Okee-dokey.

Well, I've determined that the translation is additive/subtractive from the point's location.  A simple test revealed this - set the bulge to 1.0, rotate 36d (1/10th), use a point near the midpoint of the joint angles involved.   I exported an unbulged version of the figure and a bulged version.  These had their origins set to the testing body part's and then it was just a matter of getting the point's value (I cheated so that when rotated 36d, the translation would be exactly along the Y axis).  The expected value if from the body part origin would have been 100 or less (scaling of 1000 into check app).  But it was 111.  And the point was 20 from the origin.  Since the point isn't exactly at the midpoint, this discrepancy is insignificant and fits the expectation - it is from the point not the origin.

But I'm still vague about the algorithm employed for bulges.  Thought that I had it, but it appears to be different than expected.  With a sufficient bulge setting, moving the joint angles around almost acts like a lensing effect on the vertices.  It is apparent that the distance of the point from the joint angle planes imparts more or less weighting and that weighting follows a sinusoidal wave (trough to crest to through or 0 -1-0 curved).

C makes it easy to shoot yourself in the foot. C++ makes it harder, but when you do, you blow your whole leg off.

 -- Bjarne Stroustrup

Contact Me | Kuroyume's DevelopmentZone