Hi Marc, well it certainly looks promising!

- Michael

Cool, looks like the bug has finally been fixed. It would have been nice if Radiance had acknowledged the problem/fix in this thread, but I can understand him wanting to put things behind him. I hope there are no hard feelings about this issue. I don't know what took place with Paq in the private octane forums, but I assume he was only sticking to his guns while everyone was brushing him off. I think in the end Octane is a better product for his efforts, and it can now be fully embraced by the CAD community (which I would think to be a huge userbase for this kind of product).

That dual 480gtx video was nice, btw. These gpu based bi-directional path tracers are definitely going to change things over the next few years. Fun times.

The new release candidate is out now, I've tried Paq's superball and it seems to render fine as far as I can tell. But I don't have a lot of experience in this area so I hope other people with Octane license will jump to testing. The test cylinder shows no seams.

As for the flat polygon test - it mostly looks ok, there's some weird "shadow" creeping up from some angles, I don't know if this is related to vertex normals or the lighting system or what. But at least the polygon is shaded now.

Hi candide, that is looking way better, the vertex normals are definitely coming into effect there.

There should likely be a significant increase in render quality there now.

Most likely the only thing you need to do now is to uncheck "Weld vertices along edges" when you export from MoI.


> there's some weird "shadow" creeping up from some angles,
> I don't know if this is related to vertex normals or the lighting
> system or what.

Hmm, hard to tell what that is. If there is some reflection going on instead of only diffuse lighting, that could just be some reflection from the environment or something like that.

Or it could be something like individual vertex normals flipping to the opposite direction if they end up pointing opposite of the viewing direction. If that's the case then you wouldn't really see it on a more typical kind of mesh that doesn't have so much variation between the vertex normals on just one polygon.

- Michael

Part of a model I struggled with before...



Marc

I just love when a software investment pays off - I'll give the latest version a good test this weekend! ;-)

-Will