Calypso Best Fit FAIL

[[Aa...]]

Recently, in working with an interrupted 3D Curve I ran into some issues with best fitting. The curve is effectively a gear spline, with the sharp corners left out. If I try to set Best Fit to only some degrees of freedom, the Best Fit doesn't do anything (the alignment comes out to all zeros). But when I set Best Fit to all degrees of freedom, it does something, but the result is obviously wrong. The clocking is clearly off. What it comes down to is that Calypso isn't solving all 6 degrees of freedom simultaneously or iteratively.

I could arrive at a much better result by forcing iteration. I created a best fit alignment from the curve. Then created a second copy of the curve that is based on that best fit alignment from the first copy and recalls the points from the first copy. And so forth for 4 iterations.

It seems to me that if iterative solution is necessary to accurately best fit 6 degrees of freedom, Calypso should automatically implement it.

[steps off his soapbox]...

[[Is...]]

I have the same problem with in general with all interrupted curves.. 😐
My solution is not nice, basically I generate separate curves and then I Play with the displacements of each curve in order to obtain a kind average..