Making Holes In Subd Surfaces

[southern-boy]

Hey everyone,
I have a quick question. How would I make a hole in a subd surface after it's converted from polys. Everytime I make a hole with Booleans>Difference and get the result I want and then convert it to subds it's all out of whack and not clean! How would I dot it without having to manually do it on your piece of polygon geomtry and then converting? Is there a way?

Thanks guys!!!

[drblacksnake]

booleans are most lame way to model in general. only its advance is anyone can make it quick.
accordingly subDs - no.. subDs require thoughtful modeling, so You wont be able to have a good subD without good poly, and this is not quick-boolean-way. the way could be doing booleans, then cleaning mesh, then converting to subD.

[joojaa]

People dont realy thinkabout it but booleans allmost always cerate a bunc of nsided faces. And quite at random unless youve planned your boolean very carefully to just that application (in wich cas eyou wouldnt boolean).

ok mad ascii art to rescue.

imagine

CODE
-----------------------------------
|          |            |            |
|          |            |            |
|          |            |            |
-----------------------------------
|          |            |            |
|          |            |            |
|          |            |            |
-----------------------------------
|          |            |            |
|          |            |            |
|          |            |            |
-----------------------------------

People dont realy thinkabout it but booleands allmost always cerate a bunc of nsided faces. And quite at random unless youve planned yoru boolean very carefully to just that application (in wich cas eyou wouldnt boolean).

ok mad asii art to rescue.

imagine

CODE
----------------------------------
|          |            |            |
|          |            |            |
|          |            |            |
----------------------------------
|          |            |            |
|          |            |            |
|          |            |            |
----------------------------------
|          |            |            |
|          |            |            |
|          |            |            |
----------------------------------

minus a cylinder ->

CODE
----------------------------------
|          *---------- *            |
|         /             \           |
|        /               \          |
-------                 ---------
|      /                   \        |
|      *                    *       |
|      \                   /        |
-------                 *---------|
|        \               /          |
|          ------------           |
|          |            |           |
---------------------------------

Now see how allmost all the faces broke into nonquads. Now because both normal AND subd interplation ahoppens along edges (and subds even worse along teh quad) youve broken whatever you had!

In ganaral bboolean only makes sense on a solid modeller that uses exclusively nurbs.