news: chr### [at] netplexaussieorg...
> But it is still less accurate than radiosity could be, as far as I can
> tell.
Accuracy here is really a non-issue given the benefits...

>Why use a high dynamic range environment map instead of sampling
> the actual environment?
> Speed? The ability to use real-world samples
> instead of coding a background?

Creating complex environments just to make them look nice in reflections and
project the right lights is rarely done :
for practical purposes, you just don't code the whole St Peters Basilic to
make a background (or you've really been raytracing too long...). There's a
example by Jaime who did a fake HDRI that way, but he used a complete scene
he had done before.
Speed is of course a big issue there too, since maps render very fast.
Also, natural light environments taken from photographs have a complexity
and a richness that is unmatched by usual 3D lighting, which is why people
go ooh and aah in front of these pics.
Now HDR isn't the solution to all lighting problems, and is only valuable in
certain circumstances, like "packshot" images (close ups of objects), but
then it is very good and extremly simple to set up using regular radiosity
parameters.

G.

--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters

Post a reply to this message