Quote - To muddy the waters a little more, these polygons aren't REALLY concave or convex - those quads are cut internally into triangles, which can only ever be planar, at least in this universe.

Now, Mssrs. Riemann and Gauss might disagree with you. Not to mention Lobachevsky. :lol:

http://en.wikipedia.org/wiki/Triangle#Non-planar_triangles

I'm a little behind on my formal geometry!  You're right, in non-Euclidian geometry, triangles can be non-planar; however in the context of 3D modeling, everything is done under Euclid's paradigm.

Quote -
Have you been methodical? Like you haven't shown us step by step what you did and what you got, so I can't help.

I don't know if I'd call it methodical, but I have tried everything I can think of. A host of different lights. Lights with raytracing, lights with depth-mapped shadows. Higher and lower irradiance caching. AO and no AO. Smoothing on and smoothing off.

Just now, I've tested with one infinite light.

Depth-mapped shadows, no AO, large and small map size (256, 1059).
Depth mapped with AO and a large map.
Ray-traced light with AO. Ray-traced light without AO.
Ray-traced light with no blur, and with high blur (17.0).

Also tried adjusting atmosphere strength on the light.

In line with what you said earlier, the only thing that seemed to make some difference (but essentially, I could only make it worse than the images above, not better): Playing with the shadow min bias on ray-traced light. I tested between 0 and (and a bunch of stops in between). When I set shadow min bias to zero, the lines are darker and I see a lot more of them. Setting the smb unnaturally high, the marks seemed to start to fade, but only up to a point. (To my eye, the marks are slightly lighter at 2.0 than at 4.0.) But shadows set at that high of min bias are surely pretty much useless for renders, anyway.

I guess it's back to the drawing board with the morph. :(

EDIT: I should add that I've also substituted head, neck, and face texture with another one, without correcting the problem.

I guess it's back to the drawing board with the morph. :(

Morph both the head and neck.

*EDIT: I should add that I've also substituted head, neck, and face texture with another one, without correcting the problem.

speaking of that, are you using any displacement to  the "head"  posersurface?

I see a divot under the chin, and some wrinkles under the eyes,  are those built  into the mesh,  or are you using a small amount of displacement here and there.

I'm not using any displacement. The wrinkles are primarily achieved via bump. The rest is affecting the mesh via the sculpt.

Not necessarily, no. REYES engines like FireFly or 3Delight an deal with non-planar quads just fine, they tesselate them to a perfectly smooth bilinear patch. As shown in the attached FireFly Render of a single non-planar quad.

Things do get tricky though when raytracing: since ray/triangle intersections are a lot faster than a ray/patch intersection, FireFly does triangulate polygons for the raytracer.

Well, as far as I can tell, I've solved the problem. I was afraid of trying to sculpt out the problem, since frankly the problem is invisible until render, so I expected to be flying blind in my 3D app. As it turned out, however, I went over the area just under the jawline a few times with the jelly smooth tool, and in all my renders so far since I'm not seeing the marks.

Whew.

It is a bit worrisome, because I realize that when I do full body sculpts, I may end up with similar issues elsewhere, and only a ton of rendering may uncover it. I guess I'll need to recruit some character testers... but I need to finish up with the textures first.