Spline Loft Sculpture

Yes, each object was extruded from single spline, and those splines were in opposite directions.
So your saying an object based on a spline inherits it's drawing direction.
Which means this inherited trait is polygon based in the end, correct?

Is there a way to reverse polygon drawing direction, or would I have to go back to the splines?
 
Last edited:
For polygon meshes it´s Polygon: Flip normal and in Creator objects it´s the property "Flip Normals" and for Splines "Reverse sequence".
In vector drawing apps there are fill rules: evenodd and nonzero - but these are just a cosmetic feature for the renderer to tell if there´s a hole where multiple paths met or not. Cheetah3d is more tolerant and takes merged splines most likely as evenodd and keep holes no matter the path direction. Here´s a total mix of directions and Cheetah3d keeps the holes&normals intact:

evenodd.png


This isn´t curable by the fill rules anymore unless you pick at least 2 paths and reverse their sequence manually. The little tick marks between the nodes shows the path direction in Inkscape:
PathDirections.png
 
I've been trying to Illustrate the behavior Frank described
as the reason for certain normal flipping Boolean Failure.

The problem is the Extrude Creator ignores the spline direction.
Clockwise and counterclockwise splines extrude with Normals in the same direction.

I've tested enough to say it doesn't seem to have any effect on the way the Boolean
sometimes flips normals of one, the other or both. And sometimes both objects disappear.

For now it's just Boolean Voodoo. Maybe it'll help if I burn some sage and do a cleansing ritual pre Boolean. ;)

SplineDirectionExtrude.jpg
 
Are you Boolean-operating without polygon cover? If so I wonder it's working at least as you don't have watertight meshes methinks. So that's might be the reason for this behaviour. In reply #118 I showed a "special" case where it's not working; 1 spline is angled by 1 degree within the compound Path.
 
I set that up in that way to force a failure. I never expected it to work at all with this configuration.
I intended to Shell the open object and fill one end of the other.

The surprise was that it did work, and left the subtracted object and flipped it's normals.

Watertight is not necessary anymore.

WaterLoose.jpg
 
Now to get back on track with the Boolean Subtraction operation.

Here's a shot of the Boolean Creator setup. Neither object is watertight
and the contacting/intersecting surfaces are not all covered.

This should not work, and because of the flipped polygons, it looks like it didn't work.
But it did work, Cheetah actually did the subtraction, it just didn't remove the subtracted object.

The lesson is check your failed Booleans carefully, they might be salvageable.
In this case it's more work if I set this up "correctly" by
shelling Element 2 and filling one end of Element 1.

B Fail.jpg
 
Here's how the "correctly" set up objects would look.

I tried flipping Normals before Frank but that was unsuccessful.
And as I said before, it did work the way I did it, so it's not needed.
But thanks anyway.

B Corrected.jpg


B flipped.jpg
 
I forgot to mention I changed object positions under the Boolean Creator as well as flipping both sets of normals.
 
AppleIDs can be a nightmare - I once set up a new mac and had a typo in my email - OMG.
At least they provide free call support here. Good luck.
I finally cleared up my Apple ID problem, and I'm looking to upgrade my OS.
It looks like Monterey 12.4 is the one I can use and that's the latest I guess.
Any known problems I should be aware of?
 
Back
Top