kojot303 it's a while since I've played with IK. When I was doing so regularly, I adopted a technique of loading two identical figures, one with IK on, and the other with IK off. Make the non-IK figure appear in outline mode. Save poses for individual frames to the library from the IK figure, and then apply them to the non-IK figure. this will give you a direct indication of what the uncalculated rotations of the IK limbs are set to. Adjust the non-IK figure's rotations to make the limbs coincide with the IK figure's limbs again, and then apply the same rotations to the IK figure. Repeat the cycle to confirm that the IK pose preserves the correct joint rotations when applied to the non-IK figure. This will give you a pose file that works whether IK is turned on or not.

I know this is unnecessarily tedious, but it's the only solution I found, since the IK solver does not properly calculate all of the IK joint rotations to allow the pose to remain identical regardless of the IK state. This is understandable from the viewpoint that translations of the endpoint of an IK limb are not constrained sufficiently to prevent the limb being stretched beyond its non-IK limits. IK also pays no regard whatever to joint rotation limits, when translating the endpoint. This is especially frustrating when dealing with human knees and elbows which do not have six degrees of freedom. Forearms and shins do not normally (OK, people with looser joints do have a small range of movement) bend side-side. That movement is achieved by twisting the upper arm or thigh. IK frustratingly ignores that restriction.

[Answer originally posted on SM Poser Forum]

---

My ShareCG Stuff

Verbosity: Profusely promulgating Graham's number epics of complete and utter verbiage by the metric monkey barrel.