Hello,

I use a python scipt to export 3d datas. Then I import thoses exported datas in my OpenGL/C++ program.
All works great, expect for normals : vertices normals seem to be ALWAYS interpolated between the adjacent faces.
If this is good for a sphere, it isn't for a cube (which is not 'smooth').
Would you know how have the normals such as defined in blender ?

Here is the way I access to the normals :
(for a cube it should print {0,1,-1}, but here it prints {-0.577349185944, 0.577349185944})

Code:

import Blender import sys from Blender import Object, NMesh, Mesh, BGL, Draw select=Object.get() for ob in select :    nomMesh=ob.data.name    mesh=Mesh.get(nomMesh)    if (mesh!=None) :       nmesh=NMesh.GetRaw(nomMesh)       numFace=0       faces=nmesh.faces       for face in faces :          for ver in face.v :             print ver.no[0]             print ver.no[1]             print ver.no[2]

You want face normals, not vertex normals, which you have to calculate yourself.

... calculating them yourself

Code:

def vec_dot(u, v):    return u[0]*v[0] + u[1]*v[1] + u[2]*v[2] def vec_cross(u, v):    #   |   i   j   k   |    #   |   u0   u1   u2   |    #   |   v0   v1   v2   |    # (u1*v2-u2*v1)i - (u0*v2-u2*v0)j + (u0*v1-u1*v0)k    # (u1*v2-u2*v1)i + (u2*v0-u0*v2)j + (u0*v1-u1*v0)k    return [u[1]*v[2]-u[2]*v[1], u[2]*v[0]-u[0]*v[2], u[0]*v[1]-u[1]*v[0]] def vec_mag(u):    return sqrt(u[0]*u[0] + u[1]*u[1] + u[2]*u[2]) def vec_norm(u):    n = vec_mag(u)    return [u[0]/n, u[1]/n, u[2]/n] def vec_between_points(u, v):    # returns the vector from point u to point v    return [v[0]-u[0], v[1]-u[1], v[2]-u[2]] # thisMesh is a NMesh object i = 0 while i<len(thisMesh.faces):    aFace = thisMesh.faces[i]    norm = vec_norm(vec_cross(vec_between_points(aFace.v[0], aFace.v[1]), vec_between_points(aFace.v[1], aFace.v[2])))

(you may use this code, even though it is from two of my scripts (only one of which has ben released)

oh, I forgot to mention that this works best on triangles as they are planar. On quads one vertex is completely ignored. It may be prefered that you average your normals on a quad