Page 14 of 20
Re: ICE
Posted: Tue Jan 18, 2011 9:10 pm
by mlody47
AWESOMENESS!
Re: ICE
Posted: Sat Jan 29, 2011 12:04 am
by face
Have written a new Compound which gives me the position, tangent, normal and binormal from a curve back.
It´s nearly the same as the buildin UVtoLocation compound, but mine divide the curve by lenght.
With an input of 0.5 i become the middle of the curve back and not, like the UVtoLocation compound, where the controlpoints are located.
The output goes to a parallel transport algorythm. That has the plus, that the normals never flip.
The Background from this test is to export perfect strands for Octane
face
[vimeo]http://www.vimeo.com/19312008[/vimeo]
Re: ICE
Posted: Sat Jan 29, 2011 12:18 pm
by r4ytrace
very nice
Re: ICE
Posted: Sat Jan 29, 2011 2:36 pm
by mlody47
Face !!
Its sooooooo goood to have You here.
+1000 from me
Re: ICE
Posted: Wed Feb 09, 2011 1:33 am
by face
And here is what you get, when it´s done
Now I will spend time for the normals *arghhhh
face
Re: ICE
Posted: Thu Feb 10, 2011 3:08 pm
by face
Next try after some problems.
Now its time to work on the normals...
face
Re: ICE
Posted: Thu Feb 10, 2011 5:00 pm
by PeterCGS
face wrote:And here is what you get, when it´s done
Now I will spend time for the normals *arghhhh
face
strand1.png
This one is awesome!
Very cool work!
Re: ICE
Posted: Thu Feb 10, 2011 11:13 pm
by face
Normals should work now...
2.4mio triangles
face
Re: ICE
Posted: Fri Feb 11, 2011 7:31 am
by andrian
Nice one, but what about intersections?? Can you avoid them?
Re: ICE
Posted: Fri Feb 11, 2011 1:53 pm
by face
If you can code a collision detection in ICE, yes
face