How to explain meshes vs nurbs to a kid

Yesterday we had a 7-8 (?) year old kid at our workshop. He’s super quick at picking up Rhino. Last time I quickly showed him where to find the buttons for drawing solids, an hour later he had drawn a rocket including launchpad.
Anyway, we’re now getting into 3D printing the stuff that he’s drawing, and got to a point where I had to explain the difference between meshes and nurbs. To adults, I often make the comparison to working with images with pixels in photoshop vs working with vectors in illustrator, but that explanation will just make it more confusing here so I was wondering:
Does anybody have a super basic way or analogy to explain the dufference between meshes and nurbs to kids?

I would just show that the tiniest possible valid surface would be a triangle, which is going to be always flat.

Then compare a sphere created of big triangles to a nurbs sphere. Faceted vs smooth.

I don’t think an analogy is really necessary, since showing meshes in an exaggerated manner will probably drive the point home. Meshes will always be created from tiny planar things. And smoothness in mesh stuff comes from manipulating the normals.

So you get to show what a normal is, too.

I guess a good analogy could potentially be something like a spiderweb or woven fishnet that is mainly defined by silk threads, rope (mesh edges) and knots (vertices), and obviously subdivided versus something like a soap film within a boundary that is derived only from the edge situation (like a boundary surface) and physical properties (math for nurbs).

That’s not true, at least for modelling, where you often see and work with quads that aren’t planar at all. It’s just the underlying, rendered tri-mesh that has all planar faces, but that’s happening behind the scenes.

As I mentioned earlier, the tiniest possible surface is a triangle. Quads (and NGons) are all also just collections of tris.

It is exactly that what is interesting, especially for 3D printing where STL is used a lot. An STL file is triangles only.

STL (file format) - Wikipedia (incidentally the only two 3D programs I know are mentioned here, yay).

Yes, but as a “designer” you’re nonetheless superficially working with the quad or ngon and not with the underlying collection of tris that the render pipeline uses to display stuff behind the scenes.

That’s true, but in my opinion merely the fact that something is composed of triangles doesn’t really explain the difference between solids, nurbs, and meshes well. Even a solid (poylsurface) could be build from triangular surfaces or a surface tessellated with a triangular grid of subsurfaces.
Compared to a triangular mesh, you could then have two identically looking geometries, which is even harder to explain to children, since again you arrive at the more abstract question what’s a surface vs. a mesh face.

For 3D printing, I prefer the superior 3MF format. STLs seems archaic nowadays, even if still popular.

Yes, but that’s probably unnecessary in the context of explaining to an eight year old how a smooth nurbs surface gets 3d printed and/or why the resulting surface needs to be smoothed post-printing.

1 Like

As for pixels and vectors for curves
I think matches and rope might be better

This can be expanded into 3D by a mosaic and a balloon or textile



I would say that the big difference is what goes on between the points. In either meshes or nurbs specific points outline the shape. In meshes the connections between the points are defined by straight lines. In nurbs the connections can also be defined by more complicated formulas to allow smoother and more complicated shapes to be defined by less information stored in the computer’s memory.

Showing smooth vs flat shading might be useful to demonstrate the difference if accompanied by the right explanation. Mentioning the role of approximation when using meshes vs exact representation using nurbs might help.

With meshes points are the only thing. More points <=> More accuracy.