Quote
Originally posted by Lorenoth:
**what are alpha channels...?
**
(I don't know how much tech talk you know, so pardon me if I talk down to you.)
"Glenn's computer graphics primer"
Background info - if you "don't know much 'bout graphics"
(edit)
More preliminary background material:
Ok, a computer image is made up of pixels (little dots of color on the screen.) The more pixels an image contains, the more detailed it can be, but also the more computer memory is required to describe it. In addition to increasing the number of pixels, a picture can be made "better" (but again, requiring more memory) by describing each pixel with more information. This amount of information is called "bit depth."
The earliest and smallest bit depth is 1 bit. Each pixel is described by 1 bit of data, where a bit is the smallest unit of computer memory, which can either be "on" or "off." Thus, each pixel can be "on", or "off", which usually means black, or white. You don't see a lot of one-bit graphics around these days. 
4 bit and 8 bit color are the next smallest, and they work on similar methods. In each case, the image has attached to it a list of exact colors - 16 colors for 4-bit, and 256 colors for 8-bit. Think of it as paint-by-numbers: each pixel has a value, which tells the system which color it should be. By changing the list of colors, you can completely change the look of the image. (This is what Ferazel's Wand does with its enemies, like the "green goblins", "gray goblins," and "purple goblins". They only had to describe the image once, and just told the computer to draw it with a different group of colors.)
Next, there's 16 bit and 24 bit. These are "true color" formats. These images have three "channels" of data: "red", "green", and "blue." What this means is, out of the 16 bits describing a pixel in a 16-bit image, five are "red" bits, five are "blue" bits, and five are "green" bits. (Yes, that leaves one extra left-over bit. I'm not sure what they do with it. ) Each five-bit color channel describes the amount of that color to add to that pixel, from none to lots. (Remember that red, blue, and green are the additive colors of light; full red plus full green plus full blue makes white; other colors are a mix of different levels of the three.). 24-bit color works the same as 16-bit, but instead of having 5 bits for each channel, it has 8. This additional space lets each channel have a more precise range of color, giving a more exact image.
The actual answer to your question:
Alpha channels come into play with 32-bit images. A 32-bit image is actually a 24-bit image with one more 8-bit channel: the "alpha" channel. While the Red, Green, and Blue channels describe the color of each pixel, the Alpha channel describes its transparency. 8 bits gives 256 levels of transparency, from fully opaque to fully transparent, with a full range of translucency in between. (By contrast, the old graphics format in Coldstone (which you can still use, if you wish) only has 1-bit transparency: either a pixel is completely transparent, or it's completely opaque.)
By using alpha masking, an image can have partially translucent edges which allow it to smoothly blend with its background. It also allows "see-through" windows which tint objects behind them, shadows which darken things underneath them, "light sources" which brighten the area around them, fog to obscure the view, translucent spell effects, ghostly enemies, and many other very cool things.
(Is that clearer?)
------------------
Back in my day, we didn't have any of this here UBB nonsense. It was DiscBoard or nothing, and we liked it!
"In all matters of opinion, our adversaries are insane." - Mark Twain
"The answer is yes or no , depending on the interpretation." - Albert Einstein
(This message has been edited by Glenn (edited 05-31-2001).)
