Keying (graphics)
From Wikipedia, the free encyclopedia
In graphics and visual effects, keying is an informal term for compositing two full frame images together, by discriminating the visual information into values of color and light.
Contents |
[edit] Chroma key
A chroma key is the removal of a color from one image to reveal another "behind" it.
Tutorial: How to do basic keying in Adobe After Effects
[edit] Luma key
A luma key similarly replaces color from an image which falls into a particular range of brightness. This technique is less controllable, but can be used on graphic elements. It is particularly useful for realistic fire keying.
[edit] Difference key
A difference key uses a background plate of the scene that the foreground object is being keyed out of and the software then assesses the source video and any pixels that don't match the grid are obviously meant to be keyed out.
For example, if your subject is standing in front of a wall, a photo taken from the camera of the same wall, is used. This must be taken from the same camera angle, focus & distance. The software then compares the video to be keyed with the original photo and generates a mask based upon that difference, hence the name.
[edit] Matte key
A matte key uses three images: the two images that will be composited, and a black-and-white third image (called "mask") that dictates the blending of the two, with white revealing one image, black the other, and grey revealing a blend of the two together.
Generally, the "bottom" image is called the beauty, the image that appears on top is the fill and the discriminating element (chroma, luma or matte) is called the key or matte.
[edit] Downstream key
A downstream key (or DSK) is a method of Matte keying, so you use three image or video signals. You have the base signal, where the fill signal is keyed on to, using the key signal to control the opacity of the fill signal. This results in a new signal, that you can use as an input for another DSK. This technique is used in television production, where you show the station name in a corner (DSK 1), and need a name title for a guest (DSK 2), while showing subtitles for translation of his/her speech (DSK 3).
