What is 3D animation exactly?
What is 3D animation exactly?
Film is illusion: a sequence of still images is played so fast, that the viewer experiences a movement. That principle is being used by animation: by not recording real life (live action), but creating every image, impossible things become possible. Cartoons, claymation, both are examples of this.
In the early eighties the computer arrived as a means to make images. First scientific, later also artistic. And if you can make one image, you can make more. If you adjust these images to one another, it's called animation.
3D animation is about simulating space. The illusion is of objects that are not flat but have depth. To achieve this effect a couple of steps are necessary:
Modeling
Texturing
Scene setup
Lighting
Animation
Rendering
Compositing
Modeling
Everything starts with geometry, or spatial algebra. That might sound familiar: it is part of school math. It is the mathematical way to describe spatial structures. A circle is the line of points that on a flat surface are always at the same distance to the centerpoint. X-squared plus Y-squared equals R-squared: Pythagoras's theorem.
And a ball is the curved surface that in 3 dimensions is always at the same distance to the centerpoint. X-squared plus Y-squared plus Z-squared equals R-squared.
Like this, a mathematical description can be made for any form. And if a computer is good at anything, it's math.
The software I use, makes it very easy for me to formulate these descriptions: I "draw" cubes, balls, cones, and more complicated shapes, where I have three views on screen: top, front and side view. At the same time I can also view the object in a perspective view, where I can rotate it.
All objects are in the end described by points (3 coordinates, for height, width and depth), and the shapes that are formed with these points. Like this a cube has 8 points, and 6 square polygons. Every polygon shares its 4 points with 4 neighbours. It is like those oldfashioned cardboard model kits: a piece of printed cardboard, and you cut out pieces which you glue together to make a three dimensional model.
This process, the building of the object, is called modeling.
Texturing
After the modeling phase comes texturing. That means that all the polygons that together form the object, will each get a color. And that color can also be a wood structure, or marble, or textile, etc. Such a color image is called a texturemap. Furthermore the amount of glossines is determined, reflection, and transparency. And there's lots more possible, to get exactly what you want.
That can be realistic, but for instance also cartoonlike, or stylised in another way.
Scene setup
Positioning all objects is also done in three dimensions. Height (Y), width (X) and depth (Z). Both the objects and the camera. This is done in such a way as to achieve the right setup, where the camera sees exactly that which has to be seen.
Texturing a ball.
Underneath is a simple ball, in the following 9 stages:
1. Standard grey
2. Colored blue
3. Matte specular highlight
4. Sharp Specular highlight
5. Bump texture
6. Image encircling the ball
7. Environment and reflection
8. Transparency and refraction
9. Positioning
Lighting
After that the scene is lit. Virtual lights work about the same as real lights: you put them somewhere, and that's where the light comes from. You can even point them. Without lights the image will be black, with too bright lights the image will be overexposed. But simulated lights offer a lot more flexibility: shadows can be turned on and off, certain objects can be excluded from certain lights, lights can change color, etc. For simple scenes 3 lights usually suffice, but sometimes hundreds are needed to achieve the desired effect.
Animation
Animation is in principle not more difficult than making an extra scene setup. With the difference that you define that this particular setup occurs later in time. E.g.: a cube is on the left at the beginning, and after two seconds it is on the right. The software interpolates the steps in between.
Everything can be animated like this: objects, the camera, the lights, the shape of the objects, the colors, the light intensity, you name it. Every kind of animation demands something different from the animator: a flying logo is much easier to animate than a character that should act out emotions.
Rendering
After all these steps are done, where we actually only did the setup for the animation, the scenes have to be "shot". Because animation consists of a sequence of images, these images must now be constructed. For a movie of 10 seconds, 240 images (or frames) are needed, as a movie plays 24 frames a second.
Every frame is calculated: in fact the workings of a filmcamera are simulated. Enormous amounts of light rays are being calculated, their interaction with the objects (reflection, transparency, color...). This rendering uses very intensive calculations, and is in general very time consuming. It's a good thing the animator is no longer needed for this task, so it can be done at night or over the weekend. For big projects things called render farms are used: a group of computers that only calculate images.
The endresult is a series of digital images, that can be put onto film or video, or can be processed further.
Compositing
With the rendered images a lot can be done in post production. Only think of fitting them into live action.
But the most beautiful thing of compositing 3D animation, is treating all aspects of the image in a different way. With 3D animation, it is possible to separate color, lighting, shadow, gloss (specular reflection), transparency, reflection, edge shape, etc. Every aspect gets its own image sequence. Even objects or parts of objects can be rendered separately. With these sequences in post production the final image can be constructed and finetuned. With all freedoms that come with separate parts.
The outside world, visible through the window, a bit brighter? We slide a slider. Just this one martian greener, and the rest a bit more gloss? We just slide two sliders.
A lot of these tings can be done before rendering. But this way the animation can be rendered once, in stead of many times after each change. And also there are possibilities that are impossible, or very slow, in 3D software.
plz say thanks
Reply With Quote
Bookmarks