2. Basics of Image Processing / 2.3 Layers
An image layer contains a raster image.
Image layers are split into color channels. Color channels are logical collections of all pixel parts of the same kind. Bitmaps and gray images therefore consist of only one color channel, the K-channel. "K" means - as in CMYK - "Black".
RGB images consist of 3-color channel, which correspond to the color components of the RGB color space: the R-channel for the red component, the G-channel for the green component and the B-channel for the blue component.
CMYK images have accordingly 4 color channels: the C-, the M-, the Y- and the K-channel.
All kind of images in image layers may (but may not) have an additional channel: the alpha channel, which is also called transparency. The alpha channel determines, how transparent a given pixel is. The alpha channel has the same color depth as the accompanying image: 1 or 2 Bytes per pixel. With 1 Byte the transparency can be controlled in 256 steps. A transparency value of 0 declares a pixel to be complete invisible, and a value of 255 declares a pixel to be complete opaque.
Some functions automatically create an alpha channel (how to create an alpha channel manually, see chapter 4.11). This can increase the memory usage for a bitmap dramatically, because bitmaps only use 1 Bit per pixel, but alpha information 8 Bit in this case. This means: if an alpha channel is created for a bitmap, the image layer will need 9 times the memory it needed before.