Book Excerpt from Nonlinear/4
By Michael Rubin

"ASPECT RATIO BASICS"

The shape of a frame, whether film or video - analog or digital - is traditionally described by the ratio of its width to its height. This ratio is called the aspect ratio. Television (whether NTSC or PAL) is 4:3, which means it is four units wide by three high. Therefore, if an image is 10 inches long, the height would be 7.5 inches; if it is 640 pixels long, it would be 480 pixels high. Another way of writing this ratio is 1.33:1.

Traditional film as projected on a big screen typically has an aspect ratio of about 16:9 (actually 1.85:1). This is far more "horizontal" than television. Some say film was originally presented in this wide format because it was more like the proscenium of a stage, as you would see it from the audience during an opera, for instance. Regardless, the issue of aspect ratio becomes important when one wants to play a movie on TV, but it's the wrong shape. It's also an issue when a director shoots DV, for example, and hopes to transfer it to film. Or when producing a TV show to be seen by audiences in both NTSC and HDTV. Aspect ratio is a hot topic when the subject of a digital, universal master format is discussed.

Here are the two principal aspect ratios for moving pictures. The darker, more "vertical" rectangle is 4:3; the grey horizontal rectangle is 16:9 (below).

If you put a 16:9 image on a 4:3 display, it will only fit if you shrink it down about 25% and leave blanks along the top and bottom of the screen. Similarly, a 4:3 image will only fit on a 16:9 display if you shrink it down and leave blank spaces along the sides. Thus, there are compromises that must be made in order to convert material from one format to another (below).

Methods
Shrinking the image down and leaving bars at the top and bottom of the screen is called letterbox format; it's the most common way to transfer film to tape, without having to modify the image per se.

But if you decide to avoid the bars and simply crop the image, and ignore any aesthetic considerations, you might perform a center crop-which can be remarkably unsatisfying because it chops off characters and action not in the middle of the screen. In practice, the most common technique, performed in telecine and nonlinear online sessions by Hollywood studio clients, is called pan-and-scan. In this case, the operator moves the 4:3 frame around on the 16:9 image, keeping the most important material in-frame at any given moment.

Finally, you could convert shapes by using an anamorphic compression (meaning that the image is squished along one axis but not the other). But people tend to dislike having their images squished. Research has shown that a squish of 5% is undetectable, and 7% is detectable, though not objectionable. But these manipulations are a far cry from the 25% squish or stretch needed to convert these aspect ratios fully.

For the most part, going the other way (changing a 4:3 image to 16:9) involves variations on these conversion methods - called pillarbox, middle cut, and tilt - and - scan.

With frame conversions in both directions, newer technology makes it possible to mix methods. For example, a little letterbox plus a little center crop, plus a tiny bit of anamorphic squeezing can all be used together as a compromise, better approximating the original source. These blends, as they are called, reduce the necessity of having to make aesthetic judgments for each individual scene. All methods will continue to be needed as long as 4:3 and 16:9 coexist, in their uneasy truce, in the name of convergence.