Action Constraint: Optimisation attempt
Following on from the methods added specially for PoseLib to only execute the F-Curves in a given Action Group, I've attempted to use this for evaluating Action Constraints on Pose Channels. This does rely on the F-Curves being in groups named according to name of the Pose Channel of interest, hence, we may need some way to be able to fine tune this later. Preliminary tests seem to be promising, with rigs with quite a few action constraints being slightly more responsive (subjective test though). Please test thoroughly.
This commit is contained in:
@@ -134,7 +134,7 @@ void update_pose_constraint_flags(struct bPose *pose);
|
||||
void framechange_poses_clear_unkeyed(void);
|
||||
|
||||
/* Used for the Action Constraint */
|
||||
void what_does_obaction(struct Scene *scene, struct Object *ob, struct Object *workob, struct bPose *pose, struct bAction *act, float cframe);
|
||||
void what_does_obaction(struct Scene *scene, struct Object *ob, struct Object *workob, struct bPose *pose, struct bAction *act, char groupname[], float cframe);
|
||||
|
||||
/* exported for game engine */
|
||||
void blend_poses(struct bPose *dst, struct bPose *src, float srcweight, short mode);
|
||||
|
||||
Reference in New Issue
Block a user