Apple Shake 4 - Manual (Page 697)

Chapter 24 Keying 697
On the other hand, if your foreground subject has slicked back hair, a crisp suit, and
there are no translucent areas to worry about, you may be able to pull a perfectly
acceptable key.
The following example presents an impeccably shot bluescreen image, recorded using
a high-quality DV camera.
When you key the image and place it over a background (a red field is used in this
example), the result is marred by blocky, aliased-looking edges all around your foreground
subject. Unfortunately, that’s the 4:1:1 color sampling of DV making itself seen.
Why is this happening? In the RGB color space, each pixel in your picture is represented
with a value of Red, Green, and Blue.
When you shoot video—which uses the YUV (or YCrCb) color space, each pixel is
represented by a luminance value (Y) and two chrominance values: red/green
difference (Cr) and blue/green difference (Cb). This is done because green has a higher
luminance value, and is therefore more likely to display artifacts if too much
information is taken away.
In the video world, 4:4:4 color sampling means that for every four pixels in a row, you
get four pixels each for Y, Cr, Cb. This is equivalent to 8-bit RGB. 4:2:2 means that you
get four Y pixels and two each of Cr/Cb for every four pixels in a row. You get less
information in a smaller package, giving you a little more speed and a usually
acceptable loss of detail.
The DV format’s 4:1:1 color sampling means that you get four Y (luminance) pixels and
a single each of Cr/Cb (red/green and blue/green difference). In other words,
luminance can change at every pixel, but color changes only once every four pixels.
If you look at the keyed DV footage, you see that each of those blocks around the edge
of your subject is four pixels wide. A lot of data is saved. Unfortunately, bluescreen
keyers pull keys from color, not luminance; hence the artifacts.
DV bluescreen Key pulled on DV bluescreen