Chris Huff wrote:
> 
> > It needs icicles.
> 
> Isosurfaces!
> 

Or maybe blobs would also do for that.

> > Also, it looks like the snow fell straight down, does your method
> > allow you to defind a wind direction that would place the snow to one
> > side of everything?
> 
> I think he just used the trace() function going straight down from the
> sky. Slanting the direction you give trace() should give you something
> like that effect.
> 

As i said in the reply to Josh, that alone would not be that good, at
least variation of the blob component size according to the normal of the
objects would be necessary.

I also think at some places like the 'N' in 'Snow' some 'surrounding
observation' would be necessary to avoid the snow overlapping with the
objects too much (right now i only use one 'trace' per component).  Maybe
some components with negative strength could also help there.

Christoph

-- 
Christoph Hormann <chr### [at] gmxde>
Homepage: http://www.schunter.etc.tu-bs.de/~chris/

Post a reply to this message