Well, no. When I first started looking for something to learn 3D I found Anim8or, and that’s what I used for the rough model of the head. Unfortunately it has many serious shortcomings, so I looked and looked and eventually found Blender.
Joan of Arc (her true name is Jeanne d’Arc, BTW) tutorial is made for MAX. Perhaps it does not behave badly with triangles. Blender does especially when they are poles like the one on the cheek. Any triangle is hazard in blender.
When organic modelling, you have have only 2 choices possible without ressorting to ngons, using tris or poles (vertex with more than 4 incident faces). Blender handle the latters much better than the former, that’s all.
Look at my own head (see signature link) you wont find any triangle elsewhere than on the boundary of the 2 halves, and when welded they will form a quad. My personnal maximum number of tris allowed in a mesh is 0, i even took 1/4h to get rid of the invisible one closing the inner part of the ear
One of the tris you highlited is a quad is shape of tri. It would be better in a diamond formation but should not mess too much. any other can induce crease or ripple effect when subsurfed. This may be controlled for a still, but when animating, it’s a problem.
Now, more worrying imho is that, although your mesh is quite clean, you have in many areas a mix of quads of differents sizes, and quite a number of them. taking again shamelessly my own head as ref, you will notice i took much care of having roughly equally sized quads everywhere with progressive size changes. Again for a still, not a serious problem but one for animation.
Also you only half followed the loop concept. Was it intentional ?
Once again, it is a key for animation.
are you using reference images or just eyeballing it?
I’ve modified the actual ref’s pretty heavily, but these are the base images
The second pic is a cropped out portion of a picture I drew on a Palm, but as you can see it needed modification to fit it to an actual physiology. The first’s a good old Gimp job fixed to match the second.
Though at this point I’m pretty much just eyeballing in Aerinn’s finer features.
Ah, that’s what I wanted to hear.
I see what you mean. I’ve previously noticed how your head seems to have very few faces. Also, I have been wondering about the edge loops merging under the corners of the mouth (as in the d’Arc tut). Seemed to defeat the point of putting them there(?). I want to keep a tiny bit of detail for the little muscle bulges under the corners of the mouth (maybe I’ll complete one edge loop) and see if I can’t elminate most of the other chaff.
I’ll post updates when I get there…
As we subsurf after (level 2 in my case) having few faces is not a problem, it is even an advantage if you do it right.
I try to keep my full models (with very detailled hands) under 6k vertices non-subsurfed. Now, what was published so far is a bit deceptive as the very few vertices being full edgeloops, i will certainly add some loops ( it’s a matter of 5s) at texturing stage. What imports is to define shapes.
but you can only do that if you keep your mesh very clean and take great care about topology. Dont think that, because there is fewer vertices , it’s less work, it is more.
The other way of doing it is to make a more dense level 0 mesh and subsurf at level 1 only. But i feel this approach a bit more cumbersome.
Merging loops is not a baddie, but this must be done in sensible places. I’m a bit dubious about the ones under the mouth indeed.
Hope you’ve all got a good hi-speed connection… (I’ll shrink these sometime in the future)
As suggested, I’ve cleaned up the mess…er mesh…quite a bit. I’ve been modelling primarily at SubSerf level 2, planned
to bump it up to 4 when producing images. I still wonder about that diamond under the corners of the mouth… Should I
just extend the edge loop all the way across the chin?
I’ve fixed those sharp points at the top of the ear since snapping these pics.
Wow. That looks quite good. There is a little jerkiness in the shoulder movement though… every once in a while it seems to “snap”… don’t know a better word :-? You were very successful in making it look like there are real bones under her skin.
Ah, yes, you’re right. That’s caused because I wasn’t too careful about the IK in the arms for this animation (sorry, I should have been!). The snap travels back from the arm whenever the IK target exceeds the range of the arm. You’ll notice that the arm and shoulders snap at the same times (twice).
Did it with an action constraint to adjust the position of the scapula and teres bones.
When you stare at something long enough you miss those sorts of things. The first pass was really mechanical…