Hi all!

As of the last days, time is getting short till I run
into several examns (one each week for the next
6 weeks or so, the last ones less "in frequency", all
in all 7 of 'em nasty buggers), so I've not gotten much
further than cleaning some code (which of course isn't
really visible in the image).

Thus I thought, maybe it might be interesting for some
to get some insight into the technicals:
I'm using photons and radiosity for this image, which
would be entirely insane if I'd actually use the
stacked-planes-cloud-technique and all details.

So, what do I do?

I trace an image, with some details set to "off" or "low",
and replacing the glass-window with a solid don't-see-through
white placeholder. This is done to avoid radiosity affecting
the outside, which isn't really necessary. The radiosity-data
is saved to disk.
As the meshes in the image are required to have the same
detail as in the final version for radiosity to work correctly,
I can also generate the required photons in that pass, also
saved to disk.

At last then, I switch everything to "on" and "high detail", load
the radiosity and photon-data (as well as the meshes I
generated and saved to disk at some point), and trace that
one. Parsing is far below, as well as picking radiosity-samples
(none) or calculating photons (none), since everything gets loaded.
The same save-and-load procedure is applied to the raintrails
on the window as well, which still need some tweaking to look right.

Anyways, for those that haven't got the bandwidth, I've decided
to not attach all images to this post, but rather, reply to this post
four times with the different images.

Enjoy,
Tim


--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html
Email: Tim### [at] gmxde

Post a reply to this message