[Bf-cycles] Volumetric patch update (now with new shiny bidirectional integrator)

[Thomas Dinges]

Hi storm,
I really appreciate your work, thanks for that. :)

But I wonder if it wouldn't be time to focus on the Volume part? Bi-Dir 
is nice, but I doubt it will be usable for trunk in the next few months.
It would be nice if you could release a clean volume patch (I mean only 
the volume features). This way we could look into getting this into 
trunk step by step, maybe start with a small subset like only 
homogeneous volumes.
I looked at the patch myself already, but it's a big guessing what 
belongs to Bi-Dir and what to Volume in a 10k line patch.

I know you work on the things you find interesting, which is fine, but I 
can only ask you to re-consider this position and work on a small 
subset, which then can be made "final".

Best regards,
Thomas

Am 19.01.2013 15:01, schrieb storm:
> As Thomas Dinges already answered, yes, it is very experimental modified
> Cycles, you better stay of if you not coder and interested in that
> bidirectional sampling. It have so many bugs and produce wrong pictures
> except very small subset then only i am can make, as i know patch
> limitations. Actually, with that last patch I already have found very
> wrong shades when number of bounces more then 1. When it will be
> finished, you expect to get something like
> Luxrender/Mitshuba/Yafaray/etc bidirectional mode. The reason i post
> that unfinished patch is i found that geom factor used was wrong, only
> one cosine term was applied (correct formula is (cos(a)*cos(b)/t*t) ,
> was cos(b)/t*t ), and after i fixed that i get much more pleasent result
> then before, that is all.
>
> _______________________________________________
> Bf-cycles mailing list
> Bf-cycles at blender.org
> http://lists.blender.org/mailman/listinfo/bf-cycles


-- 
Thomas Dinges
Blender Developer, Artist and Musician

www.dingto.org