It seems to me that one of the big elements of reality is the imperfections. A texturemap has slight variations in colour which are harder to get with a material setting. A flat surface isn't quite smooth. Straight lines aren't. Symmetry isn't quite perfect. So, a little challenge. Make a picture of something simple, which isn't quite perfect.

http://www.this-wonderful-life.com/gallery.htm This individual seems to have done well with getting characters to look real.

i used to think that the ultimate goal in 3D was to 'fake' reality, and get the best and highest resolution photorealistic texture imagineable. ive since come around in my thinking... i see texturers that have been around awhile buying higher and higher resolution cameras in order to make more and more photorealistic textures. i think that 3D should have some style and polish to it... anyone can just keep taking higher and higher resolution photos and pasting them together - especially if you do them of people who are middle-aged or older with pronounced pore detail. you could teach a chimpanzee to do it. if you look at the top animation studios like square - and even the guy whose link you included in this post - youll notice that their textures use the human skin as a 'visual' reference, theyre not uber-photorealistic in the sense that they are composited from 6000x6000 pixel macro photos of the human face. most of their detail is actually quite subtle, and is from specularity/bump mapping rather than a 10000x10000px photo texture. super high-res does not necessarily look good when rendered... all it is doing essentially is taking a photograph, mapping it onto a surface, and from some angles if the texture is high-res enough then it will 'look' like a photo... but if thats what you want then why not bypass the 3D stage altogether and take a photo :) suffice to say, all my future textures are going to take an altogether different approach... im sick of the way we are headed with the photorealism thing. i used to scoff at handpainted textures and think that a photo would offer superior detail - but should the primary goal of 3D really have to be 'mimicing photos' - or should it rather be to convey a sense of style, and render aesthetically? /end rant cheers, -gabriel

on a side note - i applaud this Liam Kemp guy for actually choosing a more realistic body style. im tired of the same old giraffe-like 3D characters i keep seeing everywhere - especially when it comes to stick-legs.

I tend to agree with Blackhearted on this, I myself prefer painted textures because they have style. While I applaud those technically perfect renders that nearly fool you into thinking it's a live person, I am more drawn to the other kind precisely because they are not of this world. I like the artistic style and expression that those textures bring to characters.

Extreme detail is wasted anyway. Human perception isn't a dot-by-dot scanner. It looks first for motion, then for edges, then for forms. Details are largely filled in from memory of similar things. So the job of a texture isn't to map each pixel of the original onto one photoreceptor of the retina; rather it should evoke the "right kind" of similar thing. (Picking the "right kind" is where the art comes in, of course....)

My feeling is that consistency is the key, whether you're going for real or feel. :O) If elements of the pic look like they don't match with other elements it kills the illusion.

just watched the movies he had on his site. he really did do an excellent job - what breaks the illusion for me is when she opens her mouth... for one, her teeth are almost comically gigantic, and she seems to be suffering from the 'glowing mouth' that plagues most poser renders. its very rare for me to see a 3D character who looks good with their mouths open - most seem like they have lightbulbs in their mouths.

I'd agree that perception is important. And transferring something to a flat image can alter what matters. If I have a figure resting a hand against a wall, it doesn't have to be exactly at the wall surface, but watch for the shadow. As I recall, one of the classic examples in cinema was a short piece by Eisenstein, where he showed an actor reacting to various events. Strong emotions. He showed film of the events, and then cut to the actor's face. The only thing was, each time he showed the actor's face it was the same image, printed from the same piece of film. Nobody noticed. They saw the emotion they expected. I've been using the Lemurtek figures, humans with heads from various Poser animals. They're short on the expression of emotion, even when you remember to use their tails. So you have to set up an emotional context, and tell the audience what to see by some other means.

I'd agree that perception is important. And transferring something to a flat image can alter what matters. If I have a figure resting a hand against a wall, it doesn't have to be exactly at the wall surface, but watch for the shadow. As I recall, one of the classic examples in cinema was a short piece by Eisenstein, where he showed an actor reacting to various events. Strong emotions. He showed film of the events, and then cut to the actor's face. The only thing was, each time he showed the actor's face it was the same image, printed from the same piece of film. Nobody noticed. They saw the emotion they expected. I've been using the Lemurtek figures, humans with heads from various Poser animals. They're short on the expression of emotion, even when you remember to use their tails. So you have to set up an emotional context, and tell the audience what to see by some other means.