Posts by TeddyBear

  • W.I.P. Critique

    Quote from Sa74n

    ive seen this before. here is another take on this scene by a fellow mapper. hope it serves as some inspiration for you. http://littlelostpoly.co.uk/images/gall ... /nd039.jpgbtw. i think that crossfire is probably your strongest map atm. i wouldve liked to see a slightly different take on it but i guess you decided to stick as close as possible to it for the blackmesa mod. still looks good though

    Now there's a great map!

  • totally random texture thread

    Haha

    Well because of the simplicity of the process, I am writing an article that I will probably publish for the game development community to read and adapt. I will have to wait for any patent approval/denial before I release it though (just to be on the safe side).

    Here's a small introductory excerpt (still rough):

    "In recent time quite a few new approaches to model and environment texturing have arisen; techniques that have characterized the standard of graphics in the game industry of today. Materials suddenly behave more realistically – with the help of advanced so called “shaders” the texture(s) of an object can calculate complex behaviours such as correct refraction (du/dv) and reflection (normal & specular) of light. In “next generation” projects (currently yet to be released) certain shaders and texture maps (named “relief maps”) can even collaborate to curve pixel geometry. A comparatively primitive precursor to this technique is the “parallax map” which manipulates the pixels of a texture in two dimensions to resemble 3D curvature.


    In this guide I will lecture you in the fundamentals of creating highly complex texture information with fairly simple methods. When you are through with this guide you will be able to produce better texture maps than most game development studios of today. Why is this?


    Texture information such as normal- and displacement (height-, parallax-, relief-) maps is rather simply put characterized by a genius merge between the complexity of model geometry and simplicity of material information. In the normal map case you are enabled to replace (theoretically) an unlimited amount of triangles with a single normal map and still produce very similar results. The normal map consists of R-, G-, and B channel information representing welds of shadows and highlights which are used to manipulate the normals of a model’s vertices.


    There are two common cases – one for tiled environment texturing (seemingly flat surfaces such as walls, floors and the like) where you often transform a high polygonal model (containing dents, scratches and other details you might find) into a normal map which instead applies to a flat two polygon surface. The other common case is converting the normal information of a high-poly model into a normal map and then applying it to a corresponding model with fewer polygons. Depending on the quality of the graphics engine the only thing that can tell the two models apart is the outline edges of the geometry – all corresponding surfaces within these lines are identical because of the two alternative ways of rendering normal information; polygonal contra textural.


    My goal when inventing new methods of creating realistic texture information is strict but simple: a process that is uncomplicated, economical, time-efficient and timeless. By timeless I mean a method which produces a result never growing old or obsolete – the product is only allowed to improve as the art of rendering technology develops. However, in a few years from now the kinds of texture materials described may even be replaced with more rendering efficient techniques – the future is hard to predict. Nonetheless, you simply cannot produce more realistic results than the ones I’m going to teach you to produce today. Why is this?


    It is because your results are derived directly from a photographical reality. You have read me writing about computer models producing the materials I am describing. There are already hundreds of tutorials concerning this, why write another one? Well, I’m not. Think about the term “model”. What is it? That’s correct, a representation of reality. There is also something called “the human factor”. When trying to represent reality we can never be 100% precise (perhaps unless we devote our lives to training ourselves for the task). My goal is to cut the detour."

    Is there an interest for this? I feel that not only professors and doctoral students should be free to use this information, game developers can benefit from it as well.

  • Gearbox selling computers for dell?

    Quote from Bic-B@ll

    he looks like a character from seventh heaven or one of those shows on wb that has a bunch of teens and only 40 year olds watch

    Yeah well this Dell deal is handled in a pretty sleazy way. Questions like "What did you spend your annual bonus on?" were for example only displayed for the employes who identically answered "DELL Ultrasharp 24'' TFT 10024i1084012840812x PiXMA blabalala"... things like that annoy me.

  • totally random texture thread

    Final height product from my first real-world computing experiment (an entirely unique personal invention):

    [Blocked Image: http://img197.imageshack.us/img197/9858/handheightvu3.jpg]

    It differs a bit from the other one posted earlier in the way that it contains 428 frames of footage instead of 38. The result is proof I can take any real world object and turn it into 3D (with height and normal information) with one digital camera recording and one click in Photoshop. This can save a great sum of money as the studios producing the same results today do it with terribly expensive and complex equipment. A patent application is in the works