_tars, /* flush constraint targets */
constrname_get_tarmat, /* get target matrix */
constrname_evaluate /* evaluate */
};
#endif
/* This function should be used for the get_target_matrix member of all
* constraints that are not picky about what happens to their target matrix.
*/
2011-06-15 14:06:25 +00:00
static void default_get_tarmat (bConstraint *con, bConstraintOb *UNUSED(cob), bConstraintTarget *ct, float UNUSED(ctime))
{
if (VALID_CONS_TARGET(ct))
2011-06-15 14:06:25 +00:00
constraint_target_to_mat4(ct->tar, ct->subtarget, ct->matrix, CONSTRAINT_SPACE_WORLD, ct->space, con->headtail);
else if (ct)
unit_m4(ct->matrix);
}
/* This following macro should be used for all standard single-target *_get_tars functions
2012-03-01 12:20:18 +00:00
* to save typing and reduce maintenance woes.
* (Hopefully all compilers will be happy with the lines with just a space on them. Those are
* really just to help this code easier to read)
*/
2009-09-16 17:43:09 +00:00
// TODO: cope with getting rotation order...
#define SINGLETARGET_GET_TARS(con, datatar, datasubtarget, ct, list) \
{ \
ct= MEM_callocN(sizeof(bConstraintTarget), "tempConstraintTarget"); \
\
ct->tar= datatar; \
BLI_strncpy(ct->subtarget, datasubtarget, sizeof(ct->subtarget)); \
ct->space= con->tarspace; \
ct->flag= CONSTRAINT_TAR_TEMP; \
\
if (ct->tar) { \
if ((ct->tar->type==OB_ARMATURE) && (ct->subtarget[0])) { \
bPoseChannel *pchan= get_pose_channel(ct->tar->pose, ct->subtarget); \
ct->type = CONSTRAINT_OBTYPE_BONE; \
ct->rotOrder= (pchan) ? (pchan->rotmode) : EULER_ORDER_DEFAULT; \
}\
else if (OB_TYPE_SUPPORT_VGROUP(ct->tar->type) && (ct->subtarget[0])) { \
ct->type = CONSTRAINT_OBTYPE_VERT; \
ct->rotOrder = EULER_ORDER_DEFAULT; \
} \
else {\
ct->type = CONSTRAINT_OBTYPE_OBJECT; \
ct->rotOrder= ct->tar->rotmode; \
} \
} \
\
BLI_addtail(list, ct); \
}
/* This following macro should be used for all standard single-target *_get_tars functions
2012-03-01 12:20:18 +00:00
* to save typing and reduce maintenance woes. It does not do the subtarget related operations
* (Hopefully all compilers will be happy with the lines with just a space on them. Those are
* really just to help this code easier to read)
*/
2009-09-16 17:43:09 +00:00
// TODO: cope with getting rotation order...
#define SINGLETARGETNS_GET_TARS(con, datatar, ct, list) \
{ \
ct= MEM_callocN(sizeof(bConstraintTarget), "tempConstraintTarget"); \
\
ct->tar= datatar; \
ct->space= con->tarspace; \
ct->flag= CONSTRAINT_TAR_TEMP; \
\
if (ct->tar) ct->type = CONSTRAINT_OBTYPE_OBJECT; \
\
BLI_addtail(list, ct); \
}
/* This following macro should be used for all standard single-target *_flush_tars functions
2012-03-01 12:20:18 +00:00
* to save typing and reduce maintenance woes.
* Note: the pointer to ct will be changed to point to the next in the list (as it gets removed)
* (Hopefully all compilers will be happy with the lines with just a space on them. Those are
* really just to help this code easier to read)
*/
#define SINGLETARGET_FLUSH_TARS(con, datatar, datasubtarget, ct, list, nocopy) \
{ \
if (ct) { \
bConstraintTarget *ctn = ct->next; \
if (nocopy == 0) { \
datatar= ct->tar; \
BLI_strncpy(datasubtarget, ct->subtarget, sizeof(datasubtarget)); \
con->tarspace= (char)ct->space; \
} \
\
BLI_freelinkN(list, ct); \
ct= ctn; \
} \
}
/* This following macro should be used for all standard single-target *_flush_tars functions
2012-03-01 12:20:18 +00:00
* to save typing and reduce maintenance woes. It does not do the subtarget related operations.
* Note: the pointer to ct will be changed to point to the next in the list (as it gets removed)
* (Hopefully all compilers will be happy with the lines with just a space on them. Those are
* really just to help this code easier to read)
*/
#define SINGLETARGETNS_FLUSH_TARS(con, datatar, ct, list, nocopy) \
{ \
if (ct) { \
bConstraintTarget *ctn = ct->next; \
if (nocopy == 0) { \
datatar= ct->tar; \
con->tarspace= (char)ct->space; \
} \
\
BLI_freelinkN(list, ct); \
ct= ctn; \
} \
}
/* --------- ChildOf Constraint ------------ */
static void childof_new_data (void *cdata)
{
bChildOfConstraint *data= (bChildOfConstraint *)cdata;
data->flag = (CHILDOF_LOCX | CHILDOF_LOCY | CHILDOF_LOCZ |
CHILDOF_ROTX |CHILDOF_ROTY | CHILDOF_ROTZ |
CHILDOF_SIZEX | CHILDOF_SIZEY | CHILDOF_SIZEZ);
unit_m4(data->invmat);
}
static void childof_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bChildOfConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int childof_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bChildOfConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void childof_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bChildOfConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void childof_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bChildOfConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
/* only evaluate if there is a target */
if (VALID_CONS_TARGET(ct)) {
float parmat[4][4];
/* simple matrix parenting */
if (data->flag == CHILDOF_ALL) {
/* multiply target (parent matrix) by offset (parent inverse) to get
* the effect of the parent that will be exherted on the owner
*/
mult_m4_m4m4(parmat, ct->matrix, data->invmat);
/* now multiply the parent matrix by the owner matrix to get the
* the effect of this constraint (i.e. owner is 'parented' to parent)
*/
mult_m4_m4m4(cob->matrix, parmat, cob->matrix);
}
else {
float invmat[4][4], tempmat[4][4];
float loc[3], eul[3], size[3];
float loco[3], eulo[3], sizo[3];
/* get offset (parent-inverse) matrix */
copy_m4_m4(invmat, data->invmat);
/* extract components of both matrices */
copy_v3_v3(loc, ct->matrix[3]);
mat4_to_eulO(eul, ct->rotOrder, ct->matrix);
mat4_to_size(size, ct->matrix);
copy_v3_v3(loco, invmat[3]);
mat4_to_eulO(eulo, cob->rotOrder, invmat);
mat4_to_size(sizo, invmat);
/* disable channels not enabled */
if (!(data->flag & CHILDOF_LOCX)) loc[0]= loco[0]= 0.0f;
if (!(data->flag & CHILDOF_LOCY)) loc[1]= loco[1]= 0.0f;
if (!(data->flag & CHILDOF_LOCZ)) loc[2]= loco[2]= 0.0f;
if (!(data->flag & CHILDOF_ROTX)) eul[0]= eulo[0]= 0.0f;
if (!(data->flag & CHILDOF_ROTY)) eul[1]= eulo[1]= 0.0f;
if (!(data->flag & CHILDOF_ROTZ)) eul[2]= eulo[2]= 0.0f;
if (!(data->flag & CHILDOF_SIZEX)) size[0]= sizo[0]= 1.0f;
if (!(data->flag & CHILDOF_SIZEY)) size[1]= sizo[1]= 1.0f;
if (!(data->flag & CHILDOF_SIZEZ)) size[2]= sizo[2]= 1.0f;
/* make new target mat and offset mat */
loc_eulO_size_to_mat4(ct->matrix, loc, eul, size, ct->rotOrder);
loc_eulO_size_to_mat4(invmat, loco, eulo, sizo, cob->rotOrder);
/* multiply target (parent matrix) by offset (parent inverse) to get
* the effect of the parent that will be exherted on the owner
*/
mult_m4_m4m4(parmat, ct->matrix, invmat);
/* now multiply the parent matrix by the owner matrix to get the
* the effect of this constraint (i.e. owner is 'parented' to parent)
*/
copy_m4_m4(tempmat, cob->matrix);
mult_m4_m4m4(cob->matrix, parmat, tempmat);
/* without this, changes to scale and rotation can change location
* of a parentless bone or a disconnected bone. Even though its set
* to zero above. */
if (!(data->flag & CHILDOF_LOCX)) cob->matrix[3][0]= tempmat[3][0];
if (!(data->flag & CHILDOF_LOCY)) cob->matrix[3][1]= tempmat[3][1];
if (!(data->flag & CHILDOF_LOCZ)) cob->matrix[3][2]= tempmat[3][2];
}
}
}
/* XXX note, con->flag should be CONSTRAINT_SPACEONCE for bone-childof, patched in readfile.c */
static bConstraintTypeInfo CTI_CHILDOF = {
CONSTRAINT_TYPE_CHILDOF, /* type */
sizeof(bChildOfConstraint), /* size */
"ChildOf", /* name */
"bChildOfConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
childof_id_looper, /* id looper */
NULL, /* copy data */
childof_new_data, /* new data */
childof_get_tars, /* get constraint targets */
childof_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get a target matrix */
childof_evaluate /* evaluate */
};
/* -------- TrackTo Constraint ------- */
static void trackto_new_data (void *cdata)
{
bTrackToConstraint *data= (bTrackToConstraint *)cdata;
data->reserved1 = TRACK_Y;
data->reserved2 = UP_Z;
}
static void trackto_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bTrackToConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int trackto_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bTrackToConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void trackto_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bTrackToConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static int basis_cross (int n, int m)
{
switch (n-m) {
case 1:
case -2:
return 1;
case -1:
case 2:
return -1;
default:
return 0;
}
}
static void vectomat (float *vec, float *target_up, short axis, short upflag, short flags, float m[][3])
{
float n[3];
float u[3]; /* vector specifying the up axis */
float proj[3];
float right[3];
float neg = -1;
int right_index;
2010-08-15 15:14:08 +00:00
if (normalize_v3_v3(n, vec) == 0.0f) {
n[0] = 0.0f;
n[1] = 0.0f;
n[2] = 1.0f;
}
if (axis > 2) axis -= 3;
else negate_v3(n);
/* n specifies the transformation of the track axis */
if (flags & TARGET_Z_UP) {
/* target Z axis is the global up axis */
2010-08-15 15:14:08 +00:00
copy_v3_v3(u, target_up);
}
else {
/* world Z axis is the global up axis */
u[0] = 0;
u[1] = 0;
u[2] = 1;
}
/* project the up vector onto the plane specified by n */
project_v3_v3v3(proj, u, n); /* first u onto n... */
sub_v3_v3v3(proj, u, proj); /* then onto the plane */
/* proj specifies the transformation of the up axis */
if (normalize_v3(proj) == 0.0f) { /* degenerate projection */
proj[0] = 0.0f;
proj[1] = 1.0f;
proj[2] = 0.0f;
}
/* Normalized cross product of n and proj specifies transformation of the right axis */
cross_v3_v3v3(right, proj, n);
normalize_v3(right);
if (axis != upflag) {
right_index = 3 - axis - upflag;
neg = (float)basis_cross(axis, upflag);
/* account for up direction, track direction */
m[right_index][0] = neg * right[0];
m[right_index][1] = neg * right[1];
m[right_index][2] = neg * right[2];
2010-08-15 15:14:08 +00:00
copy_v3_v3(m[upflag], proj);
2010-08-15 15:14:08 +00:00
copy_v3_v3(m[axis], n);
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* identity matrix - don't do anything if the two axes are the same */
else {
2010-08-15 15:14:08 +00:00
unit_m3(m);
}
}
static void trackto_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
2002-10-12 11:37:38 +00:00
{
bTrackToConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
if (VALID_CONS_TARGET(ct)) {
float size[3], vec[3];
float totmat[3][3];
float tmat[4][4];
/* Get size property, since ob->size is only the object's own relative size, not its global one */
mat4_to_size(size, cob->matrix);
/* Clear the object's rotation */
cob->matrix[0][0]=size[0];
cob->matrix[0][1]=0;
cob->matrix[0][2]=0;
cob->matrix[1][0]=0;
cob->matrix[1][1]=size[1];
cob->matrix[1][2]=0;
cob->matrix[2][0]=0;
cob->matrix[2][1]=0;
cob->matrix[2][2]=size[2];
/* targetmat[2] instead of ownermat[2] is passed to vectomat
* for backwards compatibility it seems... (Aligorith)
*/
sub_v3_v3v3(vec, cob->matrix[3], ct->matrix[3]);
vectomat(vec, ct->matrix[2],
(short)data->reserved1, (short)data->reserved2,
data->flags, totmat);
copy_m4_m4(tmat, cob->matrix);
mul_m4_m3m4(cob->matrix, totmat, tmat);
}
}
2002-10-12 11:37:38 +00:00
static bConstraintTypeInfo CTI_TRACKTO = {
CONSTRAINT_TYPE_TRACKTO, /* type */
sizeof(bTrackToConstraint), /* size */
"TrackTo", /* name */
"bTrackToConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
trackto_id_looper, /* id looper */
NULL, /* copy data */
trackto_new_data, /* new data */
trackto_get_tars, /* get constraint targets */
trackto_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
trackto_evaluate /* evaluate */
};
/* --------- Inverse-Kinemetics --------- */
static void kinematic_new_data (void *cdata)
{
bKinematicConstraint *data= (bKinematicConstraint *)cdata;
data->weight= 1.0f;
data->orientweight= 1.0f;
data->iterations = 500;
data->dist= 1.0f;
data->flag= CONSTRAINT_IK_TIP|CONSTRAINT_IK_STRETCH|CONSTRAINT_IK_POS;
}
static void kinematic_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bKinematicConstraint *data= con->data;
/* chain target */
func(con, (ID**)&data->tar, userdata);
/* poletarget */
func(con, (ID**)&data->poletar, userdata);
}
static int kinematic_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bKinematicConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints is used twice here */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
SINGLETARGET_GET_TARS(con, data->poletar, data->polesubtarget, ct, list)
return 2;
}
return 0;
}
static void kinematic_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bKinematicConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
SINGLETARGET_FLUSH_TARS(con, data->poletar, data->polesubtarget, ct, list, nocopy)
}
}
static void kinematic_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float UNUSED(ctime))
{
bKinematicConstraint *data= con->data;
if (VALID_CONS_TARGET(ct))
2011-06-15 14:06:25 +00:00
constraint_target_to_mat4(ct->tar, ct->subtarget, ct->matrix, CONSTRAINT_SPACE_WORLD, ct->space, con->headtail);
else if (ct) {
if (data->flag & CONSTRAINT_IK_AUTO) {
Object *ob= cob->ob;
if (ob == NULL) {
unit_m4(ct->matrix);
}
else {
float vec[3];
/* move grabtarget into world space */
mul_v3_m4v3(vec, ob->obmat, data->grabtarget);
copy_m4_m4(ct->matrix, ob->obmat);
copy_v3_v3(ct->matrix[3], vec);
2002-10-12 11:37:38 +00:00
}
}
else
unit_m4(ct->matrix);
}
}
static bConstraintTypeInfo CTI_KINEMATIC = {
CONSTRAINT_TYPE_KINEMATIC, /* type */
sizeof(bKinematicConstraint), /* size */
"IK", /* name */
"bKinematicConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
kinematic_id_looper, /* id looper */
NULL, /* copy data */
kinematic_new_data, /* new data */
kinematic_get_tars, /* get constraint targets */
kinematic_flush_tars, /* flush constraint targets */
kinematic_get_tarmat, /* get target matrix */
NULL /* evaluate - solved as separate loop */
};
/* -------- Follow-Path Constraint ---------- */
static void followpath_new_data (void *cdata)
{
bFollowPathConstraint *data= (bFollowPathConstraint *)cdata;
data->trackflag = TRACK_Y;
data->upflag = UP_Z;
data->offset = 0;
data->followflag = 0;
}
static void followpath_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bFollowPathConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int followpath_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bFollowPathConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints without subtargets */
SINGLETARGETNS_GET_TARS(con, data->tar, ct, list)
return 1;
}
return 0;
}
static void followpath_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bFollowPathConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGETNS_FLUSH_TARS(con, data->tar, ct, list, nocopy)
}
}
static void followpath_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float UNUSED(ctime))
{
bFollowPathConstraint *data= con->data;
if (VALID_CONS_TARGET(ct)) {
Curve *cu= ct->tar->data;
float vec[4], dir[3], radius;
float totmat[4][4]= MAT4_UNITY;
float curvetime;
unit_m4(ct->matrix);
/* note: when creating constraints that follow path, the curve gets the CU_PATH set now,
* currently for paths to work it needs to go through the bevlist/displist system (ton)
*/
/* only happens on reload file, but violates depsgraph still... fix! */
2009-09-16 17:43:09 +00:00
if (cu->path==NULL || cu->path->data==NULL)
makeDispListCurveTypes(cob->scene, ct->tar, 0);
if (cu->path && cu->path->data) {
float quat[4];
if ((data->followflag & FOLLOWPATH_STATIC) == 0) {
2009-09-16 17:43:09 +00:00
/* animated position along curve depending on time */
Nurb *nu = cu->nurb.first;
curvetime= cu->ctime - data->offset;
2009-09-16 17:43:09 +00:00
/* ctime is now a proper var setting of Curve which gets set by Animato like any other var that's animated,
* but this will only work if it actually is animated...
*
* we divide the curvetime calculated in the previous step by the length of the path, to get a time
* factor, which then gets clamped to lie within 0.0 - 1.0 range
2009-09-16 17:43:09 +00:00
*/
curvetime /= cu->pathlen;
if (nu && nu->flagu & CU_NURB_CYCLIC) {
/* If the curve is cyclic, enable looping around if the time is
* outside the bounds 0..1 */
if ((curvetime < 0.0f) || (curvetime > 1.0f)) {
2012-02-04 06:55:29 +00:00
curvetime -= floorf(curvetime);
}
}
else {
/* The curve is not cyclic, so clamp to the begin/end points. */
CLAMP(curvetime, 0.0f, 1.0f);
}
}
2009-09-16 17:43:09 +00:00
else {
/* fixed position along curve */
curvetime= data->offset_fac;
2009-09-16 17:43:09 +00:00
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
if ( where_on_path(ct->tar, curvetime, vec, dir, (data->followflag & FOLLOWPATH_FOLLOW) ? quat : NULL, &radius, NULL) ) { /* quat_pt is quat or NULL*/
2009-09-16 17:43:09 +00:00
if (data->followflag & FOLLOWPATH_FOLLOW) {
#if 0
float x1, q[4];
vec_to_quat(quat, dir, (short)data->trackflag, (short)data->upflag);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
normalize_v3(dir);
q[0]= (float)cos(0.5*vec[3]);
x1= (float)sin(0.5*vec[3]);
q[1]= -x1*dir[0];
q[2]= -x1*dir[1];
q[3]= -x1*dir[2];
mul_qt_qtqt(quat, q, quat);
#else
quat_apply_track(quat, data->trackflag, data->upflag);
#endif
quat_to_mat4(totmat, quat);
}
2009-09-16 17:43:09 +00:00
if (data->followflag & FOLLOWPATH_RADIUS) {
float tmat[4][4], rmat[4][4];
scale_m4_fl(tmat, radius);
mult_m4_m4m4(rmat, tmat, totmat);
copy_m4_m4(totmat, rmat);
2009-09-16 17:43:09 +00:00
}
2010-08-15 15:14:08 +00:00
copy_v3_v3(totmat[3], vec);
mul_serie_m4(ct->matrix, ct->tar->obmat, totmat, NULL, NULL, NULL, NULL, NULL, NULL);
}
}
}
else if (ct)
unit_m4(ct->matrix);
}
static void followpath_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bConstraintTarget *ct= targets->first;
/* only evaluate if there is a target */
if (VALID_CONS_TARGET(ct)) {
float obmat[4][4];
2009-09-16 17:43:09 +00:00
float size[3];
bFollowPathConstraint *data= con->data;
/* get Object transform (loc/rot/size) to determine transformation from path */
// TODO: this used to be local at one point, but is probably more useful as-is
copy_m4_m4(obmat, cob->matrix);
/* get scaling of object before applying constraint */
mat4_to_size(size, cob->matrix);
/* apply targetmat - containing location on path, and rotation */
mul_serie_m4(cob->matrix, ct->matrix, obmat, NULL, NULL, NULL, NULL, NULL, NULL);
/* un-apply scaling caused by path */
2009-09-16 17:43:09 +00:00
if ((data->followflag & FOLLOWPATH_RADIUS)==0) { /* XXX - assume that scale correction means that radius will have some scale error in it - Campbell */
float obsize[3];
2012-04-29 15:47:02 +00:00
mat4_to_size(obsize, cob->matrix);
2009-09-16 17:43:09 +00:00
if (obsize[0])
mul_v3_fl(cob->matrix[0], size[0] / obsize[0]);
2009-09-16 17:43:09 +00:00
if (obsize[1])
mul_v3_fl(cob->matrix[1], size[1] / obsize[1]);
2009-09-16 17:43:09 +00:00
if (obsize[2])
mul_v3_fl(cob->matrix[2], size[2] / obsize[2]);
2009-09-16 17:43:09 +00:00
}
}
}
static bConstraintTypeInfo CTI_FOLLOWPATH = {
CONSTRAINT_TYPE_FOLLOWPATH, /* type */
sizeof(bFollowPathConstraint), /* size */
"Follow Path", /* name */
"bFollowPathConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
followpath_id_looper, /* id looper */
NULL, /* copy data */
followpath_new_data, /* new data */
followpath_get_tars, /* get constraint targets */
followpath_flush_tars, /* flush constraint targets */
followpath_get_tarmat, /* get target matrix */
followpath_evaluate /* evaluate */
};
/* --------- Limit Location --------- */
static void loclimit_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *UNUSED(targets))
{
bLocLimitConstraint *data = con->data;
if (data->flag & LIMIT_XMIN) {
if (cob->matrix[3][0] < data->xmin)
cob->matrix[3][0] = data->xmin;
}
if (data->flag & LIMIT_XMAX) {
if (cob->matrix[3][0] > data->xmax)
cob->matrix[3][0] = data->xmax;
}
if (data->flag & LIMIT_YMIN) {
if (cob->matrix[3][1] < data->ymin)
cob->matrix[3][1] = data->ymin;
}
if (data->flag & LIMIT_YMAX) {
if (cob->matrix[3][1] > data->ymax)
cob->matrix[3][1] = data->ymax;
}
if (data->flag & LIMIT_ZMIN) {
if (cob->matrix[3][2] < data->zmin)
cob->matrix[3][2] = data->zmin;
}
if (data->flag & LIMIT_ZMAX) {
if (cob->matrix[3][2] > data->zmax)
cob->matrix[3][2] = data->zmax;
}
}
static bConstraintTypeInfo CTI_LOCLIMIT = {
CONSTRAINT_TYPE_LOCLIMIT, /* type */
sizeof(bLocLimitConstraint), /* size */
"Limit Location", /* name */
"bLocLimitConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
NULL, /* id looper */
NULL, /* copy data */
NULL, /* new data */
NULL, /* get constraint targets */
NULL, /* flush constraint targets */
NULL, /* get target matrix */
loclimit_evaluate /* evaluate */
};
/* -------- Limit Rotation --------- */
static void rotlimit_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *UNUSED(targets))
{
bRotLimitConstraint *data = con->data;
float loc[3];
float eul[3];
float size[3];
2010-08-15 15:14:08 +00:00
copy_v3_v3(loc, cob->matrix[3]);
mat4_to_size(size, cob->matrix);
mat4_to_eulO(eul, cob->rotOrder, cob->matrix);
/* constraint data uses radians internally */
/* limiting of euler values... */
if (data->flag & LIMIT_XROT) {
if (eul[0] < data->xmin)
eul[0] = data->xmin;
if (eul[0] > data->xmax)
eul[0] = data->xmax;
}
if (data->flag & LIMIT_YROT) {
if (eul[1] < data->ymin)
eul[1] = data->ymin;
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
if (eul[1] > data->ymax)
eul[1] = data->ymax;
}
if (data->flag & LIMIT_ZROT) {
if (eul[2] < data->zmin)
eul[2] = data->zmin;
if (eul[2] > data->zmax)
eul[2] = data->zmax;
}
loc_eulO_size_to_mat4(cob->matrix, loc, eul, size, cob->rotOrder);
}
static bConstraintTypeInfo CTI_ROTLIMIT = {
CONSTRAINT_TYPE_ROTLIMIT, /* type */
sizeof(bRotLimitConstraint), /* size */
"Limit Rotation", /* name */
"bRotLimitConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
NULL, /* id looper */
NULL, /* copy data */
NULL, /* new data */
NULL, /* get constraint targets */
NULL, /* flush constraint targets */
NULL, /* get target matrix */
rotlimit_evaluate /* evaluate */
};
/* --------- Limit Scaling --------- */
static void sizelimit_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *UNUSED(targets))
{
bSizeLimitConstraint *data = con->data;
float obsize[3], size[3];
2012-04-29 15:47:02 +00:00
mat4_to_size(size, cob->matrix);
mat4_to_size(obsize, cob->matrix);
if (data->flag & LIMIT_XMIN) {
if (size[0] < data->xmin)
size[0] = data->xmin;
}
if (data->flag & LIMIT_XMAX) {
if (size[0] > data->xmax)
size[0] = data->xmax;
2002-10-12 11:37:38 +00:00
}
if (data->flag & LIMIT_YMIN) {
if (size[1] < data->ymin)
size[1] = data->ymin;
}
if (data->flag & LIMIT_YMAX) {
if (size[1] > data->ymax)
size[1] = data->ymax;
}
if (data->flag & LIMIT_ZMIN) {
if (size[2] < data->zmin)
size[2] = data->zmin;
}
if (data->flag & LIMIT_ZMAX) {
if (size[2] > data->zmax)
size[2] = data->zmax;
}
if (obsize[0])
mul_v3_fl(cob->matrix[0], size[0]/obsize[0]);
if (obsize[1])
mul_v3_fl(cob->matrix[1], size[1]/obsize[1]);
if (obsize[2])
mul_v3_fl(cob->matrix[2], size[2]/obsize[2]);
}
2002-10-12 11:37:38 +00:00
static bConstraintTypeInfo CTI_SIZELIMIT = {
CONSTRAINT_TYPE_SIZELIMIT, /* type */
sizeof(bSizeLimitConstraint), /* size */
"Limit Scaling", /* name */
"bSizeLimitConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
NULL, /* id looper */
NULL, /* copy data */
NULL, /* new data */
NULL, /* get constraint targets */
NULL, /* flush constraint targets */
NULL, /* get target matrix */
sizelimit_evaluate /* evaluate */
};
/* ----------- Copy Location ------------- */
static void loclike_new_data (void *cdata)
{
bLocateLikeConstraint *data= (bLocateLikeConstraint *)cdata;
data->flag = LOCLIKE_X|LOCLIKE_Y|LOCLIKE_Z;
2002-10-12 11:37:38 +00:00
}
static void loclike_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bLocateLikeConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int loclike_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bLocateLikeConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
static void loclike_flush_tars (bConstraint *con, ListBase *list, short nocopy)
2002-10-12 11:37:38 +00:00
{
if (con && list) {
bLocateLikeConstraint *data= con->data;
bConstraintTarget *ct= list->first;
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
2002-10-12 11:37:38 +00:00
static void loclike_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bLocateLikeConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
if (VALID_CONS_TARGET(ct)) {
float offset[3] = {0.0f, 0.0f, 0.0f};
if (data->flag & LOCLIKE_OFFSET)
2010-08-15 15:14:08 +00:00
copy_v3_v3(offset, cob->matrix[3]);
if (data->flag & LOCLIKE_X) {
cob->matrix[3][0] = ct->matrix[3][0];
if (data->flag & LOCLIKE_X_INVERT) cob->matrix[3][0] *= -1;
cob->matrix[3][0] += offset[0];
}
if (data->flag & LOCLIKE_Y) {
cob->matrix[3][1] = ct->matrix[3][1];
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
if (data->flag & LOCLIKE_Y_INVERT) cob->matrix[3][1] *= -1;
cob->matrix[3][1] += offset[1];
}
if (data->flag & LOCLIKE_Z) {
cob->matrix[3][2] = ct->matrix[3][2];
if (data->flag & LOCLIKE_Z_INVERT) cob->matrix[3][2] *= -1;
cob->matrix[3][2] += offset[2];
}
}
}
2002-10-12 11:37:38 +00:00
static bConstraintTypeInfo CTI_LOCLIKE = {
CONSTRAINT_TYPE_LOCLIKE, /* type */
sizeof(bLocateLikeConstraint), /* size */
"Copy Location", /* name */
"bLocateLikeConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
loclike_id_looper, /* id looper */
NULL, /* copy data */
loclike_new_data, /* new data */
loclike_get_tars, /* get constraint targets */
loclike_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
loclike_evaluate /* evaluate */
};
/* ----------- Copy Rotation ------------- */
static void rotlike_new_data (void *cdata)
{
bRotateLikeConstraint *data= (bRotateLikeConstraint *)cdata;
data->flag = ROTLIKE_X|ROTLIKE_Y|ROTLIKE_Z;
}
static void rotlike_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bChildOfConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int rotlike_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bRotateLikeConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void rotlike_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bRotateLikeConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void rotlike_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bRotateLikeConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
if (VALID_CONS_TARGET(ct)) {
float loc[3];
float eul[3], obeul[3];
float size[3];
2010-08-15 15:14:08 +00:00
copy_v3_v3(loc, cob->matrix[3]);
mat4_to_size(size, cob->matrix);
/* to allow compatible rotations, must get both rotations in the order of the owner... */
mat4_to_eulO(obeul, cob->rotOrder, cob->matrix);
/* we must get compatible eulers from the beginning because some of them can be modified below (see bug #21875) */
mat4_to_compatible_eulO(eul, obeul, cob->rotOrder, ct->matrix);
if ((data->flag & ROTLIKE_X)==0)
eul[0] = obeul[0];
else {
if (data->flag & ROTLIKE_OFFSET)
rotate_eulO(eul, cob->rotOrder, 'X', obeul[0]);
if (data->flag & ROTLIKE_X_INVERT)
eul[0] *= -1;
}
if ((data->flag & ROTLIKE_Y)==0)
eul[1] = obeul[1];
else {
if (data->flag & ROTLIKE_OFFSET)
rotate_eulO(eul, cob->rotOrder, 'Y', obeul[1]);
if (data->flag & ROTLIKE_Y_INVERT)
eul[1] *= -1;
}
if ((data->flag & ROTLIKE_Z)==0)
eul[2] = obeul[2];
else {
if (data->flag & ROTLIKE_OFFSET)
rotate_eulO(eul, cob->rotOrder, 'Z', obeul[2]);
if (data->flag & ROTLIKE_Z_INVERT)
eul[2] *= -1;
}
/* good to make eulers compatible again, since we don't know how much they were changed above */
compatible_eul(eul, obeul);
loc_eulO_size_to_mat4(cob->matrix, loc, eul, size, cob->rotOrder);
}
}
static bConstraintTypeInfo CTI_ROTLIKE = {
CONSTRAINT_TYPE_ROTLIKE, /* type */
sizeof(bRotateLikeConstraint), /* size */
"Copy Rotation", /* name */
"bRotateLikeConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
rotlike_id_looper, /* id looper */
NULL, /* copy data */
rotlike_new_data, /* new data */
rotlike_get_tars, /* get constraint targets */
rotlike_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
rotlike_evaluate /* evaluate */
};
/* ---------- Copy Scaling ---------- */
static void sizelike_new_data (void *cdata)
{
bSizeLikeConstraint *data= (bSizeLikeConstraint *)cdata;
data->flag = SIZELIKE_X|SIZELIKE_Y|SIZELIKE_Z;
}
static void sizelike_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bSizeLikeConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int sizelike_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bSizeLikeConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void sizelike_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bSizeLikeConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void sizelike_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bSizeLikeConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
if (VALID_CONS_TARGET(ct)) {
float obsize[3], size[3];
mat4_to_size(size, ct->matrix);
mat4_to_size(obsize, cob->matrix);
if ((data->flag & SIZELIKE_X) && (obsize[0] != 0)) {
if (data->flag & SIZELIKE_OFFSET) {
size[0] += (obsize[0] - 1.0f);
mul_v3_fl(cob->matrix[0], size[0] / obsize[0]);
}
else
mul_v3_fl(cob->matrix[0], size[0] / obsize[0]);
}
if ((data->flag & SIZELIKE_Y) && (obsize[1] != 0)) {
if (data->flag & SIZELIKE_OFFSET) {
size[1] += (obsize[1] - 1.0f);
mul_v3_fl(cob->matrix[1], size[1] / obsize[1]);
}
else
mul_v3_fl(cob->matrix[1], size[1] / obsize[1]);
}
if ((data->flag & SIZELIKE_Z) && (obsize[2] != 0)) {
if (data->flag & SIZELIKE_OFFSET) {
size[2] += (obsize[2] - 1.0f);
mul_v3_fl(cob->matrix[2], size[2] / obsize[2]);
}
else
mul_v3_fl(cob->matrix[2], size[2] / obsize[2]);
}
}
}
static bConstraintTypeInfo CTI_SIZELIKE = {
CONSTRAINT_TYPE_SIZELIKE, /* type */
sizeof(bSizeLikeConstraint), /* size */
"Copy Scale", /* name */
"bSizeLikeConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
sizelike_id_looper, /* id looper */
NULL, /* copy data */
sizelike_new_data, /* new data */
sizelike_get_tars, /* get constraint targets */
sizelike_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
sizelike_evaluate /* evaluate */
};
/* ----------- Copy Transforms ------------- */
static void translike_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bTransLikeConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int translike_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bTransLikeConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void translike_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bTransLikeConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void translike_evaluate (bConstraint *UNUSED(con), bConstraintOb *cob, ListBase *targets)
{
bConstraintTarget *ct= targets->first;
if (VALID_CONS_TARGET(ct)) {
/* just copy the entire transform matrix of the target */
copy_m4_m4(cob->matrix, ct->matrix);
}
}
static bConstraintTypeInfo CTI_TRANSLIKE = {
CONSTRAINT_TYPE_TRANSLIKE, /* type */
sizeof(bTransLikeConstraint), /* size */
"Copy Transforms", /* name */
"bTransLikeConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
translike_id_looper, /* id looper */
NULL, /* copy data */
NULL, /* new data */
translike_get_tars, /* get constraint targets */
translike_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
translike_evaluate /* evaluate */
};
/* ---------- Maintain Volume ---------- */
static void samevolume_new_data (void *cdata)
{
bSameVolumeConstraint *data= (bSameVolumeConstraint *)cdata;
data->flag = SAMEVOL_Y;
data->volume = 1.0f;
}
static void samevolume_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *UNUSED(targets))
{
bSameVolumeConstraint *data= con->data;
float volume = data->volume;
float fac = 1.0f;
float obsize[3];
mat4_to_size(obsize, cob->matrix);
/* calculate normalizing scale factor for non-essential values */
if (obsize[data->flag] != 0)
fac = sqrtf(volume / obsize[data->flag]) / obsize[data->flag];
/* apply scaling factor to the channels not being kept */
switch (data->flag) {
case SAMEVOL_X:
mul_v3_fl(cob->matrix[1], fac);
mul_v3_fl(cob->matrix[2], fac);
break;
case SAMEVOL_Y:
mul_v3_fl(cob->matrix[0], fac);
mul_v3_fl(cob->matrix[2], fac);
break;
case SAMEVOL_Z:
mul_v3_fl(cob->matrix[0], fac);
mul_v3_fl(cob->matrix[1], fac);
break;
}
}
static bConstraintTypeInfo CTI_SAMEVOL = {
CONSTRAINT_TYPE_SAMEVOL, /* type */
sizeof(bSameVolumeConstraint), /* size */
"Maintain Volume", /* name */
"bSameVolumeConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
NULL, /* id looper */
NULL, /* copy data */
samevolume_new_data, /* new data */
NULL, /* get constraint targets */
NULL, /* flush constraint targets */
NULL, /* get target matrix */
samevolume_evaluate /* evaluate */
};
/* ----------- Python Constraint -------------- */
static void pycon_free (bConstraint *con)
{
bPythonConstraint *data= con->data;
/* id-properties */
IDP_FreeProperty(data->prop);
MEM_freeN(data->prop);
/* multiple targets */
BLI_freelistN(&data->targets);
}
static void pycon_relink (bConstraint *con)
{
bPythonConstraint *data= con->data;
ID_NEW(data->text);
}
static void pycon_copy (bConstraint *con, bConstraint *srccon)
{
bPythonConstraint *pycon = (bPythonConstraint *)con->data;
bPythonConstraint *opycon = (bPythonConstraint *)srccon->data;
pycon->prop = IDP_CopyProperty(opycon->prop);
BLI_duplicatelist(&pycon->targets, &opycon->targets);
}
static void pycon_new_data (void *cdata)
{
bPythonConstraint *data= (bPythonConstraint *)cdata;
/* everything should be set correctly by calloc, except for the prop->type constant.*/
data->prop = MEM_callocN(sizeof(IDProperty), "PyConstraintProps");
data->prop->type = IDP_GROUP;
}
static int pycon_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bPythonConstraint *data= con->data;
list->first = data->targets.first;
list->last = data->targets.last;
return data->tarnum;
}
return 0;
}
static void pycon_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bPythonConstraint *data= con->data;
bConstraintTarget *ct;
/* targets */
for (ct= data->targets.first; ct; ct= ct->next)
func(con, (ID**)&ct->tar, userdata);
/* script */
func(con, (ID**)&data->text, userdata);
}
/* Whether this approach is maintained remains to be seen (aligorith) */
static void pycon_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float UNUSED(ctime))
{
#ifdef WITH_PYTHON
bPythonConstraint *data= con->data;
#endif
if (VALID_CONS_TARGET(ct)) {
/* special exception for curves - depsgraph issues */
if (ct->tar->type == OB_CURVE) {
Curve *cu= ct->tar->data;
2002-10-12 11:37:38 +00:00
/* this check is to make sure curve objects get updated on file load correctly.*/
if (cu->path==NULL || cu->path->data==NULL) /* only happens on reload file, but violates depsgraph still... fix! */
makeDispListCurveTypes(cob->scene, ct->tar, 0);
}
/* firstly calculate the matrix the normal way, then let the py-function override
* this matrix if it needs to do so
*/
2011-06-15 14:06:25 +00:00
constraint_target_to_mat4(ct->tar, ct->subtarget, ct->matrix, CONSTRAINT_SPACE_WORLD, ct->space, con->headtail);
/* only execute target calculation if allowed */
#ifdef WITH_PYTHON
if (G.f & G_SCRIPT_AUTOEXEC)
BPY_pyconstraint_target(data, ct);
#endif
}
else if (ct)
unit_m4(ct->matrix);
}
static void pycon_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
#ifndef WITH_PYTHON
2010-10-23 16:03:31 +00:00
(void)con; (void)cob; (void)targets; /* unused */
return;
#else
bPythonConstraint *data= con->data;
/* only evaluate in python if we're allowed to do so */
if ((G.f & G_SCRIPT_AUTOEXEC)==0) return;
/* currently removed, until I this can be re-implemented for multiple targets */
#if 0
/* Firstly, run the 'driver' function which has direct access to the objects involved
* Technically, this is potentially dangerous as users may abuse this and cause dependency-problems,
* but it also allows certain 'clever' rigging hacks to work.
*/
BPY_pyconstraint_driver(data, cob, targets);
#endif
/* Now, run the actual 'constraint' function, which should only access the matrices */
BPY_pyconstraint_exec(data, cob, targets);
#endif /* WITH_PYTHON */
}
static bConstraintTypeInfo CTI_PYTHON = {
CONSTRAINT_TYPE_PYTHON, /* type */
sizeof(bPythonConstraint), /* size */
"Script", /* name */
"bPythonConstraint", /* struct name */
pycon_free, /* free data */
pycon_relink, /* relink data */
pycon_id_looper, /* id looper */
pycon_copy, /* copy data */
pycon_new_data, /* new data */
pycon_get_tars, /* get constraint targets */
NULL, /* flush constraint targets */
pycon_get_tarmat, /* get target matrix */
pycon_evaluate /* evaluate */
};
/* -------- Action Constraint ----------- */
static void actcon_relink (bConstraint *con)
{
bActionConstraint *data= con->data;
ID_NEW(data->act);
}
static void actcon_new_data (void *cdata)
{
bActionConstraint *data= (bActionConstraint *)cdata;
/* set type to 20 (Loc X), as 0 is Rot X for backwards compatibility */
data->type = 20;
}
static void actcon_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bActionConstraint *data= con->data;
/* target */
func(con, (ID**)&data->tar, userdata);
/* action */
func(con, (ID**)&data->act, userdata);
}
static int actcon_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bActionConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void actcon_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bActionConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void actcon_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float UNUSED(ctime))
{
bActionConstraint *data = con->data;
if (VALID_CONS_TARGET(ct)) {
float tempmat[4][4], vec[3];
float s, t;
short axis;
/* initialize return matrix */
unit_m4(ct->matrix);
/* get the transform matrix of the target */
2011-06-15 14:06:25 +00:00
constraint_target_to_mat4(ct->tar, ct->subtarget, tempmat, CONSTRAINT_SPACE_WORLD, ct->space, con->headtail);
/* determine where in transform range target is */
/* data->type is mapped as follows for backwards compatibility:
* 00,01,02 - rotation (it used to be like this)
* 10,11,12 - scaling
* 20,21,22 - location
*/
if (data->type < 10) {
/* extract rotation (is in whatever space target should be in) */
mat4_to_eul(vec, tempmat);
mul_v3_fl(vec, RAD2DEGF(1.0f)); /* rad -> deg */
axis= data->type;
}
else if (data->type < 20) {
/* extract scaling (is in whatever space target should be in) */
mat4_to_size(vec, tempmat);
axis= data->type - 10;
}
else {
/* extract location */
2010-08-15 15:14:08 +00:00
copy_v3_v3(vec, tempmat[3]);
axis= data->type - 20;
}
/* Target defines the animation */
s = (vec[axis]-data->min) / (data->max-data->min);
CLAMP(s, 0, 1);
t = (s * (data->end-data->start)) + data->start;
if (G.debug & G_DEBUG)
printf("do Action Constraint %s - Ob %s Pchan %s\n", con->name, cob->ob->id.name+2, (cob->pchan)?cob->pchan->name:NULL);
/* Get the appropriate information from the action */
if (cob->type == CONSTRAINT_OBTYPE_BONE) {
Object workob;
bPose *pose;
bPoseChannel *pchan, *tchan;
/* make a temporary pose and evaluate using that */
pose = MEM_callocN(sizeof(bPose), "pose");
/* make a copy of the bone of interest in the temp pose before evaluating action, so that it can get set
* - we need to manually copy over a few settings, including rotation order, otherwise this fails
*/
pchan = cob->pchan;
tchan= verify_pose_channel(pose, pchan->name);
tchan->rotmode= pchan->rotmode;
/* evaluate action using workob (it will only set the PoseChannel in question) */
2011-06-15 14:06:25 +00:00
what_does_obaction(cob->ob, &workob, pose, data->act, pchan->name, t);
/* convert animation to matrices for use here */
pchan_calc_mat(tchan);
copy_m4_m4(ct->matrix, tchan->chan_mat);
/* Clean up */
== Bone Groups == I'm committing some work-in-progress code for "bone groups" now, as I there have been are some major bugs caused by the timeoffset stuff (some of my test files were not loading, and other files were showing all sorts of weird problems). Anyway, in this commit, the following things for "bone groups" have been done: * Bone groups are stored per armature (internally, this is per bPose block) * Added controls for editing bone-groups per armature - "add", "remove", "rename". These can be found in the "Links and Materials" panel in PoseMode, beside the settings for PoseLib. * Reorganised buttons for editing selected bones in PoseMode. I've replaced the "dist" and "weight" buttons (they existed in EditMode anyway) with a menu to choose the bone-group and the custom-shape-ob field. In the place of the old custom-shape-ob field, I've restored the "Hide" button. This might break muscle-memory a bit, but there isn't a lot of space to play with there. Some stuff I'd been originally planning to do before committing: * When adding keyframes for bones, an action-group with the same name as the bone's group will be added to the action, and the action-channel will be made a member of that. * New action/bone groups have unique names (renaming/adding new should check if name exists before assigning it) * There's a setting under Bone-Groups stuff which sets which custom-colour set is used to colour that group's bones. Currently, this is non-functional, as the necessary drawing code for armatures is not in place yet.
2008-01-20 02:55:35 +00:00
free_pose(pose);
}
else if (cob->type == CONSTRAINT_OBTYPE_OBJECT) {
Object workob;
/* evaluate using workob */
// FIXME: we don't have any consistent standards on limiting effects on object...
2011-06-15 14:06:25 +00:00
what_does_obaction(cob->ob, &workob, NULL, data->act, NULL, t);
BKE_object_to_mat4(&workob, ct->matrix);
}
else {
2012-03-01 12:20:18 +00:00
/* behavior undefined... */
puts("Error: unknown owner type for Action Constraint");
}
}
}
static void actcon_evaluate (bConstraint *UNUSED(con), bConstraintOb *cob, ListBase *targets)
{
bConstraintTarget *ct= targets->first;
if (VALID_CONS_TARGET(ct)) {
float temp[4][4];
/* Nice and simple... we just need to multiply the matrices, as the get_target_matrix
* function has already taken care of everything else.
*/
copy_m4_m4(temp, cob->matrix);
mult_m4_m4m4(cob->matrix, temp, ct->matrix);
}
}
static bConstraintTypeInfo CTI_ACTION = {
CONSTRAINT_TYPE_ACTION, /* type */
sizeof(bActionConstraint), /* size */
"Action", /* name */
"bActionConstraint", /* struct name */
NULL, /* free data */
actcon_relink, /* relink data */
actcon_id_looper, /* id looper */
NULL, /* copy data */
actcon_new_data, /* new data */
actcon_get_tars, /* get constraint targets */
actcon_flush_tars, /* flush constraint targets */
actcon_get_tarmat, /* get target matrix */
actcon_evaluate /* evaluate */
};
/* --------- Locked Track ---------- */
static void locktrack_new_data (void *cdata)
{
bLockTrackConstraint *data= (bLockTrackConstraint *)cdata;
data->trackflag = TRACK_Y;
2007-10-26 11:37:29 +00:00
data->lockflag = LOCK_Z;
}
static void locktrack_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bLockTrackConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int locktrack_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bLockTrackConstraint *data= con->data;
bConstraintTarget *ct;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void locktrack_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bLockTrackConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void locktrack_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bLockTrackConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
if (VALID_CONS_TARGET(ct)) {
2012-04-29 15:47:02 +00:00
float vec[3], vec2[3];
float totmat[3][3];
float tmpmat[3][3];
float invmat[3][3];
float tmat[4][4];
float mdet;
/* Vector object -> target */
sub_v3_v3v3(vec, ct->matrix[3], cob->matrix[3]);
switch (data->lockflag) {
case LOCK_X: /* LOCK X */
{
switch (data->trackflag) {
case TRACK_Y: /* LOCK X TRACK Y */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[0]);
sub_v3_v3v3(totmat[1], vec, vec2);
normalize_v3(totmat[1]);
/* the x axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[0], cob->matrix[0]);
/* the z axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[2], totmat[0], totmat[1]);
}
break;
case TRACK_Z: /* LOCK X TRACK Z */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[0]);
sub_v3_v3v3(totmat[2], vec, vec2);
normalize_v3(totmat[2]);
2007-10-26 11:37:29 +00:00
/* the x axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[0], cob->matrix[0]);
2007-10-26 11:37:29 +00:00
/* the z axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[1], totmat[2], totmat[0]);
}
break;
case TRACK_nY: /* LOCK X TRACK -Y */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[0]);
sub_v3_v3v3(totmat[1], vec, vec2);
normalize_v3(totmat[1]);
negate_v3(totmat[1]);
/* the x axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[0], cob->matrix[0]);
/* the z axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[2], totmat[0], totmat[1]);
}
break;
case TRACK_nZ: /* LOCK X TRACK -Z */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[0]);
sub_v3_v3v3(totmat[2], vec, vec2);
normalize_v3(totmat[2]);
negate_v3(totmat[2]);
/* the x axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[0], cob->matrix[0]);
/* the z axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[1], totmat[2], totmat[0]);
}
break;
default:
{
2010-08-15 15:14:08 +00:00
unit_m3(totmat);
}
break;
2002-10-12 11:37:38 +00:00
}
}
break;
case LOCK_Y: /* LOCK Y */
{
switch (data->trackflag) {
case TRACK_X: /* LOCK Y TRACK X */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[1]);
sub_v3_v3v3(totmat[0], vec, vec2);
normalize_v3(totmat[0]);
/* the y axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[1], cob->matrix[1]);
/* the z axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[2], totmat[0], totmat[1]);
}
break;
case TRACK_Z: /* LOCK Y TRACK Z */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[1]);
sub_v3_v3v3(totmat[2], vec, vec2);
normalize_v3(totmat[2]);
/* the y axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[1], cob->matrix[1]);
/* the z axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[0], totmat[1], totmat[2]);
}
break;
case TRACK_nX: /* LOCK Y TRACK -X */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[1]);
sub_v3_v3v3(totmat[0], vec, vec2);
normalize_v3(totmat[0]);
negate_v3(totmat[0]);
/* the y axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[1], cob->matrix[1]);
/* the z axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[2], totmat[0], totmat[1]);
}
break;
case TRACK_nZ: /* LOCK Y TRACK -Z */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[1]);
sub_v3_v3v3(totmat[2], vec, vec2);
normalize_v3(totmat[2]);
negate_v3(totmat[2]);
/* the y axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[1], cob->matrix[1]);
/* the z axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[0], totmat[1], totmat[2]);
}
break;
default:
{
2010-08-15 15:14:08 +00:00
unit_m3(totmat);
}
break;
}
}
break;
case LOCK_Z: /* LOCK Z */
{
switch (data->trackflag) {
case TRACK_X: /* LOCK Z TRACK X */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[2]);
sub_v3_v3v3(totmat[0], vec, vec2);
normalize_v3(totmat[0]);
/* the z axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[2], cob->matrix[2]);
/* the x axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[1], totmat[2], totmat[0]);
}
break;
case TRACK_Y: /* LOCK Z TRACK Y */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[2]);
sub_v3_v3v3(totmat[1], vec, vec2);
normalize_v3(totmat[1]);
/* the z axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[2], cob->matrix[2]);
/* the x axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[0], totmat[1], totmat[2]);
}
break;
case TRACK_nX: /* LOCK Z TRACK -X */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[2]);
sub_v3_v3v3(totmat[0], vec, vec2);
normalize_v3(totmat[0]);
negate_v3(totmat[0]);
/* the z axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[2], cob->matrix[2]);
/* the x axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[1], totmat[2], totmat[0]);
}
break;
case TRACK_nY: /* LOCK Z TRACK -Y */
{
/* Projection of Vector on the plane */
project_v3_v3v3(vec2, vec, cob->matrix[2]);
sub_v3_v3v3(totmat[1], vec, vec2);
normalize_v3(totmat[1]);
negate_v3(totmat[1]);
/* the z axis is fixed */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[2], cob->matrix[2]);
/* the x axis gets mapped onto a third orthogonal vector */
cross_v3_v3v3(totmat[0], totmat[1], totmat[2]);
}
break;
default:
{
2010-08-15 15:14:08 +00:00
unit_m3(totmat);
}
break;
}
}
break;
default:
2011-05-11 09:31:00 +00:00
{
unit_m3(totmat);
}
break;
}
/* Block to keep matrix heading */
2010-08-15 15:14:08 +00:00
copy_m3_m4(tmpmat, cob->matrix);
normalize_m3(tmpmat);
invert_m3_m3(invmat, tmpmat);
mul_m3_m3m3(tmpmat, totmat, invmat);
totmat[0][0] = tmpmat[0][0];totmat[0][1] = tmpmat[0][1];totmat[0][2] = tmpmat[0][2];
totmat[1][0] = tmpmat[1][0];totmat[1][1] = tmpmat[1][1];totmat[1][2] = tmpmat[1][2];
totmat[2][0] = tmpmat[2][0];totmat[2][1] = tmpmat[2][1];totmat[2][2] = tmpmat[2][2];
copy_m4_m4(tmat, cob->matrix);
2012-04-29 15:47:02 +00:00
mdet = determinant_m3(totmat[0][0], totmat[0][1], totmat[0][2],
totmat[1][0], totmat[1][1], totmat[1][2],
totmat[2][0], totmat[2][1], totmat[2][2]);
if (mdet==0) {
2010-08-15 15:14:08 +00:00
unit_m3(totmat);
}
/* apply out transformaton to the object */
mul_m4_m3m4(cob->matrix, totmat, tmat);
}
}
static bConstraintTypeInfo CTI_LOCKTRACK = {
CONSTRAINT_TYPE_LOCKTRACK, /* type */
sizeof(bLockTrackConstraint), /* size */
"Locked Track", /* name */
"bLockTrackConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
locktrack_id_looper, /* id looper */
NULL, /* copy data */
locktrack_new_data, /* new data */
locktrack_get_tars, /* get constraint targets */
locktrack_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
locktrack_evaluate /* evaluate */
};
/* ---------- Limit Distance Constraint ----------- */
static void distlimit_new_data (void *cdata)
{
bDistLimitConstraint *data= (bDistLimitConstraint *)cdata;
data->dist= 0.0f;
}
static void distlimit_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bDistLimitConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int distlimit_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bDistLimitConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void distlimit_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bDistLimitConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void distlimit_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bDistLimitConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
/* only evaluate if there is a target */
if (VALID_CONS_TARGET(ct)) {
float dvec[3], dist=0.0f, sfac=1.0f;
short clamp_surf= 0;
/* calculate our current distance from the target */
dist= len_v3v3(cob->matrix[3], ct->matrix[3]);
/* set distance (flag is only set when user demands it) */
if (data->dist == 0)
data->dist= dist;
/* check if we're which way to clamp from, and calculate interpolation factor (if needed) */
if (data->mode == LIMITDIST_OUTSIDE) {
/* if inside, then move to surface */
if (dist <= data->dist) {
clamp_surf= 1;
if (dist != 0.0f) sfac= data->dist / dist;
}
/* if soft-distance is enabled, start fading once owner is dist+softdist from the target */
else if (data->flag & LIMITDIST_USESOFT) {
if (dist <= (data->dist + data->soft)) {
}
}
}
else if (data->mode == LIMITDIST_INSIDE) {
/* if outside, then move to surface */
if (dist >= data->dist) {
clamp_surf= 1;
if (dist != 0.0f) sfac= data->dist / dist;
}
/* if soft-distance is enabled, start fading once owner is dist-soft from the target */
else if (data->flag & LIMITDIST_USESOFT) {
// FIXME: there's a problem with "jumping" when this kicks in
if (dist >= (data->dist - data->soft)) {
sfac = (float)( data->soft*(1.0f - expf(-(dist - data->dist)/data->soft)) + data->dist );
if (dist != 0.0f) sfac /= dist;
clamp_surf= 1;
}
}
}
else {
if (IS_EQF(dist, data->dist)==0) {
clamp_surf= 1;
if (dist != 0.0f) sfac= data->dist / dist;
}
}
/* clamp to 'surface' (i.e. move owner so that dist == data->dist) */
if (clamp_surf) {
/* simply interpolate along line formed by target -> owner */
interp_v3_v3v3(dvec, ct->matrix[3], cob->matrix[3], sfac);
/* copy new vector onto owner */
2010-08-15 15:14:08 +00:00
copy_v3_v3(cob->matrix[3], dvec);
}
}
}
static bConstraintTypeInfo CTI_DISTLIMIT = {
CONSTRAINT_TYPE_DISTLIMIT, /* type */
sizeof(bDistLimitConstraint), /* size */
"Limit Distance", /* name */
"bDistLimitConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
distlimit_id_looper, /* id looper */
NULL, /* copy data */
distlimit_new_data, /* new data */
distlimit_get_tars, /* get constraint targets */
distlimit_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get a target matrix */
distlimit_evaluate /* evaluate */
};
/* ---------- Stretch To ------------ */
static void stretchto_new_data (void *cdata)
{
bStretchToConstraint *data= (bStretchToConstraint *)cdata;
data->volmode = 0;
data->plane = 0;
data->orglength = 0.0;
data->bulge = 1.0;
}
static void stretchto_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bStretchToConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int stretchto_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bStretchToConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void stretchto_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bStretchToConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void stretchto_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bStretchToConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
/* only evaluate if there is a target */
if (VALID_CONS_TARGET(ct)) {
float size[3], scale[3], vec[3], xx[3], zz[3], orth[3];
float totmat[3][3];
float tmat[4][4];
float dist;
/* store scaling before destroying obmat */
mat4_to_size(size, cob->matrix);
/* store X orientation before destroying obmat */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(xx, cob->matrix[0]);
/* store Z orientation before destroying obmat */
2010-08-15 15:14:08 +00:00
normalize_v3_v3(zz, cob->matrix[2]);
/* XXX That makes the constraint buggy with asymmetrically scaled objects, see #29940. */
/* sub_v3_v3v3(vec, cob->matrix[3], ct->matrix[3]);*/
/* vec[0] /= size[0];*/
/* vec[1] /= size[1];*/
/* vec[2] /= size[2];*/
/* dist = normalize_v3(vec);*/
dist = len_v3v3(cob->matrix[3], ct->matrix[3]);
/* Only Y constrained object axis scale should be used, to keep same length when scaling it. */
dist /= size[1];
/* data->orglength==0 occurs on first run, and after 'R' button is clicked */
if (data->orglength == 0)
data->orglength = dist;
if (data->bulge == 0)
data->bulge = 1.0;
scale[1] = dist/data->orglength;
switch (data->volmode) {
/* volume preserving scaling */
case VOLUME_XZ :
scale[0] = 1.0f - (float)sqrt(data->bulge) + (float)sqrt(data->bulge*(data->orglength/dist));
scale[2] = scale[0];
break;
case VOLUME_X:
scale[0] = 1.0f + data->bulge * (data->orglength /dist - 1);
scale[2] = 1.0;
break;
case VOLUME_Z:
scale[0] = 1.0;
scale[2] = 1.0f + data->bulge * (data->orglength /dist - 1);
break;
/* don't care for volume */
case NO_VOLUME:
scale[0] = 1.0;
scale[2] = 1.0;
break;
default: /* should not happen, but in case*/
return;
} /* switch (data->volmode) */
/* Clear the object's rotation and scale */
cob->matrix[0][0]=size[0]*scale[0];
cob->matrix[0][1]=0;
cob->matrix[0][2]=0;
cob->matrix[1][0]=0;
cob->matrix[1][1]=size[1]*scale[1];
cob->matrix[1][2]=0;
cob->matrix[2][0]=0;
cob->matrix[2][1]=0;
cob->matrix[2][2]=size[2]*scale[2];
sub_v3_v3v3(vec, cob->matrix[3], ct->matrix[3]);
normalize_v3(vec);
/* new Y aligns object target connection*/
2010-08-15 15:14:08 +00:00
negate_v3_v3(totmat[1], vec);
switch (data->plane) {
case PLANE_X:
/* build new Z vector */
/* othogonal to "new Y" "old X! plane */
cross_v3_v3v3(orth, vec, xx);
normalize_v3(orth);
/* new Z*/
2010-08-15 15:14:08 +00:00
copy_v3_v3(totmat[2], orth);
/* we decided to keep X plane*/
cross_v3_v3v3(xx, orth, vec);
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[0], xx);
break;
case PLANE_Z:
/* build new X vector */
/* othogonal to "new Y" "old Z! plane */
cross_v3_v3v3(orth, vec, zz);
normalize_v3(orth);
/* new X */
2010-08-15 15:14:08 +00:00
negate_v3_v3(totmat[0], orth);
/* we decided to keep Z */
cross_v3_v3v3(zz, orth, vec);
2010-08-15 15:14:08 +00:00
normalize_v3_v3(totmat[2], zz);
break;
} /* switch (data->plane) */
copy_m4_m4(tmat, cob->matrix);
mul_m4_m3m4(cob->matrix, totmat, tmat);
}
}
static bConstraintTypeInfo CTI_STRETCHTO = {
CONSTRAINT_TYPE_STRETCHTO, /* type */
sizeof(bStretchToConstraint), /* size */
"Stretch To", /* name */
"bStretchToConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
stretchto_id_looper, /* id looper */
NULL, /* copy data */
stretchto_new_data, /* new data */
stretchto_get_tars, /* get constraint targets */
stretchto_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
stretchto_evaluate /* evaluate */
};
/* ---------- Floor ------------ */
static void minmax_new_data (void *cdata)
{
bMinMaxConstraint *data= (bMinMaxConstraint *)cdata;
data->minmaxflag = TRACK_Z;
data->offset = 0.0f;
2012-03-23 20:18:09 +00:00
zero_v3(data->cache);
data->flag = 0;
}
static void minmax_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bMinMaxConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int minmax_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bMinMaxConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void minmax_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bMinMaxConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void minmax_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bMinMaxConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
/* only evaluate if there is a target */
if (VALID_CONS_TARGET(ct)) {
float obmat[4][4], imat[4][4], tarmat[4][4], tmat[4][4];
float val1, val2;
int index;
copy_m4_m4(obmat, cob->matrix);
copy_m4_m4(tarmat, ct->matrix);
if (data->flag & MINMAX_USEROT) {
/* take rotation of target into account by doing the transaction in target's localspace */
invert_m4_m4(imat, tarmat);
mult_m4_m4m4(tmat, imat, obmat);
copy_m4_m4(obmat, tmat);
unit_m4(tarmat);
}
switch (data->minmaxflag) {
case TRACK_Z:
val1 = tarmat[3][2];
val2 = obmat[3][2]-data->offset;
index = 2;
break;
case TRACK_Y:
val1 = tarmat[3][1];
val2 = obmat[3][1]-data->offset;
index = 1;
break;
case TRACK_X:
val1 = tarmat[3][0];
val2 = obmat[3][0]-data->offset;
index = 0;
break;
case TRACK_nZ:
val2 = tarmat[3][2];
val1 = obmat[3][2]-data->offset;
index = 2;
break;
case TRACK_nY:
val2 = tarmat[3][1];
val1 = obmat[3][1]-data->offset;
index = 1;
break;
case TRACK_nX:
val2 = tarmat[3][0];
val1 = obmat[3][0]-data->offset;
index = 0;
break;
default:
return;
}
if (val1 > val2) {
obmat[3][index] = tarmat[3][index] + data->offset;
if (data->flag & MINMAX_STICKY) {
if (data->flag & MINMAX_STUCK) {
2010-08-15 15:14:08 +00:00
copy_v3_v3(obmat[3], data->cache);
}
else {
2010-08-15 15:14:08 +00:00
copy_v3_v3(data->cache, obmat[3]);
data->flag |= MINMAX_STUCK;
}
}
if (data->flag & MINMAX_USEROT) {
/* get out of localspace */
mult_m4_m4m4(tmat, ct->matrix, obmat);
copy_m4_m4(cob->matrix, tmat);
}
else {
2010-08-15 15:14:08 +00:00
copy_v3_v3(cob->matrix[3], obmat[3]);
}
}
else {
data->flag &= ~MINMAX_STUCK;
}
}
}
static bConstraintTypeInfo CTI_MINMAX = {
CONSTRAINT_TYPE_MINMAX, /* type */
sizeof(bMinMaxConstraint), /* size */
"Floor", /* name */
"bMinMaxConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
minmax_id_looper, /* id looper */
NULL, /* copy data */
minmax_new_data, /* new data */
minmax_get_tars, /* get constraint targets */
minmax_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
minmax_evaluate /* evaluate */
};
/* ------- RigidBody Joint ---------- */
static void rbj_new_data (void *cdata)
{
bRigidBodyJointConstraint *data= (bRigidBodyJointConstraint *)cdata;
// removed code which set target of this constraint
data->type=1;
}
static void rbj_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bRigidBodyJointConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
func(con, (ID**)&data->child, userdata);
}
static int rbj_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bRigidBodyJointConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints without subtargets */
SINGLETARGETNS_GET_TARS(con, data->tar, ct, list)
return 1;
}
return 0;
}
static void rbj_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bRigidBodyJointConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGETNS_FLUSH_TARS(con, data->tar, ct, list, nocopy)
}
}
static bConstraintTypeInfo CTI_RIGIDBODYJOINT = {
CONSTRAINT_TYPE_RIGIDBODYJOINT, /* type */
sizeof(bRigidBodyJointConstraint), /* size */
"Rigid Body Joint", /* name */
"bRigidBodyJointConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
rbj_id_looper, /* id looper */
NULL, /* copy data */
rbj_new_data, /* new data */
rbj_get_tars, /* get constraint targets */
rbj_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
NULL /* evaluate - this is not solved here... is just an interface for game-engine */
};
/* -------- Clamp To ---------- */
static void clampto_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bClampToConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int clampto_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bClampToConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints without subtargets */
SINGLETARGETNS_GET_TARS(con, data->tar, ct, list)
return 1;
}
return 0;
}
static void clampto_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bClampToConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGETNS_FLUSH_TARS(con, data->tar, ct, list, nocopy)
}
}
static void clampto_get_tarmat (bConstraint *UNUSED(con), bConstraintOb *cob, bConstraintTarget *ct, float UNUSED(ctime))
{
if (VALID_CONS_TARGET(ct)) {
Curve *cu= ct->tar->data;
/* note: when creating constraints that follow path, the curve gets the CU_PATH set now,
* currently for paths to work it needs to go through the bevlist/displist system (ton)
*/
/* only happens on reload file, but violates depsgraph still... fix! */
2009-09-16 17:43:09 +00:00
if (cu->path==NULL || cu->path->data==NULL)
makeDispListCurveTypes(cob->scene, ct->tar, 0);
}
/* technically, this isn't really needed for evaluation, but we don't know what else
* might end up calling this...
*/
if (ct)
unit_m4(ct->matrix);
}
static void clampto_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bClampToConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
/* only evaluate if there is a target and it is a curve */
if (VALID_CONS_TARGET(ct) && (ct->tar->type == OB_CURVE)) {
Curve *cu= data->tar->data;
float obmat[4][4], ownLoc[3];
float curveMin[3], curveMax[3];
float targetMatrix[4][4]= MAT4_UNITY;
copy_m4_m4(obmat, cob->matrix);
2010-08-15 15:14:08 +00:00
copy_v3_v3(ownLoc, obmat[3]);
INIT_MINMAX(curveMin, curveMax)
BKE_object_minmax(ct->tar, curveMin, curveMax);
/* get targetmatrix */
if (cu->path && cu->path->data) {
float vec[4], dir[3], totmat[4][4];
float curvetime;
short clamp_axis;
/* find best position on curve */
/* 1. determine which axis to sample on? */
if (data->flag == CLAMPTO_AUTO) {
float size[3];
sub_v3_v3v3(size, curveMax, curveMin);
/* find axis along which the bounding box has the greatest
* extent. Otherwise, default to the x-axis, as that is quite
* frequently used.
*/
if ((size[2]>size[0]) && (size[2]>size[1]))
clamp_axis= CLAMPTO_Z - 1;
else if ((size[1]>size[0]) && (size[1]>size[2]))
clamp_axis= CLAMPTO_Y - 1;
else
clamp_axis = CLAMPTO_X - 1;
}
else
clamp_axis= data->flag - 1;
/* 2. determine position relative to curve on a 0-1 scale based on bounding box */
if (data->flag2 & CLAMPTO_CYCLIC) {
/* cyclic, so offset within relative bounding box is used */
float len= (curveMax[clamp_axis] - curveMin[clamp_axis]);
float offset;
/* check to make sure len is not so close to zero that it'll cause errors */
if (IS_EQ(len, 0) == 0) {
/* find bounding-box range where target is located */
if (ownLoc[clamp_axis] < curveMin[clamp_axis]) {
/* bounding-box range is before */
offset = curveMin[clamp_axis] - ceil((curveMin[clamp_axis] - ownLoc[clamp_axis]) / len) * len;
/* now, we calculate as per normal, except using offset instead of curveMin[clamp_axis] */
curvetime = (ownLoc[clamp_axis] - offset) / (len);
}
else if (ownLoc[clamp_axis] > curveMax[clamp_axis]) {
/* bounding-box range is after */
offset= curveMax[clamp_axis] + (int)((ownLoc[clamp_axis] - curveMax[clamp_axis]) / len) * len;
/* now, we calculate as per normal, except using offset instead of curveMax[clamp_axis] */
curvetime = (ownLoc[clamp_axis] - offset) / (len);
}
else {
/* as the location falls within bounds, just calculate */
curvetime = (ownLoc[clamp_axis] - curveMin[clamp_axis]) / (len);
}
}
else {
/* as length is close to zero, curvetime by default should be 0 (i.e. the start) */
curvetime= 0.0f;
}
}
else {
/* no cyclic, so position is clamped to within the bounding box */
if (ownLoc[clamp_axis] <= curveMin[clamp_axis])
curvetime = 0.0f;
else if (ownLoc[clamp_axis] >= curveMax[clamp_axis])
curvetime = 1.0f;
else if ( IS_EQ((curveMax[clamp_axis] - curveMin[clamp_axis]), 0) == 0 )
curvetime = (ownLoc[clamp_axis] - curveMin[clamp_axis]) / (curveMax[clamp_axis] - curveMin[clamp_axis]);
else
curvetime = 0.0f;
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* 3. position on curve */
if (where_on_path(ct->tar, curvetime, vec, dir, NULL, NULL, NULL) ) {
unit_m4(totmat);
2010-08-15 15:14:08 +00:00
copy_v3_v3(totmat[3], vec);
mul_serie_m4(targetMatrix, ct->tar->obmat, totmat, NULL, NULL, NULL, NULL, NULL, NULL);
}
}
/* obtain final object position */
2010-08-15 15:14:08 +00:00
copy_v3_v3(cob->matrix[3], targetMatrix[3]);
}
}
static bConstraintTypeInfo CTI_CLAMPTO = {
CONSTRAINT_TYPE_CLAMPTO, /* type */
sizeof(bClampToConstraint), /* size */
"Clamp To", /* name */
"bClampToConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
clampto_id_looper, /* id looper */
NULL, /* copy data */
NULL, /* new data */
clampto_get_tars, /* get constraint targets */
clampto_flush_tars, /* flush constraint targets */
clampto_get_tarmat, /* get target matrix */
clampto_evaluate /* evaluate */
};
/* ---------- Transform Constraint ----------- */
static void transform_new_data (void *cdata)
{
bTransformConstraint *data= (bTransformConstraint *)cdata;
data->map[0]= 0;
data->map[1]= 1;
data->map[2]= 2;
}
static void transform_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bTransformConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int transform_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bTransformConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void transform_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bTransformConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void transform_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bTransformConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
/* only evaluate if there is a target */
if (VALID_CONS_TARGET(ct)) {
float loc[3], eul[3], size[3];
float dvec[3], sval[3];
int i;
/* obtain target effect */
switch (data->from) {
case 2: /* scale */
2012-04-29 15:47:02 +00:00
mat4_to_size(dvec, ct->matrix);
break;
case 1: /* rotation (convert to degrees first) */
mat4_to_eulO(dvec, cob->rotOrder, ct->matrix);
mul_v3_fl(dvec, RAD2DEGF(1.0f)); /* rad -> deg */
break;
default: /* location */
copy_v3_v3(dvec, ct->matrix[3]);
break;
}
/* extract components of owner's matrix */
2010-08-15 15:14:08 +00:00
copy_v3_v3(loc, cob->matrix[3]);
mat4_to_eulO(eul, cob->rotOrder, cob->matrix);
mat4_to_size(size, cob->matrix);
/* determine where in range current transforms lie */
if (data->expo) {
for (i=0; i<3; i++) {
if (data->from_max[i] - data->from_min[i])
sval[i]= (dvec[i] - data->from_min[i]) / (data->from_max[i] - data->from_min[i]);
else
sval[i]= 0.0f;
}
}
else {
/* clamp transforms out of range */
for (i=0; i<3; i++) {
CLAMP(dvec[i], data->from_min[i], data->from_max[i]);
if (data->from_max[i] - data->from_min[i])
sval[i]= (dvec[i] - data->from_min[i]) / (data->from_max[i] - data->from_min[i]);
else
sval[i]= 0.0f;
}
}
/* apply transforms */
switch (data->to) {
case 2: /* scaling */
for (i=0; i<3; i++)
size[i]= data->to_min[i] + (sval[(int)data->map[i]] * (data->to_max[i] - data->to_min[i]));
break;
case 1: /* rotation */
for (i=0; i<3; i++) {
float tmin, tmax;
tmin= data->to_min[i];
tmax= data->to_max[i];
/* all values here should be in degrees */
eul[i]= tmin + (sval[(int)data->map[i]] * (tmax - tmin));
/* now convert final value back to radians */
eul[i] = DEG2RADF(eul[i]);
}
break;
default: /* location */
/* get new location */
for (i=0; i<3; i++)
loc[i]= (data->to_min[i] + (sval[(int)data->map[i]] * (data->to_max[i] - data->to_min[i])));
/* add original location back on (so that it can still be moved) */
add_v3_v3v3(loc, cob->matrix[3], loc);
break;
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
/* apply to matrix */
loc_eulO_size_to_mat4(cob->matrix, loc, eul, size, cob->rotOrder);
}
}
static bConstraintTypeInfo CTI_TRANSFORM = {
CONSTRAINT_TYPE_TRANSFORM, /* type */
sizeof(bTransformConstraint), /* size */
"Transform", /* name */
"bTransformConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
transform_id_looper, /* id looper */
NULL, /* copy data */
transform_new_data, /* new data */
transform_get_tars, /* get constraint targets */
transform_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get a target matrix */
transform_evaluate /* evaluate */
};
/* ---------- Shrinkwrap Constraint ----------- */
static void shrinkwrap_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bShrinkwrapConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->target, userdata);
}
static int shrinkwrap_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bShrinkwrapConstraint *data = con->data;
bConstraintTarget *ct;
SINGLETARGETNS_GET_TARS(con, data->target, ct, list)
return 1;
}
return 0;
}
static void shrinkwrap_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bShrinkwrapConstraint *data = con->data;
bConstraintTarget *ct= list->first;
SINGLETARGETNS_FLUSH_TARS(con, data->target, ct, list, nocopy)
}
}
static void shrinkwrap_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float UNUSED(ctime))
{
bShrinkwrapConstraint *scon = (bShrinkwrapConstraint *) con->data;
if ( VALID_CONS_TARGET(ct) && (ct->tar->type == OB_MESH) ) {
int fail = FALSE;
float co[3] = {0.0f, 0.0f, 0.0f};
float no[3] = {0.0f, 0.0f, 0.0f};
float dist;
SpaceTransform transform;
DerivedMesh *target = object_get_derived_final(ct->tar);
BVHTreeRayHit hit;
BVHTreeNearest nearest;
BVHTreeFromMesh treeData= {NULL};
nearest.index = -1;
nearest.dist = FLT_MAX;
hit.index = -1;
hit.dist = 100000.0f; //TODO should use FLT_MAX.. but normal projection doenst yet supports it
unit_m4(ct->matrix);
if (target != NULL) {
space_transform_from_matrixs(&transform, cob->matrix, ct->tar->obmat);
switch (scon->shrinkType) {
case MOD_SHRINKWRAP_NEAREST_SURFACE:
case MOD_SHRINKWRAP_NEAREST_VERTEX:
if (scon->shrinkType == MOD_SHRINKWRAP_NEAREST_VERTEX)
bvhtree_from_mesh_verts(&treeData, target, 0.0, 2, 6);
else
bvhtree_from_mesh_faces(&treeData, target, 0.0, 2, 6);
if (treeData.tree == NULL) {
fail = TRUE;
break;
}
space_transform_apply(&transform, co);
BLI_bvhtree_find_nearest(treeData.tree, co, &nearest, treeData.nearest_callback, &treeData);
dist = len_v3v3(co, nearest.co);
if (dist != 0.0f) {
interp_v3_v3v3(co, co, nearest.co, (dist - scon->dist)/dist); /* linear interpolation */
}
space_transform_invert(&transform, co);
break;
case MOD_SHRINKWRAP_PROJECT:
if (scon->projAxis & MOD_SHRINKWRAP_PROJECT_OVER_X_AXIS) no[0] = 1.0f;
if (scon->projAxis & MOD_SHRINKWRAP_PROJECT_OVER_Y_AXIS) no[1] = 1.0f;
if (scon->projAxis & MOD_SHRINKWRAP_PROJECT_OVER_Z_AXIS) no[2] = 1.0f;
if (dot_v3v3(no, no) < FLT_EPSILON) {
fail = TRUE;
break;
}
normalize_v3(no);
bvhtree_from_mesh_faces(&treeData, target, scon->dist, 4, 6);
if (treeData.tree == NULL) {
fail = TRUE;
break;
}
if (normal_projection_project_vertex(0, co, no, &transform, treeData.tree, &hit, treeData.raycast_callback, &treeData) == FALSE) {
fail = TRUE;
break;
}
2010-08-15 15:14:08 +00:00
copy_v3_v3(co, hit.co);
break;
}
free_bvhtree_from_mesh(&treeData);
target->release(target);
2012-03-23 20:18:09 +00:00
if (fail == TRUE) {
/* Don't move the point */
2012-03-23 20:18:09 +00:00
zero_v3(co);
}
/* co is in local object coordinates, change it to global and update target position */
mul_m4_v3(cob->matrix, co);
2010-08-15 15:14:08 +00:00
copy_v3_v3(ct->matrix[3], co);
}
}
}
static void shrinkwrap_evaluate (bConstraint *UNUSED(con), bConstraintOb *cob, ListBase *targets)
{
bConstraintTarget *ct= targets->first;
/* only evaluate if there is a target */
if (VALID_CONS_TARGET(ct)) {
2010-08-15 15:14:08 +00:00
copy_v3_v3(cob->matrix[3], ct->matrix[3]);
}
}
static bConstraintTypeInfo CTI_SHRINKWRAP = {
CONSTRAINT_TYPE_SHRINKWRAP, /* type */
sizeof(bShrinkwrapConstraint), /* size */
"Shrinkwrap", /* name */
"bShrinkwrapConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
shrinkwrap_id_looper, /* id looper */
NULL, /* copy data */
NULL, /* new data */
shrinkwrap_get_tars, /* get constraint targets */
shrinkwrap_flush_tars, /* flush constraint targets */
shrinkwrap_get_tarmat, /* get a target matrix */
shrinkwrap_evaluate /* evaluate */
};
/* --------- Damped Track ---------- */
static void damptrack_new_data (void *cdata)
{
bDampTrackConstraint *data= (bDampTrackConstraint *)cdata;
data->trackflag = TRACK_Y;
}
static void damptrack_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bDampTrackConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int damptrack_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bDampTrackConstraint *data= con->data;
bConstraintTarget *ct;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void damptrack_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bDampTrackConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
/* array of direction vectors for the tracking flags */
static const float track_dir_vecs[6][3] = {
2012-04-29 15:47:02 +00:00
{+1, 0, 0}, {0, +1, 0}, {0, 0, +1}, /* TRACK_X, TRACK_Y, TRACK_Z */
{-1, 0, 0}, {0, -1, 0}, {0, 0, -1} /* TRACK_NX, TRACK_NY, TRACK_NZ */
};
static void damptrack_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bDampTrackConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
if (VALID_CONS_TARGET(ct)) {
float obvec[3], tarvec[3], obloc[3];
float raxis[3], rangle;
float rmat[3][3], tmat[4][4];
/* find the (unit) direction that the axis we're interested in currently points
* - mul_mat3_m4_v3() only takes the 3x3 (rotation+scaling) components of the 4x4 matrix
* - the normalization step at the end should take care of any unwanted scaling
* left over in the 3x3 matrix we used
*/
2010-08-15 15:14:08 +00:00
copy_v3_v3(obvec, track_dir_vecs[data->trackflag]);
mul_mat3_m4_v3(cob->matrix, obvec);
if (normalize_v3(obvec) == 0.0f) {
/* exceptional case - just use the track vector as appropriate */
2010-08-15 15:14:08 +00:00
copy_v3_v3(obvec, track_dir_vecs[data->trackflag]);
}
/* find the (unit) direction vector going from the owner to the target */
2010-08-15 15:14:08 +00:00
copy_v3_v3(obloc, cob->matrix[3]);
sub_v3_v3v3(tarvec, ct->matrix[3], obloc);
if (normalize_v3(tarvec) == 0.0f) {
/* the target is sitting on the owner, so just make them use the same direction vectors */
// FIXME: or would it be better to use the pure direction vector?
2010-08-15 15:14:08 +00:00
copy_v3_v3(tarvec, obvec);
//copy_v3_v3(tarvec, track_dir_vecs[data->trackflag]);
}
/* determine the axis-angle rotation, which represents the smallest possible rotation
* between the two rotation vectors (i.e. the 'damping' referred to in the name)
* - we take this to be the rotation around the normal axis/vector to the plane defined
* by the current and destination vectors, which will 'map' the current axis to the
* destination vector
* - the min/max wrappers around (obvec . tarvec) result (stored temporarily in rangle)
* are used to ensure that the smallest angle is chosen
*/
cross_v3_v3v3(raxis, obvec, tarvec);
rangle= dot_v3v3(obvec, tarvec);
rangle= acos( MAX2(-1.0f, MIN2(1.0f, rangle)) );
/* construct rotation matrix from the axis-angle rotation found above
* - this call takes care to make sure that the axis provided is a unit vector first
*/
axis_angle_to_mat3(rmat, raxis, rangle);
/* rotate the owner in the way defined by this rotation matrix, then reapply the location since
* we may have destroyed that in the process of multiplying the matrix
*/
unit_m4(tmat);
mul_m4_m3m4(tmat, rmat, cob->matrix); // m1, m3, m2
copy_m4_m4(cob->matrix, tmat);
2010-08-15 15:14:08 +00:00
copy_v3_v3(cob->matrix[3], obloc);
}
}
static bConstraintTypeInfo CTI_DAMPTRACK = {
CONSTRAINT_TYPE_DAMPTRACK, /* type */
sizeof(bDampTrackConstraint), /* size */
"Damped Track", /* name */
"bDampTrackConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
damptrack_id_looper, /* id looper */
NULL, /* copy data */
damptrack_new_data, /* new data */
damptrack_get_tars, /* get constraint targets */
damptrack_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
damptrack_evaluate /* evaluate */
};
/* ----------- Spline IK ------------ */
static void splineik_free (bConstraint *con)
{
bSplineIKConstraint *data= con->data;
/* binding array */
if (data->points)
MEM_freeN(data->points);
}
static void splineik_copy (bConstraint *con, bConstraint *srccon)
{
bSplineIKConstraint *src= srccon->data;
bSplineIKConstraint *dst= con->data;
/* copy the binding array */
dst->points= MEM_dupallocN(src->points);
}
static void splineik_new_data (void *cdata)
{
bSplineIKConstraint *data= (bSplineIKConstraint *)cdata;
data->chainlen= 1;
}
static void splineik_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bSplineIKConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int splineik_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bSplineIKConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints without subtargets */
SINGLETARGETNS_GET_TARS(con, data->tar, ct, list)
return 1;
}
return 0;
}
static void splineik_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bSplineIKConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGETNS_FLUSH_TARS(con, data->tar, ct, list, nocopy)
}
}
static void splineik_get_tarmat (bConstraint *UNUSED(con), bConstraintOb *cob, bConstraintTarget *ct, float UNUSED(ctime))
{
if (VALID_CONS_TARGET(ct)) {
Curve *cu= ct->tar->data;
/* note: when creating constraints that follow path, the curve gets the CU_PATH set now,
* currently for paths to work it needs to go through the bevlist/displist system (ton)
*/
/* only happens on reload file, but violates depsgraph still... fix! */
if (cu->path==NULL || cu->path->data==NULL)
makeDispListCurveTypes(cob->scene, ct->tar, 0);
}
/* technically, this isn't really needed for evaluation, but we don't know what else
* might end up calling this...
*/
if (ct)
unit_m4(ct->matrix);
}
static bConstraintTypeInfo CTI_SPLINEIK = {
CONSTRAINT_TYPE_SPLINEIK, /* type */
sizeof(bSplineIKConstraint), /* size */
"Spline IK", /* name */
"bSplineIKConstraint", /* struct name */
splineik_free, /* free data */
NULL, /* relink data */
splineik_id_looper, /* id looper */
splineik_copy, /* copy data */
splineik_new_data, /* new data */
splineik_get_tars, /* get constraint targets */
splineik_flush_tars, /* flush constraint targets */
splineik_get_tarmat, /* get target matrix */
NULL /* evaluate - solved as separate loop */
};
== Pivot Constraint == This constraint allows an object or bone to have their rotations applied as if their origin/pivot-point was located elsewhere. The most obvious uses include foot-roll, see-saws, but could also include more complicated rolling-box examples. == Usage Examples == === Foot Roll === 1. Add 'Pivot' Constraint to the bone without any target. 2. Set the 'Y' value of the offset to the length of the bone. Usually this should be negative (if you rig with feet facing 'forwards' along -Y axis). This gives you a pivot point relative to the bone's (preconstraint) location, which should be at the tip of the bone here. Disabling the 'Use Relative Offset' would make this offset be relative to 0,0,0 instead of to the owner/bone-head. 3. Ensure that the 'Pivot When' setting is set to '-X Rot', (default) which means that the pivot will only used when the rotation on the X-Axis is negative to get tip-toe 'roll'. === See Saw === 1. Add a 'Pivot' constraint too see-saw plank object, this time with a target that you wish to have as the pivot-point. It's possible to do this without too (as before), but is less intuitive. 2. Optionally, if you want the plank slightly raised, set the z-offset value, which should make the pivot-point used to be relative to the target with the z-offset applied. 3. Ensure that 'Pivot When' is set to 'Always', which means that the pivot will always be used, irrespective of the rotation. == Notes == * The 'Pivot When' setting has been integrated in the constraint, since this is something that will often be required for these setups. Having to set up additional drivers to drive the constraint to do this kindof beats the purpose of providing this. * The 'Offset' functionality is probably not presented as clearly as it could be. We may need to go over this again. * For foot-roll - if any scaling of the foot is required, simply set up a driver on the y-offset to make this dynamically respond to the "scale" RNA property of the bones (don't use the "Transform Channel" vartype since that won't work correct here). However, this shouldn't be common enough to warrant special treatment.
2010-05-27 10:50:06 +00:00
/* ----------- Pivot ------------- */
static void pivotcon_id_looper (bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bPivotConstraint *data= con->data;
/* target only */
func(con, (ID**)&data->tar, userdata);
}
static int pivotcon_get_tars (bConstraint *con, ListBase *list)
{
if (con && list) {
bPivotConstraint *data= con->data;
bConstraintTarget *ct;
/* standard target-getting macro for single-target constraints */
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
return 1;
}
return 0;
}
static void pivotcon_flush_tars (bConstraint *con, ListBase *list, short nocopy)
{
if (con && list) {
bPivotConstraint *data= con->data;
bConstraintTarget *ct= list->first;
/* the following macro is used for all standard single-target constraints */
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
}
}
static void pivotcon_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
{
bPivotConstraint *data= con->data;
bConstraintTarget *ct= targets->first;
float pivot[3], vec[3];
float rotMat[3][3];
/* pivot correction */
float axis[3], angle;
== Pivot Constraint == This constraint allows an object or bone to have their rotations applied as if their origin/pivot-point was located elsewhere. The most obvious uses include foot-roll, see-saws, but could also include more complicated rolling-box examples. == Usage Examples == === Foot Roll === 1. Add 'Pivot' Constraint to the bone without any target. 2. Set the 'Y' value of the offset to the length of the bone. Usually this should be negative (if you rig with feet facing 'forwards' along -Y axis). This gives you a pivot point relative to the bone's (preconstraint) location, which should be at the tip of the bone here. Disabling the 'Use Relative Offset' would make this offset be relative to 0,0,0 instead of to the owner/bone-head. 3. Ensure that the 'Pivot When' setting is set to '-X Rot', (default) which means that the pivot will only used when the rotation on the X-Axis is negative to get tip-toe 'roll'. === See Saw === 1. Add a 'Pivot' constraint too see-saw plank object, this time with a target that you wish to have as the pivot-point. It's possible to do this without too (as before), but is less intuitive. 2. Optionally, if you want the plank slightly raised, set the z-offset value, which should make the pivot-point used to be relative to the target with the z-offset applied. 3. Ensure that 'Pivot When' is set to 'Always', which means that the pivot will always be used, irrespective of the rotation. == Notes == * The 'Pivot When' setting has been integrated in the constraint, since this is something that will often be required for these setups. Having to set up additional drivers to drive the constraint to do this kindof beats the purpose of providing this. * The 'Offset' functionality is probably not presented as clearly as it could be. We may need to go over this again. * For foot-roll - if any scaling of the foot is required, simply set up a driver on the y-offset to make this dynamically respond to the "scale" RNA property of the bones (don't use the "Transform Channel" vartype since that won't work correct here). However, this shouldn't be common enough to warrant special treatment.
2010-05-27 10:50:06 +00:00
/* firstly, check if pivoting should take place based on the current rotation */
if (data->rotAxis != PIVOTCON_AXIS_NONE) {
float rot[3];
/* extract euler-rotation of target */
mat4_to_eulO(rot, cob->rotOrder, cob->matrix);
/* check which range might be violated */
if (data->rotAxis < PIVOTCON_AXIS_X) {
/* negative rotations (data->rotAxis = 0 -> 2) */
if (rot[data->rotAxis] > 0.0f)
return;
}
else {
/* positive rotations (data->rotAxis = 3 -> 5 */
if (rot[data->rotAxis - PIVOTCON_AXIS_X] < 0.0f)
return;
}
}
/* find the pivot-point to use */
if (VALID_CONS_TARGET(ct)) {
/* apply offset to target location */
add_v3_v3v3(pivot, ct->matrix[3], data->offset);
}
else {
/* no targets to worry about... */
if ((data->flag & PIVOTCON_FLAG_OFFSET_ABS) == 0) {
/* offset is relative to owner */
add_v3_v3v3(pivot, cob->matrix[3], data->offset);
}
else {
/* directly use the 'offset' specified as an absolute position instead */
2010-08-15 15:14:08 +00:00
copy_v3_v3(pivot, data->offset);
== Pivot Constraint == This constraint allows an object or bone to have their rotations applied as if their origin/pivot-point was located elsewhere. The most obvious uses include foot-roll, see-saws, but could also include more complicated rolling-box examples. == Usage Examples == === Foot Roll === 1. Add 'Pivot' Constraint to the bone without any target. 2. Set the 'Y' value of the offset to the length of the bone. Usually this should be negative (if you rig with feet facing 'forwards' along -Y axis). This gives you a pivot point relative to the bone's (preconstraint) location, which should be at the tip of the bone here. Disabling the 'Use Relative Offset' would make this offset be relative to 0,0,0 instead of to the owner/bone-head. 3. Ensure that the 'Pivot When' setting is set to '-X Rot', (default) which means that the pivot will only used when the rotation on the X-Axis is negative to get tip-toe 'roll'. === See Saw === 1. Add a 'Pivot' constraint too see-saw plank object, this time with a target that you wish to have as the pivot-point. It's possible to do this without too (as before), but is less intuitive. 2. Optionally, if you want the plank slightly raised, set the z-offset value, which should make the pivot-point used to be relative to the target with the z-offset applied. 3. Ensure that 'Pivot When' is set to 'Always', which means that the pivot will always be used, irrespective of the rotation. == Notes == * The 'Pivot When' setting has been integrated in the constraint, since this is something that will often be required for these setups. Having to set up additional drivers to drive the constraint to do this kindof beats the purpose of providing this. * The 'Offset' functionality is probably not presented as clearly as it could be. We may need to go over this again. * For foot-roll - if any scaling of the foot is required, simply set up a driver on the y-offset to make this dynamically respond to the "scale" RNA property of the bones (don't use the "Transform Channel" vartype since that won't work correct here). However, this shouldn't be common enough to warrant special treatment.
2010-05-27 10:50:06 +00:00
}
}
/* get rotation matrix representing the rotation of the owner */
// TODO: perhaps we might want to include scaling based on the pivot too?
copy_m3_m4(rotMat, cob->matrix);
normalize_m3(rotMat);
/* correct the pivot by the rotation axis otherwise the pivot translates when it shouldnt */
mat3_to_axis_angle(axis, &angle, rotMat);
if (angle) {
float dvec[3];
sub_v3_v3v3(vec, pivot, cob->matrix[3]);
project_v3_v3v3(dvec, vec, axis);
sub_v3_v3(pivot, dvec);
}
== Pivot Constraint == This constraint allows an object or bone to have their rotations applied as if their origin/pivot-point was located elsewhere. The most obvious uses include foot-roll, see-saws, but could also include more complicated rolling-box examples. == Usage Examples == === Foot Roll === 1. Add 'Pivot' Constraint to the bone without any target. 2. Set the 'Y' value of the offset to the length of the bone. Usually this should be negative (if you rig with feet facing 'forwards' along -Y axis). This gives you a pivot point relative to the bone's (preconstraint) location, which should be at the tip of the bone here. Disabling the 'Use Relative Offset' would make this offset be relative to 0,0,0 instead of to the owner/bone-head. 3. Ensure that the 'Pivot When' setting is set to '-X Rot', (default) which means that the pivot will only used when the rotation on the X-Axis is negative to get tip-toe 'roll'. === See Saw === 1. Add a 'Pivot' constraint too see-saw plank object, this time with a target that you wish to have as the pivot-point. It's possible to do this without too (as before), but is less intuitive. 2. Optionally, if you want the plank slightly raised, set the z-offset value, which should make the pivot-point used to be relative to the target with the z-offset applied. 3. Ensure that 'Pivot When' is set to 'Always', which means that the pivot will always be used, irrespective of the rotation. == Notes == * The 'Pivot When' setting has been integrated in the constraint, since this is something that will often be required for these setups. Having to set up additional drivers to drive the constraint to do this kindof beats the purpose of providing this. * The 'Offset' functionality is probably not presented as clearly as it could be. We may need to go over this again. * For foot-roll - if any scaling of the foot is required, simply set up a driver on the y-offset to make this dynamically respond to the "scale" RNA property of the bones (don't use the "Transform Channel" vartype since that won't work correct here). However, this shouldn't be common enough to warrant special treatment.
2010-05-27 10:50:06 +00:00
/* perform the pivoting... */
/* 1. take the vector from owner to the pivot */
sub_v3_v3v3(vec, cob->matrix[3], pivot);
== Pivot Constraint == This constraint allows an object or bone to have their rotations applied as if their origin/pivot-point was located elsewhere. The most obvious uses include foot-roll, see-saws, but could also include more complicated rolling-box examples. == Usage Examples == === Foot Roll === 1. Add 'Pivot' Constraint to the bone without any target. 2. Set the 'Y' value of the offset to the length of the bone. Usually this should be negative (if you rig with feet facing 'forwards' along -Y axis). This gives you a pivot point relative to the bone's (preconstraint) location, which should be at the tip of the bone here. Disabling the 'Use Relative Offset' would make this offset be relative to 0,0,0 instead of to the owner/bone-head. 3. Ensure that the 'Pivot When' setting is set to '-X Rot', (default) which means that the pivot will only used when the rotation on the X-Axis is negative to get tip-toe 'roll'. === See Saw === 1. Add a 'Pivot' constraint too see-saw plank object, this time with a target that you wish to have as the pivot-point. It's possible to do this without too (as before), but is less intuitive. 2. Optionally, if you want the plank slightly raised, set the z-offset value, which should make the pivot-point used to be relative to the target with the z-offset applied. 3. Ensure that 'Pivot When' is set to 'Always', which means that the pivot will always be used, irrespective of the rotation. == Notes == * The 'Pivot When' setting has been integrated in the constraint, since this is something that will often be required for these setups. Having to set up additional drivers to drive the constraint to do this kindof beats the purpose of providing this. * The 'Offset' functionality is probably not presented as clearly as it could be. We may need to go over this again. * For foot-roll - if any scaling of the foot is required, simply set up a driver on the y-offset to make this dynamically respond to the "scale" RNA property of the bones (don't use the "Transform Channel" vartype since that won't work correct here). However, this shouldn't be common enough to warrant special treatment.
2010-05-27 10:50:06 +00:00
/* 2. rotate this vector by the rotation of the object... */
mul_m3_v3(rotMat, vec);
/* 3. make the rotation in terms of the pivot now */
add_v3_v3v3(cob->matrix[3], pivot, vec);
}
static bConstraintTypeInfo CTI_PIVOT = {
CONSTRAINT_TYPE_PIVOT, /* type */
sizeof(bPivotConstraint), /* size */
"Pivot", /* name */
"bPivotConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
pivotcon_id_looper, /* id looper */
NULL, /* copy data */
2012-03-01 12:20:18 +00:00
NULL, /* new data */ // XXX: might be needed to get 'normal' pivot behavior...
== Pivot Constraint == This constraint allows an object or bone to have their rotations applied as if their origin/pivot-point was located elsewhere. The most obvious uses include foot-roll, see-saws, but could also include more complicated rolling-box examples. == Usage Examples == === Foot Roll === 1. Add 'Pivot' Constraint to the bone without any target. 2. Set the 'Y' value of the offset to the length of the bone. Usually this should be negative (if you rig with feet facing 'forwards' along -Y axis). This gives you a pivot point relative to the bone's (preconstraint) location, which should be at the tip of the bone here. Disabling the 'Use Relative Offset' would make this offset be relative to 0,0,0 instead of to the owner/bone-head. 3. Ensure that the 'Pivot When' setting is set to '-X Rot', (default) which means that the pivot will only used when the rotation on the X-Axis is negative to get tip-toe 'roll'. === See Saw === 1. Add a 'Pivot' constraint too see-saw plank object, this time with a target that you wish to have as the pivot-point. It's possible to do this without too (as before), but is less intuitive. 2. Optionally, if you want the plank slightly raised, set the z-offset value, which should make the pivot-point used to be relative to the target with the z-offset applied. 3. Ensure that 'Pivot When' is set to 'Always', which means that the pivot will always be used, irrespective of the rotation. == Notes == * The 'Pivot When' setting has been integrated in the constraint, since this is something that will often be required for these setups. Having to set up additional drivers to drive the constraint to do this kindof beats the purpose of providing this. * The 'Offset' functionality is probably not presented as clearly as it could be. We may need to go over this again. * For foot-roll - if any scaling of the foot is required, simply set up a driver on the y-offset to make this dynamically respond to the "scale" RNA property of the bones (don't use the "Transform Channel" vartype since that won't work correct here). However, this shouldn't be common enough to warrant special treatment.
2010-05-27 10:50:06 +00:00
pivotcon_get_tars, /* get constraint targets */
pivotcon_flush_tars, /* flush constraint targets */
default_get_tarmat, /* get target matrix */
pivotcon_evaluate /* evaluate */
};
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
/* ----------- Follow Track ------------- */
static void followtrack_new_data(void *cdata)
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
{
bFollowTrackConstraint *data = (bFollowTrackConstraint *)cdata;
data->clip = NULL;
data->flag |= FOLLOWTRACK_ACTIVECLIP;
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
}
static void followtrack_id_looper(bConstraint *con, ConstraintIDFunc func, void *userdata)
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
{
bFollowTrackConstraint *data = con->data;
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
func(con, (ID**)&data->clip, userdata);
func(con, (ID**)&data->camera, userdata);
func(con, (ID**)&data->depth_ob, userdata);
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
}
static void followtrack_evaluate(bConstraint *con, bConstraintOb *cob, ListBase *UNUSED(targets))
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
{
Scene *scene = cob->scene;
bFollowTrackConstraint *data = con->data;
MovieClip *clip = data->clip;
MovieTracking *tracking;
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
MovieTrackingTrack *track;
MovieTrackingObject *tracking_object;
Object *camob= data->camera ? data->camera : scene->camera;
if (data->flag & FOLLOWTRACK_ACTIVECLIP)
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
clip= scene->clip;
if (!clip || !data->track[0] || !camob)
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
return;
tracking = &clip->tracking;
if (data->object[0])
tracking_object = BKE_tracking_named_object(tracking, data->object);
else
tracking_object = BKE_tracking_get_camera_object(tracking);
if (!tracking_object)
return;
track = BKE_tracking_named_track(tracking, tracking_object, data->track);
if (!track)
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
return;
if (data->flag & FOLLOWTRACK_USE_3D_POSITION) {
if (track->flag & TRACK_HAS_BUNDLE) {
float obmat[4][4], mat[4][4];
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
copy_m4_m4(obmat, cob->matrix);
if ((tracking_object->flag & TRACKING_OBJECT_CAMERA)==0) {
float imat[4][4];
copy_m4_m4(mat, camob->obmat);
BKE_tracking_get_interpolated_camera(tracking, tracking_object, scene->r.cfra, imat);
invert_m4(imat);
mul_serie_m4(cob->matrix, obmat, mat, imat, NULL, NULL, NULL, NULL, NULL);
translate_m4(cob->matrix, track->bundle_pos[0], track->bundle_pos[1], track->bundle_pos[2]);
}
else {
BKE_get_tracking_mat(cob->scene, camob, mat);
mult_m4_m4m4(cob->matrix, obmat, mat);
translate_m4(cob->matrix, track->bundle_pos[0], track->bundle_pos[1], track->bundle_pos[2]);
}
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
}
}
else {
MovieTrackingMarker *marker;
float vec[3], disp[3], axis[3], mat[4][4];
float aspect= (scene->r.xsch * scene->r.xasp) / (scene->r.ysch * scene->r.yasp);
float len, d;
BKE_object_where_is_calc_mat4(scene, camob, mat);
/* camera axis */
vec[0] = 0.0f;
vec[1] = 0.0f;
vec[2] = 1.0f;
mul_v3_m4v3(axis, mat, vec);
/* distance to projection plane */
copy_v3_v3(vec, cob->matrix[3]);
sub_v3_v3(vec, mat[3]);
project_v3_v3v3(disp, vec, axis);
len = len_v3(disp);
if (len > FLT_EPSILON) {
CameraParams params;
float pos[2], rmat[4][4];
marker = BKE_tracking_get_marker(track, scene->r.cfra);
add_v2_v2v2(pos, marker->pos, track->offset);
2012-05-05 00:58:22 +00:00
BKE_camera_params_init(&params);
BKE_camera_params_from_object(&params, camob);
if (params.is_ortho) {
vec[0] = params.ortho_scale * (pos[0] - 0.5f + params.shiftx);
vec[1] = params.ortho_scale * (pos[1] - 0.5f + params.shifty);
vec[2] = -len;
if (aspect > 1.0f)
vec[1] /= aspect;
else
vec[0] *= aspect;
mul_v3_m4v3(disp, camob->obmat, vec);
copy_m4_m4(rmat, camob->obmat);
zero_v3(rmat[3]);
mult_m4_m4m4(cob->matrix, cob->matrix, rmat);
copy_v3_v3(cob->matrix[3], disp);
}
else {
d= (len * params.sensor_x) / (2.0f * params.lens);
vec[0] = d * (2.0f * (pos[0] + params.shiftx) - 1.0f);
vec[1] = d * (2.0f * (pos[1] + params.shifty) - 1.0f);
vec[2] = -len;
if (aspect > 1.0f)
vec[1] /= aspect;
else
vec[0] *= aspect;
mul_v3_m4v3(disp, camob->obmat, vec);
/* apply camera rotation so Z-axis would be co-linear */
copy_m4_m4(rmat, camob->obmat);
zero_v3(rmat[3]);
mult_m4_m4m4(cob->matrix, cob->matrix, rmat);
copy_v3_v3(cob->matrix[3], disp);
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
}
if (data->depth_ob && data->depth_ob->derivedFinal) {
Object *depth_ob = data->depth_ob;
BVHTreeFromMesh treeData = NULL_BVHTreeFromMesh;
BVHTreeRayHit hit;
float ray_start[3], ray_end[3], ray_nor[3], imat[4][4];
int result;
invert_m4_m4(imat, depth_ob->obmat);
mul_v3_m4v3(ray_start, imat, camob->obmat[3]);
mul_v3_m4v3(ray_end, imat, cob->matrix[3]);
sub_v3_v3v3(ray_nor, ray_end, ray_start);
bvhtree_from_mesh_faces(&treeData, depth_ob->derivedFinal, 0.0f, 4, 6);
hit.dist = FLT_MAX;
hit.index = -1;
result = BLI_bvhtree_ray_cast(treeData.tree, ray_start, ray_nor, 0.0f, &hit, treeData.raycast_callback, &treeData);
if (result != -1) {
mul_v3_m4v3(cob->matrix[3], depth_ob->obmat, hit.co);
}
free_bvhtree_from_mesh(&treeData);
}
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
}
}
}
static bConstraintTypeInfo CTI_FOLLOWTRACK = {
CONSTRAINT_TYPE_FOLLOWTRACK, /* type */
sizeof(bFollowTrackConstraint), /* size */
"Follow Track", /* name */
"bFollowTrackConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
followtrack_id_looper, /* id looper */
NULL, /* copy data */
followtrack_new_data, /* new data */
NULL, /* get constraint targets */
NULL, /* flush constraint targets */
NULL, /* get target matrix */
followtrack_evaluate /* evaluate */
};
/* ----------- Camre Solver ------------- */
static void camerasolver_new_data(void *cdata)
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
{
bCameraSolverConstraint *data = (bCameraSolverConstraint *)cdata;
data->clip = NULL;
data->flag |= CAMERASOLVER_ACTIVECLIP;
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
}
static void camerasolver_id_looper(bConstraint *con, ConstraintIDFunc func, void *userdata)
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
{
bCameraSolverConstraint *data = con->data;
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
func(con, (ID**)&data->clip, userdata);
}
static void camerasolver_evaluate(bConstraint *con, bConstraintOb *cob, ListBase *UNUSED(targets))
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
{
Scene *scene = cob->scene;
bCameraSolverConstraint *data = con->data;
MovieClip *clip = data->clip;
if (data->flag & CAMERASOLVER_ACTIVECLIP)
clip = scene->clip;
if (clip) {
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
float mat[4][4], obmat[4][4];
MovieTracking *tracking = &clip->tracking;
MovieTrackingObject *object = BKE_tracking_get_camera_object(tracking);
BKE_tracking_get_interpolated_camera(tracking, object, scene->r.cfra, mat);
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
copy_m4_m4(obmat, cob->matrix);
mult_m4_m4m4(cob->matrix, obmat, mat);
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
}
}
static bConstraintTypeInfo CTI_CAMERASOLVER = {
CONSTRAINT_TYPE_CAMERASOLVER, /* type */
sizeof(bCameraSolverConstraint), /* size */
"Camera Solver", /* name */
"bCameraSolverConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
camerasolver_id_looper, /* id looper */
NULL, /* copy data */
camerasolver_new_data, /* new data */
NULL, /* get constraint targets */
NULL, /* flush constraint targets */
NULL, /* get target matrix */
camerasolver_evaluate /* evaluate */
};
/* ----------- Object Solver ------------- */
static void objectsolver_new_data(void *cdata)
{
bObjectSolverConstraint *data = (bObjectSolverConstraint *)cdata;
data->clip = NULL;
data->flag |= OBJECTSOLVER_ACTIVECLIP;
unit_m4(data->invmat);
}
static void objectsolver_id_looper(bConstraint *con, ConstraintIDFunc func, void *userdata)
{
bObjectSolverConstraint *data= con->data;
func(con, (ID**)&data->clip, userdata);
func(con, (ID**)&data->camera, userdata);
}
static void objectsolver_evaluate(bConstraint *con, bConstraintOb *cob, ListBase *UNUSED(targets))
{
Scene *scene = cob->scene;
bObjectSolverConstraint *data = con->data;
MovieClip *clip = data->clip;
Object *camob = data->camera ? data->camera : scene->camera;
if (data->flag & OBJECTSOLVER_ACTIVECLIP)
clip= scene->clip;
if (!camob || !clip)
return;
if (clip) {
MovieTracking *tracking = &clip->tracking;
MovieTrackingObject *object;
object = BKE_tracking_named_object(tracking, data->object);
if (object) {
float mat[4][4], obmat[4][4], imat[4][4], cammat[4][4], camimat[4][4], parmat[4][4];
BKE_object_where_is_calc_mat4(scene, camob, cammat);
BKE_tracking_get_interpolated_camera(tracking, object, scene->r.cfra, mat);
invert_m4_m4(camimat, cammat);
mult_m4_m4m4(parmat, cammat, data->invmat);
copy_m4_m4(cammat, camob->obmat);
copy_m4_m4(obmat, cob->matrix);
invert_m4_m4(imat, mat);
mul_serie_m4(cob->matrix, cammat, imat, camimat, parmat, obmat, NULL, NULL, NULL);
}
}
}
static bConstraintTypeInfo CTI_OBJECTSOLVER = {
CONSTRAINT_TYPE_OBJECTSOLVER, /* type */
sizeof(bObjectSolverConstraint), /* size */
"Object Solver", /* name */
"bObjectSolverConstraint", /* struct name */
NULL, /* free data */
NULL, /* relink data */
objectsolver_id_looper, /* id looper */
NULL, /* copy data */
objectsolver_new_data, /* new data */
NULL, /* get constraint targets */
NULL, /* flush constraint targets */
NULL, /* get target matrix */
objectsolver_evaluate /* evaluate */
};
/* ************************* Constraints Type-Info *************************** */
/* All of the constraints api functions use bConstraintTypeInfo structs to carry out
* and operations that involve constraint specific code.
*/
/* These globals only ever get directly accessed in this file */
static bConstraintTypeInfo *constraintsTypeInfo[NUM_CONSTRAINT_TYPES];
static short CTI_INIT= 1; /* when non-zero, the list needs to be updated */
/* This function only gets called when CTI_INIT is non-zero */
static void constraints_init_typeinfo (void)
{
constraintsTypeInfo[0]= NULL; /* 'Null' Constraint */
constraintsTypeInfo[1]= &CTI_CHILDOF; /* ChildOf Constraint */
constraintsTypeInfo[2]= &CTI_TRACKTO; /* TrackTo Constraint */
constraintsTypeInfo[3]= &CTI_KINEMATIC; /* IK Constraint */
constraintsTypeInfo[4]= &CTI_FOLLOWPATH; /* Follow-Path Constraint */
constraintsTypeInfo[5]= &CTI_ROTLIMIT; /* Limit Rotation Constraint */
constraintsTypeInfo[6]= &CTI_LOCLIMIT; /* Limit Location Constraint */
constraintsTypeInfo[7]= &CTI_SIZELIMIT; /* Limit Scaling Constraint */
constraintsTypeInfo[8]= &CTI_ROTLIKE; /* Copy Rotation Constraint */
constraintsTypeInfo[9]= &CTI_LOCLIKE; /* Copy Location Constraint */
constraintsTypeInfo[10]= &CTI_SIZELIKE; /* Copy Scaling Constraint */
constraintsTypeInfo[11]= &CTI_PYTHON; /* Python/Script Constraint */
constraintsTypeInfo[12]= &CTI_ACTION; /* Action Constraint */
constraintsTypeInfo[13]= &CTI_LOCKTRACK; /* Locked-Track Constraint */
constraintsTypeInfo[14]= &CTI_DISTLIMIT; /* Limit Distance Constraint */
constraintsTypeInfo[15]= &CTI_STRETCHTO; /* StretchTo Constaint */
constraintsTypeInfo[16]= &CTI_MINMAX; /* Floor Constraint */
constraintsTypeInfo[17]= &CTI_RIGIDBODYJOINT; /* RigidBody Constraint */
constraintsTypeInfo[18]= &CTI_CLAMPTO; /* ClampTo Constraint */
constraintsTypeInfo[19]= &CTI_TRANSFORM; /* Transformation Constraint */
constraintsTypeInfo[20]= &CTI_SHRINKWRAP; /* Shrinkwrap Constraint */
constraintsTypeInfo[21]= &CTI_DAMPTRACK; /* Damped TrackTo Constraint */
constraintsTypeInfo[22]= &CTI_SPLINEIK; /* Spline IK Constraint */
constraintsTypeInfo[23]= &CTI_TRANSLIKE; /* Copy Transforms Constraint */
constraintsTypeInfo[24]= &CTI_SAMEVOL; /* Maintain Volume Constraint */
== Pivot Constraint == This constraint allows an object or bone to have their rotations applied as if their origin/pivot-point was located elsewhere. The most obvious uses include foot-roll, see-saws, but could also include more complicated rolling-box examples. == Usage Examples == === Foot Roll === 1. Add 'Pivot' Constraint to the bone without any target. 2. Set the 'Y' value of the offset to the length of the bone. Usually this should be negative (if you rig with feet facing 'forwards' along -Y axis). This gives you a pivot point relative to the bone's (preconstraint) location, which should be at the tip of the bone here. Disabling the 'Use Relative Offset' would make this offset be relative to 0,0,0 instead of to the owner/bone-head. 3. Ensure that the 'Pivot When' setting is set to '-X Rot', (default) which means that the pivot will only used when the rotation on the X-Axis is negative to get tip-toe 'roll'. === See Saw === 1. Add a 'Pivot' constraint too see-saw plank object, this time with a target that you wish to have as the pivot-point. It's possible to do this without too (as before), but is less intuitive. 2. Optionally, if you want the plank slightly raised, set the z-offset value, which should make the pivot-point used to be relative to the target with the z-offset applied. 3. Ensure that 'Pivot When' is set to 'Always', which means that the pivot will always be used, irrespective of the rotation. == Notes == * The 'Pivot When' setting has been integrated in the constraint, since this is something that will often be required for these setups. Having to set up additional drivers to drive the constraint to do this kindof beats the purpose of providing this. * The 'Offset' functionality is probably not presented as clearly as it could be. We may need to go over this again. * For foot-roll - if any scaling of the foot is required, simply set up a driver on the y-offset to make this dynamically respond to the "scale" RNA property of the bones (don't use the "Transform Channel" vartype since that won't work correct here). However, this shouldn't be common enough to warrant special treatment.
2010-05-27 10:50:06 +00:00
constraintsTypeInfo[25]= &CTI_PIVOT; /* Pivot Constraint */
Camera tracking integration =========================== Commiting camera tracking integration gsoc project into trunk. This commit includes: - Bundled version of libmv library (with some changes against official repo, re-sync with libmv repo a bit later) - New datatype ID called MovieClip which is optimized to work with movie clips (both of movie files and image sequences) and doing camera/motion tracking operations. - New editor called Clip Editor which is currently used for motion/tracking stuff only, but which can be easily extended to work with masks too. This editor supports: * Loading movie files/image sequences * Build proxies with different size for loaded movie clip, also supports building undistorted proxies to increase speed of playback in undistorted mode. * Manual lens distortion mode calibration using grid and grease pencil * Supervised 2D tracking using two different algorithms KLT and SAD. * Basic algorithm for feature detection * Camera motion solving. scene orientation - New constraints to "link" scene objects with solved motions from clip: * Follow Track (make object follow 2D motion of track with given name or parent object to reconstructed 3D position of track) * Camera Solver to make camera moving in the same way as reconstructed camera This commit NOT includes changes from tomato branch: - New nodes (they'll be commited as separated patch) - Automatic image offset guessing for image input node and image editor (need to do more tests and gather more feedback) - Code cleanup in libmv-capi. It's not so critical cleanup, just increasing readability and understanadability of code. Better to make this chaneg when Keir will finish his current patch. More details about this project can be found on this page: http://wiki.blender.org/index.php/User:Nazg-gul/GSoC-2011 Further development of small features would be done in trunk, bigger/experimental features would first be implemented in tomato branch.
2011-11-07 12:55:18 +00:00
constraintsTypeInfo[26]= &CTI_FOLLOWTRACK; /* Follow Track Constraint */
constraintsTypeInfo[27]= &CTI_CAMERASOLVER; /* Camera Solver Constraint */
constraintsTypeInfo[28]= &CTI_OBJECTSOLVER; /* Object Solver Constraint */
}
/* This function should be used for getting the appropriate type-info when only
* a constraint type is known
*/
bConstraintTypeInfo *get_constraint_typeinfo (int type)
{
/* initialize the type-info list? */
if (CTI_INIT) {
constraints_init_typeinfo();
CTI_INIT = 0;
}
/* only return for valid types */
if ( (type >= CONSTRAINT_TYPE_NULL) &&
(type < NUM_CONSTRAINT_TYPES ) )
{
/* there shouldn't be any segfaults here... */
return constraintsTypeInfo[type];
}
else {
printf("No valid constraint type-info data available. Type = %i\n", type);
}
return NULL;
}
/* This function should always be used to get the appropriate type-info, as it
* has checks which prevent segfaults in some weird cases.
*/
bConstraintTypeInfo *constraint_get_typeinfo (bConstraint *con)
{
/* only return typeinfo for valid constraints */
if (con)
return get_constraint_typeinfo(con->type);
else
return NULL;
}
/* ************************* General Constraints API ************************** */
/* The functions here are called by various parts of Blender. Very few (should be none if possible)
* constraint-specific code should occur here.
*/
/* ---------- Data Management ------- */
/* Free data of a specific constraint if it has any info.
* be sure to run BIK_clear_data() when freeing an IK constraint,
* unless DAG_scene_sort is called. */
void free_constraint_data(bConstraint *con)
{
if (con->data) {
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
/* perform any special freeing constraint may have */
if (cti && cti->free_data)
cti->free_data(con);
/* free constraint data now */
MEM_freeN(con->data);
}
}
/* Free all constraints from a constraint-stack */
void free_constraints(ListBase *list)
{
bConstraint *con;
/* Free constraint data and also any extra data */
for (con= list->first; con; con= con->next)
free_constraint_data(con);
/* Free the whole list */
BLI_freelistN(list);
}
/* Remove the specified constraint from the given constraint stack */
int remove_constraint(ListBase *list, bConstraint *con)
{
if (con) {
free_constraint_data(con);
BLI_freelinkN(list, con);
return 1;
}
else
return 0;
}
/* Remove all the constraints of the specified type from the given constraint stack */
void remove_constraints_type(ListBase *list, short type, short last_only)
{
bConstraint *con, *conp;
if (list == NULL)
return;
/* remove from the end of the list to make it faster to find the last instance */
for (con= list->last; con; con= conp) {
conp= con->prev;
if (con->type == type) {
remove_constraint(list, con);
if (last_only)
return;
}
}
}
/* ......... */
/* Creates a new constraint, initializes its data, and returns it */
static bConstraint *add_new_constraint_internal (const char *name, short type)
{
bConstraint *con= MEM_callocN(sizeof(bConstraint), "Constraint");
bConstraintTypeInfo *cti= get_constraint_typeinfo(type);
const char *newName;
/* Set up a generic constraint datablock */
con->type = type;
con->flag |= CONSTRAINT_EXPAND;
con->enforce = 1.0f;
/* Determine a basic name, and info */
if (cti) {
/* initialize constraint data */
con->data = MEM_callocN(cti->size, cti->structName);
/* only constraints that change any settings need this */
if (cti->new_data)
cti->new_data(con->data);
/* if no name is provided, use the type of the constraint as the name */
newName= (name && name[0]) ? name : cti->name;
}
else {
/* if no name is provided, use the generic "Const" name */
// NOTE: any constraint type that gets here really shouldn't get added...
newName= (name && name[0]) ? name : "Const";
}
/* copy the name */
BLI_strncpy(con->name, newName, sizeof(con->name));
/* return the new constraint */
return con;
}
/* if pchan is not NULL then assume we're adding a pose constraint */
static bConstraint *add_new_constraint (Object *ob, bPoseChannel *pchan, const char *name, short type)
{
bConstraint *con;
ListBase *list;
/* add the constraint */
con= add_new_constraint_internal(name, type);
/* find the constraint stack - bone or object? */
list = (pchan) ? (&pchan->constraints) : (&ob->constraints);
if (list) {
/* add new constraint to end of list of constraints before ensuring that it has a unique name
* (otherwise unique-naming code will fail, since it assumes element exists in list)
*/
BLI_addtail(list, con);
unique_constraint_name(con, list);
/* if the target list is a list on some PoseChannel belonging to a proxy-protected
* Armature layer, we must tag newly added constraints with a flag which allows them
* to persist after proxy syncing has been done
*/
if (proxylocked_constraints_owner(ob, pchan))
con->flag |= CONSTRAINT_PROXY_LOCAL;
/* make this constraint the active one */
constraints_set_active(list, con);
}
/* set type+owner specific immutable settings */
// TODO: does action constraint need anything here - i.e. spaceonce?
switch (type) {
case CONSTRAINT_TYPE_CHILDOF:
{
/* if this constraint is being added to a posechannel, make sure
* the constraint gets evaluated in pose-space */
if (pchan) {
con->ownspace = CONSTRAINT_SPACE_POSE;
con->flag |= CONSTRAINT_SPACEONCE;
}
}
break;
}
return con;
}
/* ......... */
/* Add new constraint for the given bone */
bConstraint *add_pose_constraint (Object *ob, bPoseChannel *pchan, const char *name, short type)
{
if (pchan == NULL)
return NULL;
return add_new_constraint(ob, pchan, name, type);
}
/* Add new constraint for the given object */
bConstraint *add_ob_constraint(Object *ob, const char *name, short type)
{
return add_new_constraint(ob, NULL, name, type);
}
/* ......... */
/* Reassign links that constraints have to other data (called during file loading?) */
void relink_constraints(ListBase *conlist)
{
bConstraint *con;
bConstraintTarget *ct;
for (con= conlist->first; con; con= con->next) {
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
if (cti) {
/* relink any targets */
if (cti->get_constraint_targets) {
ListBase targets = {NULL, NULL};
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
cti->get_constraint_targets(con, &targets);
for (ct= targets.first; ct; ct= ct->next) {
ID_NEW(ct->tar);
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
if (cti->flush_constraint_targets)
cti->flush_constraint_targets(con, &targets, 0);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
/* relink any other special data */
if (cti->relink_data)
cti->relink_data(con);
}
}
}
/* Run the given callback on all ID-blocks in list of constraints */
void id_loop_constraints(ListBase *conlist, ConstraintIDFunc func, void *userdata)
{
bConstraint *con;
for (con= conlist->first; con; con= con->next) {
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
if (cti) {
if (cti->id_looper)
cti->id_looper(con, func, userdata);
}
}
}
/* ......... */
/* helper for copy_constraints(), to be used for making sure that ID's are valid */
static void con_extern_cb(bConstraint *UNUSED(con), ID **idpoin, void *UNUSED(userData))
{
if (*idpoin && (*idpoin)->lib)
id_lib_extern(*idpoin);
}
/* duplicate all of the constraints in a constraint stack */
void copy_constraints(ListBase *dst, const ListBase *src, int do_extern)
{
bConstraint *con, *srccon;
dst->first= dst->last= NULL;
BLI_duplicatelist(dst, src);
for (con=dst->first, srccon=src->first; con && srccon; srccon=srccon->next, con=con->next) {
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
/* make a new copy of the constraint's data */
con->data = MEM_dupallocN(con->data);
/* only do specific constraints if required */
if (cti) {
/* perform custom copying operations if needed */
if (cti->copy_data)
cti->copy_data(con, srccon);
2012-03-18 07:38:51 +00:00
/* for proxies we don't want to make extern */
if (do_extern) {
/* go over used ID-links for this constraint to ensure that they are valid for proxies */
if (cti->id_looper)
cti->id_looper(con, con_extern_cb, NULL);
}
}
}
}
/* ......... */
bConstraint *constraints_findByName(ListBase *list, const char *name)
{
return BLI_findstring(list, name, offsetof(bConstraint, name));
}
/* finds the 'active' constraint in a constraint stack */
bConstraint *constraints_get_active (ListBase *list)
{
bConstraint *con;
/* search for the first constraint with the 'active' flag set */
if (list) {
for (con= list->first; con; con= con->next) {
if (con->flag & CONSTRAINT_ACTIVE)
return con;
}
}
/* no active constraint found */
return NULL;
}
/* Set the given constraint as the active one (clearing all the others) */
void constraints_set_active(ListBase *list, bConstraint *con)
{
bConstraint *c;
if (list) {
for (c= list->first; c; c= c->next) {
if (c == con)
c->flag |= CONSTRAINT_ACTIVE;
else
c->flag &= ~CONSTRAINT_ACTIVE;
}
}
}
/* -------- Constraints and Proxies ------- */
/* Rescue all constraints tagged as being CONSTRAINT_PROXY_LOCAL (i.e. added to bone that's proxy-synced in this file) */
void extract_proxylocal_constraints(ListBase *dst, ListBase *src)
{
bConstraint *con, *next;
/* for each tagged constraint, remove from src and move to dst */
for (con= src->first; con; con= next) {
next= con->next;
/* check if tagged */
if (con->flag & CONSTRAINT_PROXY_LOCAL) {
BLI_remlink(src, con);
BLI_addtail(dst, con);
}
}
}
/* Returns if the owner of the constraint is proxy-protected */
short proxylocked_constraints_owner(Object *ob, bPoseChannel *pchan)
{
/* Currently, constraints can only be on object or bone level */
if (ob && ob->proxy) {
if (ob->pose && pchan) {
bArmature *arm= ob->data;
/* On bone-level, check if bone is on proxy-protected layer */
if ((pchan->bone) && (pchan->bone->layer & arm->layer_protected))
return 1;
}
else {
/* FIXME: constraints on object-level are not handled well yet */
return 1;
}
}
return 0;
}
/* -------- Target-Matrix Stuff ------- */
/* This function is a relic from the prior implementations of the constraints system, when all
* constraints either had one or no targets. It used to be called during the main constraint solving
* loop, but is now only used for the remaining cases for a few constraints.
*
* None of the actual calculations of the matrices should be done here! Also, this function is
* not to be used by any new constraints, particularly any that have multiple targets.
*/
void get_constraint_target_matrix(struct Scene *scene, bConstraint *con, int n, short ownertype, void *ownerdata, float mat[][4], float ctime)
{
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
ListBase targets = {NULL, NULL};
bConstraintOb *cob;
bConstraintTarget *ct;
if (cti && cti->get_constraint_targets) {
/* make 'constraint-ob' */
cob= MEM_callocN(sizeof(bConstraintOb), "tempConstraintOb");
cob->type= ownertype;
cob->scene = scene;
switch (ownertype) {
case CONSTRAINT_OBTYPE_OBJECT: /* it is usually this case */
{
cob->ob= (Object *)ownerdata;
cob->pchan= NULL;
if (cob->ob) {
copy_m4_m4(cob->matrix, cob->ob->obmat);
copy_m4_m4(cob->startmat, cob->matrix);
}
else {
unit_m4(cob->matrix);
unit_m4(cob->startmat);
}
}
break;
case CONSTRAINT_OBTYPE_BONE: /* this may occur in some cases */
{
cob->ob= NULL; /* this might not work at all :/ */
cob->pchan= (bPoseChannel *)ownerdata;
if (cob->pchan) {
copy_m4_m4(cob->matrix, cob->pchan->pose_mat);
copy_m4_m4(cob->startmat, cob->matrix);
}
else {
unit_m4(cob->matrix);
unit_m4(cob->startmat);
}
}
break;
}
/* get targets - we only need the first one though (and there should only be one) */
cti->get_constraint_targets(con, &targets);
/* only calculate the target matrix on the first target */
ct= (bConstraintTarget *)targets.first;
while (ct && n-- > 0)
ct= ct->next;
if (ct) {
if (cti->get_target_matrix)
cti->get_target_matrix(con, cob, ct, ctime);
copy_m4_m4(mat, ct->matrix);
}
/* free targets + 'constraint-ob' */
if (cti->flush_constraint_targets)
cti->flush_constraint_targets(con, &targets, 1);
MEM_freeN(cob);
}
else {
/* invalid constraint - perhaps... */
unit_m4(mat);
2002-10-12 11:37:38 +00:00
}
}
/* Get the list of targets required for solving a constraint */
void get_constraint_targets_for_solving(bConstraint *con, bConstraintOb *cob, ListBase *targets, float ctime)
{
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
if (cti && cti->get_constraint_targets) {
bConstraintTarget *ct;
/* get targets
* - constraints should use ct->matrix, not directly accessing values
* - ct->matrix members have not yet been calculated here!
*/
cti->get_constraint_targets(con, targets);
/* set matrices
* - calculate if possible, otherwise just initialize as identity matrix
*/
if (cti->get_target_matrix) {
for (ct= targets->first; ct; ct= ct->next)
cti->get_target_matrix(con, cob, ct, ctime);
}
else {
for (ct= targets->first; ct; ct= ct->next)
unit_m4(ct->matrix);
}
}
}
/* ---------- Evaluation ----------- */
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* This function is called whenever constraints need to be evaluated. Currently, all
* constraints that can be evaluated are everytime this gets run.
*
* constraints_make_evalob and constraints_clear_evalob should be called before and
* after running this function, to sort out cob
*/
void solve_constraints(ListBase *conlist, bConstraintOb *cob, float ctime)
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
{
bConstraint *con;
float oldmat[4][4];
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
float enf;
/* check that there is a valid constraint object to evaluate */
if (cob == NULL)
return;
/* loop over available constraints, solving and blending them */
for (con= conlist->first; con; con= con->next) {
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
ListBase targets = {NULL, NULL};
/* these we can skip completely (invalid constraints...) */
if (cti == NULL) continue;
if (con->flag & (CONSTRAINT_DISABLE|CONSTRAINT_OFF)) continue;
/* these constraints can't be evaluated anyway */
if (cti->evaluate_constraint == NULL) continue;
/* influence == 0 should be ignored */
if (con->enforce == 0.0f) continue;
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* influence of constraint
2.5: Blender "Animato" - New Animation System Finally, here is the basic (functional) prototype of the new animation system which will allow for the infamous "everything is animatable", and which also addresses several of the more serious shortcomings of the old system. Unfortunately, this will break old animation files (especially right now, as I haven't written the version patching code yet), however, this is for the future. Highlights of the new system: * Scrapped IPO-Curves/IPO/(Action+Constraint-Channels)/Action system, and replaced it with F-Curve/Action. - F-Curves (animators from other packages will feel at home with this name) replace IPO-Curves. - The 'new' Actions, act as the containers for F-Curves, so that they can be reused. They are therefore more akin to the old 'IPO' blocks, except they do not have the blocktype restriction, so you can store materials/texture/geometry F-Curves in the same Action as Object transforms, etc. * F-Curves use RNA-paths for Data Access, hence allowing "every" (where sensible/editable that is) user-accessible setting from RNA to be animated. * Drivers are no longer mixed with Animation Data, so rigs will not be that easily broken and several dependency problems can be eliminated. (NOTE: drivers haven't been hooked up yet, but the code is in place) * F-Curve modifier system allows useful 'large-scale' manipulation of F-Curve values, including (I've only included implemented ones here): envelope deform (similar to lattices to allow broad-scale reshaping of curves), curve generator (polynomial or py-expression), cycles (replacing the old cyclic extrapolation modes, giving more control over this). (NOTE: currently this cannot be tested, as there's not access to them, but the code is all in place) * NLA system with 'tracks' (i.e. layers), and multiple strips per track. (NOTE: NLA system is not yet functional, as it's only partially coded still) There are more nice things that I will be preparing some nice docs for soon, but for now, check for more details: http://lists.blender.org/pipermail/bf-taskforce25/2009-January/000260.html So, what currently works: * I've implemented two basic operators for the 3D-view only to Insert and Delete Keyframes. These are tempolary ones only that will be replaced in due course with 'proper' code. * Object Loc/Rot/Scale can be keyframed. Also, the colour of the 'active' material (Note: this should really be for nth material instead, but that doesn't work yet in RNA) can also be keyframed into the same datablock. * Standard animation refresh (i.e. animation resulting from NLA and Action evaluation) is now done completely separate from drivers before anything else is done after a frame change. Drivers are handled after this in a separate pass, as dictated by depsgraph flags, etc. Notes: * Drivers haven't been hooked up yet * Only objects and data directly linked to objects can be animated. * Depsgraph will need further tweaks. Currently, I've only made sure that it will update some things in the most basic cases (i.e. frame change). * Animation Editors are currently broken (in terms of editing stuff). This will be my next target (priority to get Dopesheet working first, then F-Curve editor - i.e. old IPO Editor) * I've had to put in large chunks of XXX sandboxing for old animation system code all around the place. This will be cleaned up in due course, as some places need special review. In particular, the particles and sequencer code have far too many manual calls to calculate + flush animation info, which is really bad (this is a 'please explain yourselves' call to Physics coders!).
2009-01-17 03:12:50 +00:00
* - value should have been set from animation data already
*/
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
enf = con->enforce;
/* make copy of worldspace matrix pre-constraint for use with blending later */
copy_m4_m4(oldmat, cob->matrix);
/* move owner matrix into right space */
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
constraint_mat_convertspace(cob->ob, cob->pchan, cob->matrix, CONSTRAINT_SPACE_WORLD, con->ownspace);
/* prepare targets for constraint solving */
get_constraint_targets_for_solving(con, cob, &targets, ctime);
/* Solve the constraint and put result in cob->matrix */
cti->evaluate_constraint(con, cob, &targets);
/* clear targets after use
* - this should free temp targets but no data should be copied back
* as constraints may have done some nasty things to it...
*/
if (cti->flush_constraint_targets) {
cti->flush_constraint_targets(con, &targets, 1);
}
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
/* move owner back into worldspace for next constraint/other business */
if ((con->flag & CONSTRAINT_SPACEONCE) == 0)
constraint_mat_convertspace(cob->ob, cob->pchan, cob->matrix, con->ownspace, CONSTRAINT_SPACE_WORLD);
/* Interpolate the enforcement, to blend result of constraint into final owner transform
* - all this happens in worldspace to prevent any weirdness creeping in ([#26014] and [#25725]),
* since some constraints may not convert the solution back to the input space before blending
* but all are guaranteed to end up in good "worldspace" result
*/
/* Note: all kind of stuff here before (caused trouble), much easier to just interpolate, or did I miss something? -jahka */
if (enf < 1.0f) {
float solution[4][4];
copy_m4_m4(solution, cob->matrix);
blend_m4_m4m4(cob->matrix, oldmat, solution, enf);
== Constraints System == After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented. === What's New === * The long-awaited ``ChildOf Constraint'': This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent. * Constraint/Target Evaluation Spaces: In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were. * Action Constraint - Loc/Rot/Size Inputs The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff. * Transform - No longer in Crazy Space (TM) Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking. === General Code Changes === * solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be. * evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here. * A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better. * Made the formatting of constraint code consistent * There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly. * LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions. * Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme. * Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so. === Future Work === * Geometry to act as targets for constraints. A space has been reserved for this already. * Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
}
}
}