This is not dual contouring

[Saratii]

Please don’t describe this as “nearly the same as Dual Contouring.”
That’s a significant oversimplification. The defining feature of Dual Contouring is its ability to preserve sharp edges and corners.
As evidenced by the rounded corners in your cube example, it cannot produce the same shapes or sharp features as Dual Contouring.

The entire purpose of Dual Contouring is to achieve those sharp corners at the cost of additional computation. If sharp features aren’t needed, there are many faster algorithms than this.

[bonsairobo]

The language in question:

It is nearly the same as Dual Contouring, but instead of using hermite (derivative) data to estimate surface points, Surface Nets will do a simpler form of interpolation (average) between points where the isosurface crosses voxel cube edges.

The goal with referencing dual contouring is twofold:

1. The produced mesh topology is literally a subset of the dual graph of the voxel grid. I believe this is where the term "dual contour" originally comes from.
2. Simplify the explanation, because the algorithm is literally the same as dual contouring except for the vertex placement and normal calculation, which I've explicitly called out above.

If you have a specific suggestion for how to change the language, I'm listening.