Commit graph

1077 commits

Author SHA1 Message Date
Angus Antley
7bb353742a more tweaks to get the wrist action right 2019-02-12 07:40:18 -08:00
amantley
bb29e382f0 tweaking the hand pose correction on the pole vector 2019-02-11 18:05:29 -08:00
Angus Antley
f125e90449 worked on the swing twist decomp to get the angles of the wrist for elbow adjustments 2019-02-11 07:41:41 -08:00
amantley
822ec1c529 working on the wrist tweak, dampened the spine twist 2019-02-08 17:53:23 -08:00
amantley
9eceb1d0bd implemented the code for the heuristic elbows including code from the paper authors. to do: dampen the twist of the spine
caused by the hand azimuth and put in the constraints for the wrists on
the pole vector theta computation.
2019-02-07 17:46:18 -08:00
amantley
4d9d597b4f tweaked the weights for the arms and negated the theta for the left arm 2019-02-07 11:14:33 -08:00
amantley
e2c9058f0a first try at the new elbow code 2019-02-06 18:29:33 -08:00
amantley
9ad20b8129 adding the animcontext to the commit 2019-02-05 17:22:00 -08:00
amantley
07a4f49c58 adding the armik nodes into the json 2019-02-05 17:21:39 -08:00
amantley
3e553f015c added the arm ik files to the repo, this is for the shoulder and elbow extensions 2019-02-05 15:54:37 -08:00
amantley
af38a40b85 Merge remote-tracking branch 'origin/animSplineIK' into animSplineIK 2019-02-05 09:43:11 -08:00
amantley
031dd5639b fixed pole vector over writing with the old json 2019-02-04 14:46:52 -08:00
amantley
78d6e42fc8 made the ik enable/disable work for the old animIK node json 2019-02-04 13:42:19 -08:00
amantley
a04e6d55ff enabled/disable ik now works for the new animspline json 2019-02-04 13:31:36 -08:00
Angus Antley
a2ef7edf10 cleaned up some warnings 2019-01-31 20:43:13 -08:00
Angus Antley
dd99f93d1a fixed missplaced static const 2019-01-31 19:52:17 -08:00
amantley
25c5a2f41a fixed mac/linux build error 2019-01-31 17:23:56 -08:00
amantley
61b019d176 added new constructor for cubichermitespline that takes quat and vec3. this means we don't need computeSplineFromTipAndBase to be declared in multiple files 2019-01-31 14:13:51 -08:00
amantley
ffd374e7d4 whitespace 2019-01-31 10:05:08 -08:00
amantley
d547d5b854 changed the json reader to take an array not a string for the flex targets 2019-01-30 18:16:02 -08:00
amantley
d174fb1b5c removed print statements 2019-01-30 16:59:12 -08:00
amantley
5054b54626 removed armIK.cpp and armIK.h, they are for the next changes 2019-01-30 16:20:50 -08:00
Angus Antley
e36877a861 added the interp for the possibility of extra flex coeffs 2019-01-29 22:02:32 -08:00
amantley
fb0ad7768c removed clear map 2019-01-29 17:31:20 -08:00
amantley
e2a729b68b got the spline working in myskeleton model, need to clean up 2019-01-29 17:25:25 -08:00
Angus Antley
3994311583 starting the move of the initial head base spline to myskeletonmodel 2019-01-29 08:54:35 -08:00
amantley
1919cc3b1a disable mid joint when not valid. more work on this tomorrow 2019-01-28 17:49:46 -08:00
amantley
dffd41ecb0 chain ik debug draw works for spline now 2019-01-28 16:58:28 -08:00
amantley
f17cfbcbb1 more cleaning, need to fix debug draw ik chain 2019-01-28 15:25:12 -08:00
amantley
ffd3a24bf2 further cleaning, broke the arms 2019-01-28 13:53:30 -08:00
amantley
f8bfef6dbd cleaning up the _hipsIndex references, cleaning in general 2019-01-28 11:50:23 -08:00
Angus Antley
2679a3a30d changed the naming of the middle joint from secondary target to midJoint, also generalized the handling of the neck head rotation after the middle spline is updated 2019-01-27 16:30:13 -08:00
amantley
f2a7f37950 got rid of the lag in the spline code by setting the flex coeffs to 1.0 2019-01-25 17:11:07 -08:00
amantley
bcbd9323c0 Merge remote-tracking branch 'upstream/master' into animSplineIK 2019-01-25 14:42:24 -08:00
amantley
446d7b9514 added the flex coeff for the primary and secondary spline targets to the json 2019-01-25 14:41:18 -08:00
amantley
992820cd67 removed unnecessary hand update function for two bone IK 2019-01-25 11:34:18 -08:00
amantley
71df614989 put all the hand update code in one function that works for two bone IK
and legacy animInverseKinematics Ik
2019-01-25 11:28:51 -08:00
Shannon Romano
fd04996853
Merge pull request #14755 from hyperlogic/bug-fix/joint-mesh-name-collisions
Bug fix for avatars that have mesh/joint name duplicates
2019-01-25 09:59:44 -08:00
amantley
bc635306ea added interp from enabled to disabled and vice versa 2019-01-24 22:28:40 -08:00
amantley
2574e82184 cleaned up print statements 2019-01-24 19:47:27 -08:00
amantley
e7941f78d6 adding the shoulder rotation back in 2019-01-24 17:17:39 -08:00
amantley
37f92d2319 added code to read tip and base var info from the json 2019-01-24 17:02:30 -08:00
amantley
cdd03646c2 latest spline code with ik node removed started cleanup 2019-01-23 17:22:13 -08:00
John Conklin II
1768d91780
Merge pull request #14733 from hyperlogic/bug-fix/translation-precision
Avatar Protocol: Fix for precision issues with translation
2019-01-23 14:07:58 -08:00
amantley
dbd03e2eb4 got the spline arms and legs working without any inverse kinematic node in the json, using _withIKnode.json 2019-01-22 11:53:18 -08:00
Anthony Thibault
5146f3dfae Bug fix for avatars that have mesh/joint name duplicates
This change has two parts.  First we do a better job of identifying joints vs meshes in the FBXSerializer.
Second we use this information to better handle duplicate names in AnimSkeleton.
When a duplicate is detected, allow joints to override meshes.
This will ensure that when joints are looked up by name we will not return a mesh object with the same name by mistake.
2019-01-22 10:47:01 -08:00
amantley
8d88c627b0 nearly have the arms legs and back working with out the ik node 2019-01-21 17:41:08 -08:00
amantley
6680ca53ad added updateHands2 which deals with two bone hands animvars 2019-01-20 08:56:39 -08:00
amantley
3a2697fa8c working on the two bone ik for the hands 2019-01-18 17:55:19 -08:00
amantley
7115796fc4 version with spline from Spine2 to head 2019-01-18 16:47:55 -08:00