This one really caused me to scratch my head (and throw about 9 hours of production patterning away).
Draw a sphere.
Increase the isocurve display density to 9. This is fine and on a symmetrical object like a sphere it creates evenly spaced isocurves as you would expect.
Now move the surface seam to a new spot.
The surface isocurves are now randomly spaced. And of course the randomness is just a little too subtle to be seen by eye, so silly old me assumed everything was still evenly spaced.
Looks like it’s related to the “special case degree 2” of spheres.
Rebuilding to a high enough degree to keep the deviation to a minimum seems to shut the problem up.
It’d be nice if this could be fixed though (if mathematically possible) because it sure as hell caused me a bunch of grief.
On a sphere, is there any reason not to just rotate the the whole sphere? You can see the point arrangement may change pretty dramatically as well with SrfSeam, I don’t know that this is a bug exactly, it is how the command works, it has to add knots and spans to the surface.
No problem to rotate the whole sphere IF I know that’s what I need to do at the beginning of the design process.
It was just such a nasty and unexpected surprise to have that happen, that I thought I’d mention it in case someone else gets trapped by it. I’m certainly not likely to make that mistake again