mrz00m

I have a iso-surface with vertices on it. I sample points close to a vertex and measure their iso-level and their position... If i sample 3 equidistant points from the vertex, for example randomly points on a sphere, i should have enough info to triangulate The Exact place of the normal, the least grey area:

Say that the green points are sample points near the vertex, we have their iso value, their position, so there must be a simple formula to figure out the apex of the sphere, the point where the normal of the vertex is? 

I tried already making a plane on the same iso level as the vertex iso, by taking 2 sample points of same iso value and doing the cross product of them to find their perpendicular normal value, and it was not so accurate and computationally expensive to sample many points to find ones of same iso level. so the above option would be amazing.

Thanks for reading!!!