[Sidefx-houdini-list] Re: convert to NURBS

Andrew D Lyons alyons at anim.dreamworks.com
Fri Aug 26 15:16:46 EDT 2005


Thanks to all,

I've now solved this problem - for this type of case anyway.

Cheers



Andrew D Lyons wrote:
> Thanks Jim and Peter,
> 
> I actually tried this yesterday. This was the technique that I referred 
> to as working reasonably well for the simplified case I sent to the list 
> - but which tanked on my production model. I'll revisit it if necessary 
> though.
> 
> I'm looking at Antoines technique with great interest. There's some 
> serious old-school trickery in there. Thanks for the knowledge sharing 
> Antoine. Hopefully I can find ways to answer those questions as I go.
> 
> Cheers
> 
> 
> 
> Peter Robbinson wrote:
> 
>> Jim Rutherford wrote:
>>
>>> A real low tech approach would be to divide sop and remove the shared 
>>> edges.
>>> convert that to a nurb surface
>>> refine it a whole bunch
>>> offset it down below the original geometry and scale its Y to 0
>>> Ray sop that back up to the original (i'd divide the original up too)
>>>
>>>
>>> _______________________________________________
>>> Sidefx-houdini-list mailing list
>>> Sidefx-houdini-list at sidefx.com
>>> http://lists.sidefx.com/mailman/listinfo/sidefx-houdini-list
>>>
>>>
>>>  
>>>
>> With this method, if you turn on "interpolate through hulls" in the 
>> convert sop you don't have to res up the geometry.
>> In the ray sop I used "minimum distance" and turned on "transform 
>> points" .
>>
>> hmmm, it's not a rational surface however...How do I change that?
>>
>> PeterR
>> _______________________________________________
>> Sidefx-houdini-list mailing list
>> Sidefx-houdini-list at sidefx.com
>> http://lists.sidefx.com/mailman/listinfo/sidefx-houdini-list
> 
> 

-- 

====================================================================
Andrew D Lyons | FX Animator | Dreamworks Animation | 1 818 695 3963
====================================================================



More information about the Sidefx-houdini-list mailing list