Alpha channel

​

Users, well acquainted with the image processing basics may skip this section.  

        There are two basic types of image processing and storage – vector and raster. This program utilizes the raster type.  

        Minimal storage unit is a pixel. The pixel is an indivisible rectangular unit, characterized with a certain color.  

        Raster image consists of pixels, located in lines and columns. The more pixels there are per unit, the more detailed is the image. Maximum resolution of a bitmap image is set during its creation and can be increased. If the image scale is increased, pixels turn into large grains.

RGB model is the color representation in a numerical form.

 

         Note: There are many other ways to represent colors, CMYK, for example, but they are not used in this program and therefore we will not consider them.

​

        Color setting requires these three values within the range of 0 - 1. Values {0, 0, 0} correspond to the absolute black color, values {1, 1, 1} – the absolute white. The rest of the colors are combinations of these three values. For instance:

• {1, 0, 0} – red; {0, 1, 0} – green; {1, 0, 0} – blue;

• {1, 1, 0} – yellow; {0.5, 0, 0.5} – violet;

•  {0.5, 0.5, 0.5} – grey.

 

         Conventionally these values are called channels.

         As you can see, with any channel value a pixel will still have full-scale (i.e. non-transparent) color.

         For transparency, the forth channel is added to the existing three.  And as the main three, it ranges 0 - 1.

 

         Note: actually, the channel values ​​range from 0 to 255 with a resolution of 1, which corresponds to one byte. But in its library of standard functions Apple uses values ​​from 0 to 1 with a resolution of 1/256. The author did not wish to” break” the established tradition.

 

         This channel is called the alpha-channel.

​

Note: The term was first introduced by Alvy Ray Smith in the late 1970s.

 

         Thus, the RGBA (Red, Green, Blur, Alpha) color model is used for color setting, which is utilized in this program.

Simply put, if the user sees “alpha” in the part of the name or filter settings, this means transparency. For instance, a simple filter Alpha changes the image transparency from 0 to 1 (i.е. 0 - 100%).

 

         Note: the author hopes that the term “alpha channel” will not confuse novice users (especially after reading this chapter) and will be more understandable to the experienced users than the term “transparency”.