In article <3AA291F9.3E734EF6@gmx.de>, Christoph Hormann 
<chr### [at] gmxde> wrote:

> That's one of the things i was going to check, but i did not yet have time
> to do so.  Anyway right now there are only non adaptive blurring methods
> implemented and therefore it is quite slow when using a lot of samples.  
> 
> I already started to write a recursive algorithm only taking additional
> samples if the values differ, but i did not yet finish it.

I had done a pattern blurring patch before, but got stuck trying to 
figure out a good way to get smooth blurring without a huge number of 
samples. I don't know how suitable it is for isosurfaces, there are 
still areas with infinite gradient. I think I can figure out a way to 
get smooth blurring, though...
My patch worked differently, it was a pattern modifier, like the 
waveform modifiers. I never thought of implementing it as another 
pattern...seems like a good idea, especially in the case of block 
patterns, which wouldn't be block patterns any more after being blurred. 
I would suggest a different keyword, though..."blur_pattern", for 
example.

One question: why do you still have the "grad" pattern, which should be 
possible with the "warp_pattern" pattern?

Also, I would suggest that "visibility" be a separate pattern...it 
doesn't seem to fit in with the object pattern, and would be better off 
not being a block pattern. Incorporating distance would be a useful 
feature.

-- 
Christopher James Huff
Personal: chr### [at] maccom, http://homepage.mac.com/chrishuff/
TAG: chr### [at] tagpovrayorg, http://tag.povray.org/

<><

Post a reply to this message