Hi niko,
> Would it not help eliminate seams in more difficult blends?
No, not really - G3 and G4 are about equalizing "higher order" geometric properties where the blend surface touches the other surface edges.
The properties that they try to match are not really distinctly visible like a seam.
G2 means to have equal surface curvature where the surfaces touch. At this point there is already not any visible seam between them.
G3 means to have an equal "rate of change of curvature".
G4 means to have an equal "rate of change of the rate of change of curvature" - it's getting really very esoteric and removed from something actually visible to the eye at this point.
In addition, all of these things only make these properties equal right at the spot where the surfaces touch.
G3 and G4 require using a larger number of control points in the blend surface so there are some side effects from using them.
It's a very common misconception in the "Industrial design ultra cool surface styling" world that some higher continuity values like G4 are necessary.
It's better to have something like G2 where the shape is overall more evenly distributed throughout rather than something like G4 where the surface changes too quickly in just a close region nearby its edge. I mean the overall surface shaping is really more important than whether it's a G2 or a G4 type connection.
There has been a kind of "marketing bullet point list" arms race in this area for a while, something like G4 is a good example of something that looks nice on a marketing brochure and sounds like it should be better while it actually does not particularly serve a useful purpose.
If you draw a regular NURBS control point curve, it is actually only G2 itself between its different internal spans.
It's even likely that the G4 blend surfaces that you are talking about are only G4 where they touch the other surface but only G2 within different spans of the blend surface itself, because it's probably a cubic surface in the other direction than the blend edges.
- Michael
|