Video Signal Primer

You may hear us mumbling about "RGB this" and "RGB that" from time to time (and if you've ever been to our sister site GameSX.com you've definately heard it a few times!) but you may wonder why. Well, quite simply, RGB (red, green and blue) has no peer for picture quality. Period. Better than S-Video, better than video, and leaps beyond the picture you get from your RF switch.

Here then, is a beginner's primer on video signals. The focus is on game systems, but the information here can be applied to nearly any kind of video application. We'll start with some background and history, then discuss the five types of video signals, from best to worst with one anomaly at the end. Once you're done, you can have a look at the specifics of getting RGB from your console.

In North America, we got the happy shaft from our consumer electronics companies. The first cameras and televisions recorded and displayed only black and white signals. Standards were introduced for recording, transmitting and displaying this signal. Then, when colour cameras and TVs and recording mechanisms were first being introduced, it was discovered that storing a video signal in its component red, green and blue parts took far too much space. Mechanisms were more complicated, and the space required (and the cost of media) was triple what the old black and white signal was. For this reason alone you won't find RGB VCRs or cameras. Enter at this point the happy scientists who devised ways to compress the colour data. The old black and white signal wasn't entirely packed with data - there were gaps and spaces between each pixel (picture element). This space was put to use - it was decided that colour information would be inserted between the gaps. This was an amazingly efficient solution - older B&W TVs would ignore the extra info, and the new colour TVs would recognize the extra data and display a colour picture.

It wasn't all roses and happy thoughts though! The colour data was distorted and compressed in order for it to fit between the gaps in the B&W signal. It lost some clarity, and changes in television timing would affect the colour - this is why we have tint controls, to compensate for inaccurate timing. Artifacts were introduced - dot crawl was a serious issue, most noticable on a contrasting vertical or horizontal line where 'dots' or pixels could be seen moving along the lines. Colour bleed was another serious issue - where colours would 'leak' or bleed into one another. Most noticable with the colour red, all colours were affected by some amount. It was generally agreed that these flaws weren't significant enough to render the installed B&W TVs useless, and the NTSC video format was formalized. A similar process went on in europe, but the colour encoding and display rate were slightly different, resulting in an incompatible signal called PAL. The French, ever the global citizens, have their own format which is more or less the same as PAL but different only enough to make life difficult for TV manufacturers.

For gaming and computer applications, where the signal is not stored but constantly generated afresh, there's no real reason not to use the separated RGB signal. Because there were no consumer products using RGB video however, no TV manufacturer released a display accepting this signal, and no standard connector was developed. Europe and Asia both have RGB standards, and RGB connectors and televisions have always been available, if not common.

Visual Aids!

Here are some sample pictures of the difference between RGB, s-video and composite video. The pictures were taken at low speed, so the detail is poor, but you can clearly see the colour bleed and colour differences between RGB and the video and s-video pictures.

RGB

S-Video

Composite Video

RGB
As mentioned, RGB is the best. While generating a display, all modern computers deal with graphics internally as individual pixels. Each pixel is defined as having a certain colour value, defined by it's RED, GREEN and BLUE components. These are each assigned a numerical value, which you've probably noticed if you've ever used a paint program. This digital data is as pure as it gets. Every time you convert or 'step down' this signal, you lose some quality. You lose it not only in the conversion process, but in the subsequent cabling as well. Also, a colour TV is an RGB monitor - it has three 'light guns' generating the picture you see, and if your game system drops from RGB to something inferior your TV will have to separate the signal into RGB again. This means more conversions, and more lost picture quality.

Not only are no conversions performed by the game system or the display, but usually the cable will help maintain the signal. You not only get one wire for each colour, but there is RF (radio frequency) shielding to reduce outside interference and associated quality degredation. For a computer generated image, nothing is better. The original image is preserved as completely and accurately as possible, all the way from the source to the display. No quality degrading conversions are performed at either end, and the RGB cable provides more protection than s-video or composite video.

S-Video
In s-video, the RGB components are combined into two wires. One is the old-school black and white signal (luminance, known as 'Y') and the other is the colour data (chrominance, or 'C'). Separating the signals prevents a phenomenon known as crossover where the signals would 'cross over' their boundries and interfere with each other. Each wire again includes its own shielding to prevent outside interference. The signal and picture are very clean, and although noticably clearer than composite video (see the next paragraph) are still not perfect. Dot crawl is virtually eliminated, and colour bleeding is much reduced. See the above pics.

Composite Video
Remember the yellow wire on that weird yellow, red and white cable you got with your Super NES but never used? That was the composite video wire. This uses one wire (with its own shielding) to carry the video information. This is generally a pretty good picture, but depends greatly on the quality of the wire, and the quality of the generating and receiving equipment. Also known as NTSC video, or the V from A/V. Usually accompanied by a red and/or white AUDIO cable. Dotcrawl is a serious problem with composite video, and colour bleed, particularly with the colour red, is very noticable. See the above pics.

RF
This goes into the 'cable' or 'antenna' plug on the back of your TV. This is one wire, shielded, carrying not only the NTSC video information, but also the audio information as well. In the case of the cable coming out of your wall, this one wire contains many (In some cases more than one hundred!) channels. Sound, Video, closed-captioning, alternate audio channels, radio stations, and a lot of RF interference (noise). As you can imagine, crossover is rampant, and the signal is about as fouled as you can imagine. If you've ever visted the cable company's office, you may have seen that their TVs are so much clearer and 'snow' free than yours. This is because their signal has travelled a shorter path and has consequently picked up far less extraneous noise and has had less time for 'crossover' to muck up your shows.

That's why RGB is so beautiful - there's no degradation of the signal by the application of a 40+ year old format. There's little to no signal crossover or interference because the signals are physically separated and protected. The internal workings of the computer are preserved as much as the laws of physics allow all the way to your display. For the discerning gamer, this makes a difference.

There's one other video format, which came into being with the advent of JPG and MPG picture and video compression. This format is used to encode and store Video CDs and DVDs. It is known as...

Colour Difference
or
Component Video
or
ColorStream
Component Video is a bit of a misnomer - RGB is technically also component video, or video whose components are transmitted seperately. Usually, when someone refers to Component Video however, they're referring to Colour-Difference video. Dan's Data explains this three channel (YRB) signal thusly:

The first channel is luminance, (notated Y, the standard abbreviation for intensity). The luminance is the signal’s brightness information only, and includes no colour data. The Y signal by itself gives a black and white picture. The other two channels are called colour difference. They’re notated R-Y and B-Y, and are the difference between red and the luminance and the difference between blue and the luminance, respectively. The colour difference channels can be algebraically recombined with the luminance to give a full colour picture, without having to transmit the green data that, on most video, takes up more bandwidth than the other two colours put together (on average, green is 59% of a video signal).

Now that's not really all there is to it. DVD discs store their video data as colour-difference video, but it's compressed, where the Luminance (Y) channel is mostly left alone and the R-Y and B-Y signals are compressed. In theory this format allows for very little crossover during transmission, but there's a snag: The compression employed does not preserve all of the colour data. The storing of pictures or video in this format trades off some clarity for space saving. It's not always compressed, however - the Playstation 2 for example allows the use of uncompressed component video for the display of games and computer data - but it's DVD output is, of course, compressed. Many modern TVs accept component video but not RGB - which seems odd to me, because component video requires more hardware inside the television for decoding and display. The lack of a standardized RGB connector prohibits the widespread adoption of this standard, unfortunately.

Now you know enough to be a bore at any party - go now, and spread the news! Questions? Email us!

Lawrence.

NEW: Discuss this in the ThinkTank!