I believe this is the issue that Joey was referring to. You's need to somehow 
consider the "hole islands" rather than the polygon islands.

gray

-----Original Message-----
From: [email protected] 
[mailto:[email protected]] On Behalf Of Stefan Kubicek
Sent: Wednesday, August 01, 2012 04:37 AM
To: [email protected]
Subject: Re: A good "Cap All Holes" script?

Are the difficult scenarios you mention not ultimately a non-manifold geometry 
problem?
This reminds me on the "Any tips to fix non-manifold vertices in Soft?" thread 
as of June 26th.

We came up with:

"A vertex is non-manifold if more than two of its adjacent edges
do not share their second vertex with any other of said adjacent edge's second 
vertices."

Or as Martin Chatterjee put it slightly differently:

"I'd check if a vertex has a neighborPolygon that does *not *share an edge with 
any of the
other neighborPolygons"

According to this it should be possible to identify problematic vertices and 
split the capping process up on a "per island" basis.




> Back several years ago I spent quite a bit of time trying to develop a
> "Fill Hole" script similar to Maya's. Beware this activity when using
> something automated.
>
> I ran into immense difficulty trying to develop a method for sealing
> winged holes. In general holes which are an island upon themselves are
> easy to fill. That means that if a quad hole has an adjacent  poly
> present at each  edge and more importantly each vertex, totalling 8
> polys surrounding the quad hole, the hole is extremely easy to fill. I
> succeeded quite well at accomplishing an automated script to fill all
> holes on a mesh in one click.
>
> The problem however is if any hole has another adjacent hole present at
> a shared vertex (oxymoron? since you can't really share vertices for
> something that technically is a void), but separated by two polys winged
> at the vertex, its extremely difficult by conventional standards to not
> identify the two holes that are winged as a single hole.
>
> If memory serves me right  even Maya had an issue with this as you could
> delete a single poly on a mesh and if the sphere was selected as an
> object running Fill Hole would automatically seal all valid holes.
> However if their were two holes and they were adjacent, or winged at a
> vertex, Maya was smart enough to know how to avoid the situation and
> would prevent the winged holes from being filled with a single poly.
>
> The point is, its fairly easy to to write a script to fill the hole, but
> it must be smart enough to prevent the winged holes from becoming a
> single poly, which is very very bad. If you have a script that was
> written to do this task, test it on an example with winged holes before
> you proceed. You won't regret the extra effort.
>
> Incidentally this tool is not impossible to write, but as a script or
> plugin it requires special information about the topology be available
> for the user to query. In my experience XSI  did not provide enough
> vertex, facet and edge info to give the developer enough information to
> easily script a bulletproof fill hole tool.
>
> Joey Ponthieux
> ATOL Experiment Specialist
> LaRC Information Technology Enhanced Services (LITES)
> Science Systems and Applications, Inc.
> NASA Langley Research Center
> 15 Langley Blvd B1268 R1051
> Hampton, VA, 23681
> Phone: 757-864-6754
> EMail: [email protected]
> ____________________________________________________________
> Opinions stated here-in are strictly those of the author and
> do not represent the opinions of NASA or any other party.
>
>
> On 2/13/2012 1:12 PM, Alan Fregtman wrote:
>> Sweet! Thanks, Mr.Core! ;)
>>
>> On 2/13/2012 11:47 AM, Oleg Bliznuk wrote:
>>> Hi Alan,
>>> you can also do it with a little bit modified "cap holes" compound via
>>> ICE, we are using it in ImplosiaFX
>>> http://clip2net.com/s/1ABrt
>>> -Oleg
>>
>


-- 
-------------------------------------------
Stefan Kubicek                   Co-founder
-------------------------------------------
           keyvis digital imagery
          Wehrgasse 9 - GrĂ¼ner Hof
           1050 Vienna  Austria
         Phone:    +43/699/12614231
--- www.keyvis.at  [email protected] ---
--  This email and its attachments are
--confidential and for the recipient only--

<<attachment: winmail.dat>>

Reply via email to