The downside to a spherical map is that it maps to the surface according to angle of surface relative to the camera, and colours it black (or whatever outline effect you have). So if you have a long surface tilted slightly away from the camera, say looking along a car bonnet so the front of the bonnet is in the middle of the screen and the back of the bonnet, under the windscreen, is a little way higher than the midpoint of the screen, the whole bonnet is black. You also tend not to get a uniform outline.
If you replace the sphere in Ross's example with a cube or cylinder, you can see this problem clearly. If applied to a 'human character', front on it works, but if the character is lying down it doesn't.
The only really effective way is to use an inverted outline mesh, or use OpenGL!
If you replace the sphere in Ross's example with a cube or cylinder, you can see this problem clearly. If applied to a 'human character', front on it works, but if the character is lying down it doesn't.
The only really effective way is to use an inverted outline mesh, or use OpenGL!