When I was a kid in suburban London I used to spend hours staring through the front window of the toy shop on Penge High Street, which featured a dazzling diarama of pre-Subuteo table soccer games, HO gauge model trains and houses, Airfix models and packets of toy soldiers, and of course Britains' 'Swoppet' knights. I can still get this same feeling when I visit the German Faller site: http://www.faller.de/start.php?lang=gb&naviUp=0&naviDown=0 or thumb through any of its catalogs that are sold as collectors' items on eBay. The feeling of magic in miniature, the sensation of world-building. This is what a CG 3D artist/animator like Wagster does for a living. He builds worlds on his screen.

In the beginning was the vertex. This is a simple mathematical point that is expressed as a set of numbers in flat space. Create a second, connect them, and you have a vector line. In the old days of computer art--the mid 80s, the vast bulk of vector art was in 2 dimensions and ad agencies were full of Adobe Illustrator experts who were kept busy converting art to this format for printing. In those days, the rival of the vector, the bitmap, was still a messy and hugely RAM-intensive format--art museums and high-end magazines could afford to reproduce photographs that way, but any sort of local or 'desktop' publishing (then in its infancy) had to stick with vectored art, which could create headlines with no jaggies, perfect curves, and linear graduations and tints of color, color that could be assigned truly from a printer's color catalog like Letraset. By the early 90s, however, the bitmap, fueled by the incredible popularity of Photoshop and by faster PCs with far bigger hard drives, took off. Advances in scanning optics, digital cameras, and software more or less sealed the doom of the vector as anything but a low-memory type and specialty illustration envelope for bitmapped art. It still forms the framework for Flash--but most of the content has long since become bitmapped.

But the humble vertex survived--by moving sideways into Z-dimension, depth, and added a third point and two more vector lines to form a polygon. And thus was born 3D art on the home computer. The art of the polygon, which swiftly became a quadrangle, then a NURBS, then 'UberNURBS', all them mathematical expressions of the atomic structure of virtual forms, models that are created by the user and inhabit imaginary space. Shapes are lathed, extruded, run along 'rails', 'morphed' with other organic looking forms, smoothed, blobbed then added or subtracted to create these models, then they are repeated, mirrored, or merged together to form immensely complex ones--city buildings, cartoon characters, space-ships, flowers, trees, anything that exists or can be imagined by the artist. Next these vertices are compiled together into a mathematical map and textures are assigned; these can either be generated by complex mathematical formulas called 'shaders' or assigned directly to bitmapped files. But these files can do more--they can actually sculpt the model itself with grey-scale height mapped versions of themselves that act directly on the geometry to 'displace' it into new shapes, or simply lend that illusion through 'bumping'. Then the models can be assembled together, 'rigged' so that they walk, talk, fly, rumble around, turn their wheels, and in general emulate all the motions and actions of people, animals, and machines. The art of creating motion is a complex one; for films like the 'Polar Express' millions of dollars were spent to 'capture motion' through tiny sensors attached to actors--which then is mathematically assigned to each lowly vertex and polygon.

Nowadays in a modern studio, all this work has become too complex for a single person to do. Wagster may have begun his career dreaming of doing it all, but as he progresses in his profession, he finds he must begin to specialize a bit. There are now creative directors, conceptual artists, modellers, texturers, lighting and rendering experts, motion-capture experts, motion-matching experts--even clothing and hair specialists. The film 'Final Fantasy' for example, devoted 60% of its render-farm capacity (which was record-breaking at that time) to the physics and rendering of the lead character's hair alone. It's now possible to spend one's career in 3D as a 'digital hair stylist' and do nothing more.

But that's not enough for dreamers like Wagster. Because once you've known the thrill of playing god, of building a scene, a city, a woman, an entire world atom by atom from a void of nothingness, you find yourself mentally returning to it over and over, longing for the simplicity, the beauty of creation, the view through the toyshop window. It's a sensual addiction. In the old days, Wagster would have spent much of his career learning or even writing proprietary software--my first gig in the business was using an application called 'ElectricImage' which was created entirely on the spot for David Cameron's 'Terminator' movies. Remember the nuclear blast that took out LA? That was the birth of ElectricImage, which employed a technique called 'camera-painting' to short-cut the usual painstaking 3D process. But nowadays commercial applications have become so sophisticated that even non-proprietary ones can be used by studios with a minimum of modification--though some, like Pixar's Renderman, are marketed after and not before they have functioned in this way. Wagster has his choice of XSI, which used to be called 'SoftImage' and ran only on SiliconGraphics machines, and Autodesk's Maya and 3DMax empires. Independents and hobbyists sometimes choose Lightwave or Cinema4D. There are also a host of separate modeling applications, from the architectural (AutoCAD, Form-Z) to the more organic (Z-Brush, Truespace). All of these take years of hard work to learn--and beware! Wagster has found that if he's away from any application for too long, he often has to start learning it all over again.

Luckily for the beginner, the enthusiast, the fantasist, the 'low-end' 3D application and specialty markets have kept up with the flagships in terms of complexity and development. If you are not a model-builder, you can choose from a vast array of pre-built and pre-assembled packages, including the Evermotion (http://www.evermotion.org/) and Dosch3D (http://www.doschdesign.com/) collections. There are human models you can shop for, clothing, props, textures, scenes, even pre-defined motions that you can apply to them all. Browse here at http://www.daz3d.com/and you will have a glimpse of the digital toyshop. These are built for a popular posing and motion application called Poser (http://my.smithmicro.com/win/poserpro/index.html). Once despised by professionals, it now boasts comparable features to many high-end applications at a fraction of the time. Though Poser's native renderer is woeful, its animated scenes can be popped in their entirety into my favorite application, the amazing 'Vue XStream', http://www.e-onsoftware.com/,which now interacts as an add-on to the programs Wagster uses in his work. Vue is the premiere environment and landscape application, capable of creating digital 'ecologies' that are then scattered and 'flocked' over realistic terrain. In fact it's even possible to import genuine height-map data from the USGS for total realism.

And so, very gradually, applications that used to be incompatible with each other are groping toward a universal interface. Which means that even the rankest novice, his nose pressed up against the toy-shop pane, can at last take the toys home and learn to play with them...

--