That looks amazing, total photo-realism! How come the diffuse and normal don't clash? The diffuse seems to have a lot of detail painted on.
Btw Teddy, will you be allowed to show us the stuff you worked on now that The Darkness has been released? ![]()
That looks amazing, total photo-realism! How come the diffuse and normal don't clash? The diffuse seems to have a lot of detail painted on.
Btw Teddy, will you be allowed to show us the stuff you worked on now that The Darkness has been released? ![]()
I'm stunned!
As for the artistic thing: Although this looks fabolous I think there is no limit for creativity and perfectionism. Think of programs like zBrush and Mudbox, almost every picture in the polycount "What are you working on?" is a beautifully sculpted head. So the severity decreases but the challenge increases... therefore it is still the artist not the tool ![]()
(damn, I want to be able to program :x)
Quote from TeddyBearMaking heads by hand sure is cool but after having done 150 of them (like some of the modelers were assigned to do in The Darkness), you get pretty sick of the iterative process. In a professional production with limited time there are two things that matter: output and efficiency. This method gives a better output than 6 days of manual labour (standard time line), requires one click and about five minutes of processing. This gives modelers opportunity to focus on other time consuming tasks with greater precision.
couldn't agree more - and this goes for every aspect of everything ever, imho
Thanks
How do you mean, D3ads?
The albedo represents a fake model of diffuse light reflection (visible as colour to the human eye), whereas the normal information refracts and occludes portions of the light itself. The specular map, in contrast to albedo, acts as a mirror reflector of light. They don't clash simply because of the fact that they are all fundamentally unique materials with different ways of simulating interaction with incoming light.
e-freak, it's hard to predict. The material acqusition is completely done, so is the pre-defined basemesh. Currently I have three different approaches for deforming the mesh:
A type of laser registration process which is my own invention. It registers layers of the human face, seen as a three-dimensional finite integral (consisting of layered primitive volumes resulting in a complex body). The program then manipulates the mesh's vertices using this data.
I'm thinking about a Stereo Matching approach which by far is more accurate (generates height data within the sub-millimeter range). If I go with this method instead, the program is going to take quite some time to get finsihed.
The third and simplest approach (which is currently used as a place-holder for one of the two methods described) reads a set of color coded pins attached to the face. The model has corresponding reference points, elastically deforming pre-defined vertex networks of the base mesh.
So it depends on how you look at it. It's already done but is under heavy improvement.
D3ads, all I did for The Darkness was basically music, I got hired right when the game was about graphically finished. Look for the harmonica guy (George Hadel) in the Fulton st. subway ![]()
Teddy: That's some mightily impressive stuff. It's very cool that you work all day as an artist and go home to make technical stuff. Not many straddle the artist/technical dude divide like that.
Out of interest (and if I may ask), is this stuff entirely of your own making and, if so, how does that stand up in terms of ownership? I.e. do you retain all the intellectual & commercial rights now that you're working full time? I'd be very interested to hear how you go about inventing this stuff -- it's fascinating stuff. I've worked on a few programming projects but never invented something entirely new. Also, what kind of background do you have that helps you on the technical side (programming / maths etc.)
Also I wanted to ask, You need the reallife guy (or specific set of photos) to perform this scan right?
Buddy: I have a method for extracting SSS maps as well, although not shown in this example. One powerful thing about this technique is that, besides the naturally acquired materials, you can morph pre-defined sets of materials to go along with the new model's UV-structure. These materials could be highly advanced vein maps, freckle maps, sss enrichers etc. There simply is no limit to how much tweaking you can apply to a head to make it look extra good. The nice thing is that this tweaking process then will make all the models look better.
Defrag, actually, it's the property of Starbreeze. They wanted part of it, so I have been funded with whatever technical devices I need.
To answer your other question, I've always thrived on creativity and the feeling of achievement, when I see the potential of something I simply start working on techniques to extract that potential. Honestly I don't consider myself a programmer nor a mathematician in their true form, I'm simply helped by the fact that I've studied alot of math and programming - it renders the possibility of thinking/speaking in a technical language.
Buddy, yes that's correct. However, with the database of faces that is built up you will be able to actually make new, unique faces based on elements of that database (through 2D/3D morphing).
I felt like making some retro metal textures for a dm map (maybe ?) ![]()
A tiny fuckaround with them (I also made all the other textures in this scene. oh and sorry for the low res pics
)
[Blocked Image: http://aycu10.webshots.com/image/24529/2004447944955254208_rs.jpg]
For my de_thearts map - Now in it's 5th iteration! ![]()
![]()
Quote from MinotauroI felt like making some retro metal textures for a dm map (maybe ?)
Owned. Reminds me some texture packs from Quake 1...
i made this texture for my coast / normandie / ww2 / what ever "is-not-yet-a-project" project
[Blocked Image: http://www.thomashess.net/normandiewall.jpg]
Needs more nextgen shaders.
i don't know your latest 2 texture tries (the one with the model aswell) seem quite rushed ![]()
but on this one the jpg quality seems to kinda be low quality settish as well, so i don't know... ![]()
It looks weird and rushed... it's pretty obvious that you used 4 photos: top, middle part, transition and bricks. Besides that, the transition looks too dark to look convincing.
Quote from MinotauroIt looks weird and rushed... it's pretty obvious that you used 4 photos: top, middle part, transition and bricks. Besides that, the transition looks too dark to look convincing.
so what should i change? except for the generic normalmapping shit.
The stones look too bumpy/noisy, otherwise good imho ![]()
It's just not coherent piece, I mean I can focus my eyes on 3 parts of texture (top middle bottom) but can't find it easy to look at whole thing... if you know what I mean ![]()
[Blocked Image: http://www.button-masher.net/wip/texture/q4_eclipse_001.jpg]
[Blocked Image: http://www.button-masher.net/wip/texture/q4_eclipse_002.jpg]
Practicing modeling out textures for Q4. Diffuse is just scaled up from the original NS texture. Think I need to exaggerate the vertical bits a bit more, round off some parts a bit better, etc... but it's a start!
Don’t have an account yet? Register yourself now and be a part of our community!