In the beginning everything you build is a cube.
- Tamás Varga (http://maxrovat.sns.hu)
Today I have started building characters in 3D for the SPACEPORT game competition this Summer.
I will share some advice as to where to begin... and where to end ;o)
For that I call on the knowledge and expertise of some great digital sculptors in 3D:
Bay Raitt, one of the best 3D modelers of the world. Visit his web presence at Spiraloid for examples like these:
|Example of Raitt's modeling work for Lord of the Rings|
"Considering 3D modeling as a process of sculpting suggests powerful techniques that can simplify difficult 3D tasks, especially the task of modeling characters for animation." Bay Raitt describes the workflow of 3D sculpting as volume, surface, detail.
Alex Alvarez, the author of 'Organic Modeling' (see http://www.thegnomonworkshop.com/tutorials/organic_part1/organic_part1.html).
Dr Brett Stevens, instructor in the Computer Graphics and Virtual Reality (see http://www.tech.port.ac.uk/units/u09887/overview.htm) at the University of Portsmouth.
And finally, the excellent tutorials and references of human anatomy by Peter Ratner (see http://cal.jmu.edu/ratner/tutorials/), . He has written many human modeling tutorials for his books 3-D Human Modeling and Animation, 1 st and 2 nd Editions, Mastering 3D Animation, 1 st and 2 nd Editions and various magazine articles as well as online tutorials.
"When animating emotion, a reliance on only facial animation will have little substance if it is not combined with the relevant posture of the body. The attitude of the body should support the expression on the face." Peter Ratner
CHOOSING BETWEEN TWO POPULAR MODELING TECHNIQUES: patch modeling or subdivision surface modeling
Before starting to model I have a choice between two popular modeling techniques.
I - PATCH/POLYGONAL MODELING
The first technique, patch modeling, creates an object in sections by connecting adjoining parts that share points with their neighbor(s).
Polygon/patch, NURBS, or Subdivision modeling
"One can compare patch modeling to a crazy quilt made up of many patches" - Peter Ratner
One of the drawbacks of polygon modeling is that each edge is a straight line, and hence curved surfaces have to be approximated by breaking them into smaller pieces. This adds computational load and does not always look very realistic.
NOTE: NURBS add even more computational load and therefore they are not feasible for use in time based modeling.
Polygons by themselves form a useful representation system. However, having to produce a polygon set for every object would be incredibly inconvenient. Thus a resource of standard objects, know as the geometric primitives (such as a box, cone, cylinder, plane, pyramid and sphere) are included in all polygon-modelling applications.
Geometric primitives are familiar as objects in the real world such as beach balls, pipes, boxes, doughnuts, and ice cream cones. From these simple objects, very complex structures can be pieced together.
II - SUBDIVISION SURFACE MODELING
The second - and preferred - technique, subdivision (also known as S-Div) surface modeling, uses a smooth modifier on a blocky object, which converts it in to a much more rounded and sometimes “organic looking” object; hence the name organic modeling.
A 'cage' is created that surrounds the subdivision surface. Modifying the cage modifies the underlying surface.
Subdivision within a 'cage'
Surface Subdivision is all about saving you time. In the picture above you can see a very square angular looking shape on the left. This has been created quickly using a few primitive shapes and polygon sweeping. The picture on the right shows the exact same object but with a surface subdivision applied to it. You can see the change immediately - the object has been smoothed and the polygon count increased. [source http://gamespace.thegamecreators.com/?f=tour3]
TIP: Most modelers prefer to create subdivision surface objects as polygonal models first and then convert them to subdivision surfaces later.
Subdivision surfaces are used extensively in character animation because they can define almost any complex smooth surface without the need to stitch patches together.
STAGE ONE: CREATING VOLUME
As subdivision surface modeling has been chosen as the way to model for the characters in the Game, next we will create some volume.
Creating a human head
"A sculptor starts by rapidly defining the volume, then concentrates on refining the surface. In 3D, this method is known as volume modeling, the process of growing, extruding, and manipulating edges, faces and vertices of a polygonal primitive such as a cube. In contrast, when you work with a modeler that manipulates higher-order surfaces, including NURBS, patches, and metaballs, you start by defining an object's surface first, and then modifying its volume. To a traditional sculptor, this approach puts the cart before the horse." [Bay Raitt]
NOTE: A character designed for a game requires a low number of polygons to move well in a game environment. As the computer will have to redraw the scene, and everything in it, 24 to 30 times per second, the larger the number of polygons, the slower the game will play. A model designed to be part of a game may look somewhat boxy, but the application of realistic textures does make up for its geometric simplicity.
Volume modeling lets you focus on building an efficient low resolution form while taking advantage of the power of your software and hardware to generate higher resolution versions using derived surfaces.
STAGE TWO: SMOOTHENING SURFACE
After creating volume you can then extrude verts, faces or edges to start cutting in areas of your surface like nose and eye sockets.
.. using magnet moves to control areas of the model is a much easier way to work than pulling on single cvs or artisan's wanky influence areas.
As for the basics of sculpting, there are only two qualities that truly define your model. The silhouettes and the contour lines.
STAGE THREE: ADDING DETAILS
Think of edge loops as laying a series of rubber bands over an entire body, adding extra bands where you need more detail (e.g., around the mouth and eyes) and fewer where less detail is needed (e.g., the back of the head).
An edge loop closely mimics how real muscles work, and if built correctly, will give you control over contour and silhouette in any position.