So, Mark II of that new technique I'm working on. I've reshaped her face, given her something closer to regular clothes (which would seem to do more to demonstrate the technique than I thought) and refined it a little.
This is Wendy. To the left is the new technique, to the right for comparison is a full 3D render of the same character. Ignore the little glitches on the right image, something went a little wrong when it rendered but its 95% right.
She's still not quite right, her shoes are meant to be red rather than "shiny grey plastic", I don't know why they keep coming out like that. Also cartographer confirmed my suspicions that there are not many women who would be caught in that particular colour/style combination but it'll do for now. Overall I'm a *damned* sight happier with this version, slightly more of a three dimensional feel to it but I think it strikes a decent balance.
Next is the Professor. I've had a try at him already and he's not co-operating terribly well due to the technical limitations of the model I'm basing him off and the end result is that he has a distinct tendency to come out looking like a badly made up transvestite hooker with eyebrows made of modelling clay.
Edit: Also, her bellybutton seems to have vanished.
I think I can improve the process a little more actually and fix the jagged outlines that aren't showing up everywhere they should.
Right now it involves two parts, one renders the flat colours and has absolutely no highlighting whatsoever. Then I render a black and white version in full 3D. I layer the 3D one over the flat one, make it semi transparent and there you go. The outlines are added to the image while I render the flats but I think if I used a normal render process (I use cartoon at the moment, which automatically adds the lines) which would *not* add lines and then did some tricks in photoshop to trace lines around all the joins in the colours it might do the trick. Most of you have no idea what any of that meant, this paragraph is mostly for my own benefit.