Frames and Motion

Now, lets go back for a minute to that last definition of animate:

To make, design, or produce (a cartoon, for example) so as to create the illusion of motion.

Interesting that the definition writers should choose to throw that word illusion in there, yet entirely accurate. It happens that with just about every form of motion media, only an illusion of motion exists. Heres where we get to the concept of frames.

Virtually all visual animation media uses frames a series of still images shown very rapidly to simulate motion or change. Anything you see on a computer, television, or movie screen is based on frames. This goes back to the earliest days of cartoon animation, where the individual pictures were drawn on sheets of cellophane and became known as cels , and the earliest motion pictures, where a similar technique was used with multiple photographs.

The concept is simple: You show a bunch of images that vary slightly from one to another, and the mind blurs them together as a single, moving image. But why do we insist on calling it an illusion of motion ? If you see a man walk across the room on a movie screen, is that not motion? Of course, its only an image of a man, not the real thing, but thats not why we dont consider it to be real motion.

Remember when I talked about an object being over here and then later over there, and I said it moved through the intervening space? Well, that is real motion. Objects move through space smoothly, not in several jumps . (You quantum physicists in the audience, just be quiet.) But any and all frame-based motion does just that. It doesnt move from spot to spot; it disappears and reappears in another location in the next frame. The faster its moving, the bigger jumps it takes.

If I showed you a picture of a man on the left side of a room, and then a few seconds later another picture of the same man on the right side of the room, youd say I showed you two pictures, not an animation. If I showed you half a dozen pictures of him in the process of crossing the room, youd still say it was a series of individual photos. (See Figure 1-1 for an example of a series of still photographs.) If I presented enough photos fast enough, that wouldnt change the fact that they are still just a whole bunch of still photos, but you would no longer see it that way. Your mind would just take it in as a man moving across the room. It is no more real motion than the original two photos were, but at some point , the mind gives up and buys into the illusion. As a matter of fact, that point has been well researched by the film industry.

Figure 1-1: A series of still photographs by Eadweard Muybridge

Researchers have found that at a rate of 24 frames per second, people are perfectly happy to accept those frames as a single moving image. Go too much slower than that, and the jumpiness gets annoying and starts to break the illusion. And it seems that the human eye cant distinguish frame rates very much higher than that, so going 100 frames per second isnt going to make your movie seem any more realistic (although higher frame rates in programmed animation can result in more responsiveness in interaction).

Frames as records

The whole concept of frames makes three things possible: storage, transmission, and display. You cant really store, transmit, and display a man walking across a room. But you can store a picture, or many. And you can transmit them and display them. Thus you can show that animation almost anywhere , at any time, as long as you have or can receive the stored images and have a way to display them.

Now, lets get a little more general definition of what a frame is. So far, Ive been referring to a frame as a still image or a drawing. Lets call it a record of a system at a specific point in time. That system could be your two-year-old daughter caught mid-grin, and the record would be that image. On the other hand, that system could be a collection of virtual objects, and the record could be their shapes , sizes, colors, positions , and so on at that particular moment in time. Thus, your movie would become not a series of still images, but rather a series of descriptions of images. Instead of just displaying the image, the computer would take that description, create the image from it, and then display it. You can even go a step further by using programmed frames.

Programmed frames

Since you have a computer that can calculate things on the fly, you dont really need a long list of descriptions for your frames. You can cut down to a description of the first frame and some rules on how to build the subsequent frames. So, now the computer is not merely creating an image from a description. Its creating the description first, and then creating and displaying the image.

Consider how much file space you could save using this approach. Images take up a lot of hard disk space and bandwidth. And 24 images per second add up fast. If you can boil that down to one description and a set of rules, youve possibly reduced the file size by a factor of hundreds. Even a very complex set of rules for how the objects should move and react takes up less space than a single medium- sized image. Indeed, one of the first things people notice about scripted animation is just how small it winds up being.

Naturally, there is a trade-off. As your system gets larger and your rules get more complex, the computer must work furiously to calculate the next scene description, and then work overtime to render it. If youre trying to maintain a particular frame rate, that gives the CPU a limited amount of time ( milliseconds ) to think about it. If it cant calculate the scene in time, your frame rate will suffer. On the other hand, image-based animation doesnt care so much about whats in the scene or how complex it is. It just shows the next picture, generally right on time.

Ive used prerendered animation to my advantage at least once. I was putting together a presentation of a number of complex Flash ActionScripted animations. File size was not a problem, since the animations were going to be played from a local machine. But timing was critical, and I had no idea how smoothly the ActionScript would render the images on this unknown, untested computer. So I brought the Flash movies into Director and exported the whole thing as a giant QuickTime movie. Since the movie was now just a series of prerendered images, it didnt really matter anymore how complex they were. As long as the computer was capable of displaying a QuickTime movie, I knew it would do so smoothly. The presentation went off without a hitch.