Hi milkywaif,

> Isn't that better modeling with less points possible? Less point should mean less calculations, faster
> viewport feedback, smoother geometry, more precision, etc... Please enlighten me.

Yes that's true in general but there can also be "diminishing returns" so it's not always that productive to totally optimize for that.

Like in this case here sweep will actually not generate a minimal point surface, it will be a cubic surface in the axis direction because usually sweep is focused on doing curvy path things. So you'll see 4 control points for the vertical direction of the cone if you use the Pipe method. But since it should be considerably more convenient and it's still a simple surface result I'd think it should be fine unless you had some special reason to optimize point count above all other considerations... That's why I was asking about that.

As long as things are structured reasonably it's usually not too important to get too overly concerned about getting the absolute smallest number of surface control points possible - some other factors like surface quality (whether it has wrinkles, self intersections or awkward bunching), and having more minimal _topology_ can be kind of more important than surface control point count just 100% by itself. There can be some relation between these things like it's easier for a surface that's super dense with control points to have wrinkles or bunching in it, but that doesn't really apply so much to 2 control points versus 4 control points for example.

- Michael