well, after my first attempt at modeling a simple geometric form a few days ago i today tried myself on the beautiful face of natalie portman. I think i got the basic drift of face modelling, even if this particular mesh does not resemble the original in any way. But see for yourselves… btw, i keep hearing of edge loops and how important they are, could somebody explain and maybe point out flaws in my mesh? thanks
WildChild it’s looking good. You will probably find when you do the rest of the head that you will have to join some of those vertices together,otherwise it will look too blocky at the back of the head.
i hadn’t planned on doing the rest, i discovered that the images i used (out of a tutorial somewhere) where a bit crummy and distorted, so i asked my girlfriend to pose for some pictures. I’ll start over with her head then … oh and thank, but i still think my mesh looks strange and not at all like n. portman ^^
about the edgeloops, can someone tell about them (about their importance) or point me to an article where i can find out?
great, thank you
If you still need some good head references, try Got3d.com. The resolution is OK and there’s quite a good selection.
He’s using Natalie Portman for a head reference, remember?
“i hadn’t planned on doing the rest, i discovered that the images i used (out of a tutorial somewhere) where a bit crummy and distorted, so i asked my girlfriend to pose for some pictures. I’ll start over with her head then … oh and thank, but i still think my mesh looks strange and not at all like n. portman”
…not any more.
Oh. Right. I knew that, I was just testing your knowledge. One day you may be great enough to know my secrets.
cough
thanks, so basically you’re telling me, before i model a head i nee to think about where there need to be edgeloops before i start. so i can work in details comfortably and have an easier time animating. Is that right so far?
i tried to model my daughter from pristine ref pics, and found that moving a vertex a teeeeeeny smidge here and there, from render to render, makes a BIG difference in recognizing faces.
Wildchild: That’s pretty much dead on.
Yup. Edgeloops are the simplest most efficient way to define the contours of the face.