Animations are always limited by the fact that our monitors (or TVs and VCRs) can only display two-dimensional, flat images. The illusion of smooth motion is created by swapping still images at a faster rate than our eyes can distinguish (latent images), combined with the latency of the display.
It is easy to demonstrate how latent images work. Take a flashlight into a dark room and sweep the beam around quickly. It appears to leave a "streak" or "trail". This is because the flashlight moves too quickly for the eye to resolve the distinct images of its motion.
Similarly, television (and computer animations) uses the same sort of technique. By displaying 30 images (frames) per second, the eye is fooled into believing that it is seeing smooth motion, and not a collection of still images.
There are two basic types of animations:
Full 3-D means that the original scene actually exists in the form of a mesh or wireframe model in the memory of a computer. This allows the user to look at the model from any angle or perspective. Usually the models are created full-scale, especially when a CAD (Computer Aided Design) system is used to create the original geometry to be loaded into the rendering/animation software.
The quality of the model depends on the density of the mesh. The mesh normally consists of triangular and polygonal flat faces, and therefore rounded surfaces are only approximated - they are never truly round. A higher density mesh will more closely approximate the actual shape of the object, and also allow more detail to be included into the model. However, a denser mesh also takes up more space on computer disk and in memory, and will need more rendering time because of the greater number of polygons.
Look at an example of different mesh densities in 3-D models.
An advantage of 3D modeling is the ability to generate "real life" (or very abstract) images of nonexistent objects (like preproduction models for market research). Subsequently engineering drawings and analysis may be generated off the same model.
Another advantage is to create demonstrations which would be too dangerous, too costly, or physically impossible to create: a fly-through of a nuclear reactor in operation, a walk-through of a building not yet built .... You can work on a prototype without touching any of the materials and equipment involved with physically creating the object.
The drawback of 3D animation is the environment: All of the objects usually taken for granted need to be recreated on the computer. An example would be a house on an existing lot. All the trees, shrubs, sidewalks, and so forth have to be modeled as well before a true representation of the new structure can be obtained. The extent of this effort is dictated by the viewpoints required - if you decided to zoom into the sidewalk or grass, the ants and grass would have to be modeled!
After the mesh model is created, hidden line removal can be done. This means that the computer determines which face-outlines can be seen and which will be hidden from the current viewpoint. A hidden line removal gives you a little more insight into the model, since it is less confusing than the normal, transparent view of the mesh model. Complex, transparent models may play tricks on the eye.
Look at mesh and hidden line removal views of the same model.
Once the mesh model has been completed, it may be loaded into a rendering/animation software package. In some cases, the creation and animation/rendering of the model can be done by the same software.
This is described on the photorealism and rendering page.
A good example of this type of animation would be a cartoon. A standard, drawn cartoon consists of a series of still images which were never 3D-based. The characters may be drawn and animated using some animation software, some type of painting software, or may be completely hand-drawn (think "Disney"), and then scanned to digitally bring them to life on a computer display, or photographed to animate them in a real-time movie.
Another use of 2D animations is text added to an existing movie or animation. If the creator of a 3D-based animation wanted to add text (such as credits, subtitles, etc.) to his final product, he/she would have two options. The text could be modeled in 3D and rendered together with the rest of the model (a very time-consuming process), or the text could be added after the rendering of the model was completed (preferred). This is a process known as video-posting. In effect, the 2D (flat) text is laid on top of the 3D rendering to create the final product.
A hybrid solution to the complexity of full 3D model animations is a 3D model placed into a 2D background. By taking a picture of the environment and placing this picture in the scene as the background image, a lot of the tedious work can be circumvented. By matching the camera angles and lighting in the model to the background picture, a very convincing image can be obtained. Doing animations using this method is somewhat tricky, but not impossible. The rest of the procedure is the same as the full 3D-model based animations.
