I corrected the faulty transformation from isosurface to mesh in the 
hardground (case of a an inner shape sticking out of the isosurface, 
misleading the MeshShape() macro).

As all normals were facing inwards, I processed the meshes in Poseray. 
Through this, the mesh sizes went also from about 38 Mb to some 80 Mb each.

This image (using radiosity) rendered in 12 minutes, while the isosurfaces 
would have taken several hours.

A problem remaining are the dark triangles visible in the cliff face. As far 
as I have been able to judge, the normals are all oriented in the same, 
outward direction and double_illuminate is used. I think this has been 
discussed in the past, but I cannot remember when or where. Any suggestion?

In any case, this exercise shows how attractive Bill Pragnell's macros are 
for speeding up render time by transforming isosurfaces into meshes. Also, 
as far as I can judge presently, the meshes seem to simulate the isosurfaces 
better than the Jaap Frank c.s. macros, but I may be wrong there. I remember 
however, that a couple of years ago, I was not able to obtain as much 
detail.

Any comments or suggestions are welcome of course.

Thomas

Post a reply to this message

Attachments:
Download 'IsosurfaceTest2-2.jpg' (82 KB)

Preview of image 'IsosurfaceTest2-2.jpg'