![Image](http://users.adelphia.net/~cellstage2/images/video.jpg)
I'm not a video engineer, so my information may be off. From what I understand, color video is made up of 6 components: Red, Green, Blue, Hsync, Vsync, and ground. Analog RGB and VGA (and VGA's variants, UGA, XGA, etc.) separate all of these on different lines, except for sync, which can sometimes be composite (combined) depending on the monitor.
Now, because all those signals are separate and the monitors have dedicated hardware to steer each signal to get it where it needs to be, the picture is perfect. As high quality as it possibly can be given the abilities of the transport (the cord used).
When you start moving up the chart and compressing signals, the picture quality degrades because special hardware inside the TV/monitor needs to "unzip" the signal information so it knows where to put it. The uncompression process isn't very accurate in most cases so lots of signal crossover occurrs, making the picture worse.
Starting from the highest quality, here's a brief description of the video types:
RGB/VGA - The best. If you're a serious gamer, this video type is what you should strive for. Let the casuals have their bleedy, grainy composite and RF signals. VGA is for the most part, a standard 15-pin short dsub, but RGB is not standard, so it may come in many different guises. The pics on the chart are three of the most common. As said above, sometimes the H and V syncs need to be combined, but usually on older monitors like the good ol' Commodores. Also, analog RGB refreshes horizontally at 15khz (interlaced) and VGA at 31khz (and higher, non-interlaced). The TV/monitor must be able to accomodate those rates to get a proper picture.
Component - Slightly compressed on three lines. I'm a bit hazy on the details of component video, so if someone knows more, please set me straight. The color green is determined by an algorithm in the TV/monitor's hardware, but Red and Blue are uncompressed. I don't know how the sync is carried, what the "Y" inputs purpose is, or the difference between Cb/Cr and Pb/Pr.
S-video - Compressed on two lines, one for color data (chroma) and one for brightness (luma). This eliminates "dot-crawl" found in composite and RF video, but some color-bleeding still occurrs. Also known as S-VHS, Luma/Chroma, and Y/C, but I'm not sure what the "Y" stands for.
Composite - Ah, the standard crap TV manufacturers have been pushing on us for decades. All 6 video components are squeezed into one line, and the picture is pretty bad, especially the color red, which bleeds like that elevator in The Shining. Yikes. Dot-crawl is also rampant, which is that flickery, anthill-looking effect on sprites and backgrounds.
RF - Here we are, the lowest of the low. Not only are all the video components crammed into one line, but all sound channels take up space here as well. Coaxial lines can carry a huge amount of pathways (300+ channels for cable TV, for example), but the video coming from your console only gets one (either channel 3 or 4). I'm shocked when I see people playing in RF when they have better, unused video inputs in their TVs. Not only that, but you have to PAY MONEY to get RF switches for your systems these days since they no longer come in the box.
If I left out anything important, please let me know.