Center OSNAP is unreliable

I’m finding that the center OSNAP is hit and miss. Sometimes it works, and sometimes it doesn’t. In the video below, this object is an extrusion created from a circle. I should be able to snap to the center. Interestingly, even if I extract the edge to create a curve it still won’t snap to the center.

http://screencast.com/t/NHLPsUpW

Thanks,

Dan

Yes, I have seen that as well. A simple box with an ArrayPolar’ed amount of holes from one extrusion and the center snap will snap to some of the holes but on to all of them.

Did you make a deformable circle? I can reproduce this here but not with a default non-deformable circle. I believe the deformable circle doesn’t have a true center snap… not sure if it can or should but I’ll file the question.

http://mcneel.myjetbrains.com/youtrack/issue/RH-28969

Hi Dan, all - check the setting in Options > Modeling Aids for

“Center snap to approximate arcs, circles and ellipses”

Does that make a difference?

-Pascal

Hi Brian,

Actually, I need to correct my original statement. I’m finding the issue with extrusions created from holes, which were created with the RoundHole command (not circles as I originally stated). Extruding the resultant circular edge is leading to this issue. Not always, but sometimes.

Pascal’s solution seems to be working.

Thanks,

Dan

Yes, much better. I will keep testing to see if I find other scenarios that still fail with this option checked.

Thanks,

Dan

Cool… I hadn’t noticed that in v6!!

Well, yes…
But, again. I make a short cylinder, draw a circle somewhere within its base, extrude that, cap, ArrayPolar 8 times around the center of the main cylinder, and boolean difference the 8 holes. Now I can center snap to all 8 circular edges in the base of the cylinder but only to 4 out of 8 circular edges in the top of the cylinder. I don’t understand why those 4 should be approximate arcs, circles, or ellipses.

Hi Wim -does it make a difference - so to speak - if the smaller cylinders all protrude fully through the larger, original one rather than come exactly to the same dimension?

-Pascal

:grin:
The larger one sits on the world CPlane and the circle (singular) was drawn on the CPlane as well. That circle was extruded well above the top face of the larger one before being ArrayPolar’ed to become several (and I would think identical). So, short answer, no.
w

I’m hoping this will work a little better in a near future WIP. The idea was to try to prevent Center osnap creep by making it by default not snap to approximate shapes. The problem is currently the tolerance passed to IsArc and IsEllipse test is super tight (about 2e-10) and that means even NURBS shapes that are in fact exact arcs, circles or ellipses are not always classified as such, especially if they are big.

So, what I ended up doing was revert the tolerance back to what is used to be for all Center osnapping (0.001 or absolute tolerance, whichever is smaller), and only test if the curve/edge has varying control point weights if “Center snap to approximate arcs, circles and ellipses” is not selected.

1 Like