TITLE: Law of Chance NAME: Andreas Grates COUNTRY: Germany EMAIL: Andreas.Grates@public.uni-hamburg.de WEBPAGE: - TOPIC: Toys & Games COPYRIGHT: I SUBMIT TO THE STANDARD RAYTRACING COMPETITION COPYRIGHT. JPGFILE: ag_lawch.jpg RENDERER USED: POV-Ray for Windows 3.6 TOOLS USED: Paint Shop Pro 7 to add signature and convert to JPG RENDER TIME: about 3h, lost exact measure HARDWARE USED: AMD Athlon 650 / 320 MB RAM IMAGE DESCRIPTION: At the heart of the picture is the 11 x 11 array of semitransparent, brightly coloured dice. Random four of them contain a bright light and provide the illumination of the whole scene. Which face of the dice is up, how they are rotated, how far they are off axisparallel and ideal position in the grid, which of six possible colours the corpus and which of two possible colors the eyes of the dice have is random, so that any pattern or coincidence is due to chance. The array is positioned on a rectangular pedestal made of a plexiglas-like material, on which also four infinite-seeming stacks of marble-dice foot. Again, the way they face, their off axisparallel rotation and off center position are random. The only landscape visible is the floor - giant-dice, showing random numbers, coming up to a random heigth. All objects in the scene are in some way subject to chance and are mere stage for the wild play of light. How does it fit to the topic? Many games in a traditional sense of board games feature an element of chance. In many cases they implement it by a simple toy, the die. I wanted to show the die with its predominant role as guard of chance. DESCRIPTION OF HOW THIS IMAGE WAS CREATED: Neither modelling nor texturing are worth mentioning much. Modelling is in large proportions restricted to the dice, which all go back to one single macro, which delivers an axisparallel die of height 1 standing on the XZ-plane at 0,0. It is a simple intersection of a box of uniform length 1 on all edges and a circle with the diameter sqrt(2). (Why? Well, in 2D it's easy. I wanted the circular part of the face to touch each edge at exactly one point. From there it's Pythagoras at most simple.) The eyes are really made via difference from corpus, spheres reaching half into the corpus-volume. The pedestal is ad-hoc CSG without much planning ahead. Texturing was mainly restricted to choosing a nice texture frome "stones.inc" (chose T_Stone8, see source) and two finishes from "finish.inc" ("Shiny" together with everything complete opaque, "Dull" everything else.). Worth mentioning is the way of generating the colours for the semitransparent dice. Imagine RGB-colorspace as 3D-cube. Take vector through origin to <1,1,1> as axis, take perpendicular vector originating at <.5,.5,.5>. Now you can choose a random colour from a closed colour-circle by rotateing that perpendicular vector by a random argument! I confined that to six possible colors and generated (by and large) random numbers from 1 to 6 (with a certain chance of 0 - didn't seem to be in "rand.inc" to create discrete random numbers of even distribution. Or just didn't see.). Maybe it is worth mentioning that every last object of the scene is affected by some random element. Apart from the camera and the skysphere (not really visible, but maybe some indirect impact in reflections and radiosity) really everything is. There's the dice - they are generated by a single macro, that put's 'em out with random face up and random rotated by some 90 degrees around the y-axis, for a start. The pedestal with all that's on it is rotated by something random between 5 and 40 degrees. The way the array looks and the dice of the columns are stacked heavily points towards some random "jitter". The giant dice have a random offset how far they go down from actual 0-level. And the lighting? Four sources placed random - not completely random, but inside one of the dice of the array, but random among 121 possibilities. P.S.: Maybe someone could drop a comment on the signature? I created a file as "stamp" and would like to know wether to change something.