What is the difference between composite, component and digital?

Composite, component and digital . . . three terms that get most people confused. For those of us in production, understanding the difference between composite, component and digital signals is the difference between good images and great images. So what is the difference? Before we can get to that, we need to understand some of the basics of the electronic video signal first.

**DISCLAIMER**

This post may be a little daunting for some.  Here’s a tip: unless you really have a passion for the technical stuff you can skip over the indented or quoted material.  Don’t say I didn’t warn you!

For the sake of simplicity, and because this is a blog about the basics of video – not one on engineering – I am going to break things down extremely easily.  If you are looking to be an engineer . . .  you’re in the wrong place! If you’re an aspiring editor or an aspiring DP . . . this post will help you to understand how to use the technology in front of you, without melting your brain. I’m a creative guy with just enough technical experience to make me dangerous.  If I missed something that you feel is important please let me know by sending me an email.  I would love to make this post more enjoyable for everyone.

**DISCLAIMER**

Video signals and color theory

In video, we are dealing with the additive color of light. This means that when we mix the primary additive colors Green, Red and Blue together we get white light.

With our cameras we do this in reverse. White light, like sunlight for example, is broken down into its component colors in the camera through the use of a beam splitter.

The most important thing to know is that the image that you are shooting and or editing, is made up of a Red Channel a Green channel and a Blue channel.

This little bit of information is needed to understand the information below so make sure that you got it. Got it? Good. Now lets see if I can blow your mind with my crazy interpretation of video science . . . remember, I’m not a scientist but I am wearing a lab coat as I type this . . . for some reason.

The Composite Signal

The simplest of all video signals is the composite signal. When you think of the word composite, you need to envision all of the information from the camera getting mixed into one giant pipe.  This means that the luminance, or difference between light and dark in the image (think black and white picture), and the colors of the image are going to be mixed together and sent down that pipe.

You may see the composite signal expressed as YUV.

• Y represents the luminance value
• U represents the hue value
• V represents the saturation value

U and V combined create the chrominance of the signal – which means the color information. These three signals are combined, or composited, into one.

Let’s say we have a camera connected to a VTR (video tape recorder) using a composite cable. On the recorder we see the signal, however we only have some basic controls over the image, such as brightness, hue and saturation. This is because the signal was mixed together into the same pipe (or cable) and now will always exist as a mixed signal – you will never be able to “un-mix” the signal once it is a composite signal.  You end up losing all of that information – FOREVER!
This is a great resource for monitoring the image, however it is not very good for capturing your final image because you lose valuable information.

The Component Signal

The first type of “component” signal – Y/C or S-Video

Unlike the composite signal the S-Video or Y/C signal separates the information into two components – Y ( luminance) and C (chrominance).

Y/C gets its name from the two components that make it up; Y and C.

• Y refers to luminance (the brightness and darkness of an image)
• C refers to chrominance (the color that makes up the image)

If we change out the composite cable to the S-Video cable in the prior example, the camera is now sending the information of the luminance and the chrominance separately – using two different pipes.

This is important because we are no longer mixing the signal into one giant pipe, we are now using two,  one for the luminance (black and white image) and one for the chrominance (color image).   On the recorder we still have good control over luminance, however, the color information is much more defined because it now exists in a larger “pipe” all by itself.  Unfortunately, we still don’t have control over each individual color.

Since we only have two pipes, we have to put Red Blue and Green into one pipe together.  This means that we’re mixing signals together again.  Not good.  This is not the best situation because, generally, the color information is the information that we need to adjust the most.  Since all of the colors are now mixed, we can never split them apart again.

If you’ve noticed that the S-Video cable is only a single cable, not two, you’re very perceptive.  If you look closely at the connector on an S-Video cable you’ll see 4 pins.  Two pins are the Y and the C signals and two pins are the ground for the Y and the C.  This is considered a consumer level connection and was designed to plug in easily.  If you need to hook up an S-Video type device into a professional video rack, the cable you would use would have the S-video type connector on one end and two BNC connectors on the other.

This is a little better than composite, but it still isn’t the best . . . our color information is still mixed together.

The second type of “component” signal – YPbPr

This is where things get a little tricky, but I promise that I will make sense of everything.  Keep reading and bear with me here, okay?  Remember you can skip the indented stuff if you want.

YPbPr is a type of component signal that uses three cables to send luminance and color information from one device to another.  This is the most common type of component video signal so pay attention. Most likely, when someone asks you about component cables, they are referring to the YPbPr model.

Broken down, the signals are:

• Y – luminance and sync information
• Pb – the difference between blue and luminance (B-Y)
• Pr – the difference between red and luminance (R-Y)

Okay, what the heck does all of that mean?  Let me explain it in detail . . . remember, you can always skip to the “easy” part . . . cheater.

In this signal method, the green color information is derived using the luminance and blue and red information. Because the green color channel is an exact representation of the luminance in the signal, engineers decided to combine the two to save on precious bandwidth but still achieve a high quality signal. Most major video applications use the YPbPr method. Let’s examine the advantages . . .

Unlike a composite signal, the YPbPr signal provides a clean representation of each color channel plus the luminance. Since the luminance is separate from color, we can still achieve good control over it much like the S-Video example. However, what makes this signal type stand out is the fact that the color information is now separated into the three primary colors, Red Green and Blue. Now our recording device is capturing the full output of the camera and (depending on the recording method – which is a whole different post) recording these component signals separately.

Simply stated, the YPbPr signal utilizes three pipes (or cables) to send information from one device to another.  One pipe for the luminance signal (black and white image). One pipe for the Red image and one pipe for the Blue image.  The hardware in the TV or VTR then magically puts these three signals together and “creates” the Green image.  Cool.  This was done to save space in those “pipes” and eventually save space on the tape.   Here’s why it works if you care . . . just remember, every time you skip these technical paragraphs, you make a scientist cry.

Scientists who are smarter than I could ever be, came up with the idea that the color green actually closely resembles the luminance of any given image.  This means that the green channel in a color image looks exactly like a black and white image.   So back when technology didn’t allow for great amounts of data to travel down cables – these nerdy scientists came up with the idea to use the luminance information in combination with the color red and the color blue to interpolate the color green.  This saved HUGE amounts of bandwidth and allowed better images to be captured in smaller spaces.  This is why cameras went from HUGE three-piece units to small camcorder type devices that you put on your shoulder.  Amazing!

One other interesting thing to note – when you are dealing with analog signals you use the notation YPbPr.  However, in the digital world you would use the notation YCbCr.  Just in case you were wondering.

This provides the best possible means to capture your video information . . . well . . . next to digital . . . but we’ll get to that in a minute.

RGB – The third type of “component” signal

Whew! Thank God!  RGB actually works the way you think it works.   It is very similar to the YPbPr signal above but in RGB’s three cables (or pipes if we’re going to stick to that analogy) there is Red information, Green information and Blue information.  These three pipes allow for full color information but they weren’t very efficient for video production years ago.  This was simply because of the bandwidth on the older record methods.  So it ended up in the computer world.  You might remember those old VGA cables . . . those were a variation of an RGB signal.  You’ll probably never encounter true RGB cables . . . but here’s some more info, just in case you do:

As you can tell by the name, RGB separates the Red, Green and Blue signals. It resembles the YPbPr signal in that it keeps the color information separate, however, the signal truly carries each color without using interpolation.  For an RGB signal the Green channel is the color green. There is no interpolation or sub-sampling of the color like in YPbPr model.

Adding sync into the mix, which is essential for video because it locks timing, color and luminance together creating a stable image – RGB has three ways to go:

• RGBS – uses composite sync; the vertical and horizontal signals are together on one cable
• RGBHV – separate sync; the horizontal and vertical are on two separate cables
• RgsB – sync is located on the green channel; a composite synce signal is overlayed on the green cable

The major difference between RGB and YPrPB is that the YPbPr signal requires interpolation to derive all of the color information for the image.  The RGB signal contains all of the color information and the luminance information and requires more bandwidth.  This may seem like the best way to go, but in video applications the YPbPr model is much more efficient. You’ll have to trust me on that.  I don’t have the time (or expertise) to write out the equation that will prove that statement.

If you’re wondering what bandwidth means, think of it this way:  say you have two pipes each 1 foot long.  One that is the diameter of a 2 liter bottle and one that is the diameter of a beer bottle.  Imagine having to fill each bottle with the same amount of liquid.  This is what the YPbPr signal is doing.  It is filling the smaller pipe with the same amount of liquid (or information) as the RGB signal but it is using a computer algorithm to make up the difference.  It is roughly 1/3 less information (no Green channel) but through interpolation it is able to “reconstruct” that information using the Luminance along with Red channel and the Blue channel.

To recap

The composite signal carries a mixture of luminance and chrominance information that can not easily be extracted. The S-Video (or Y/C) signal keeps the luminance and color information separate, however it does not separate the color information into individual colors. The YPbPr signal separates luminance and the color information into three signals,  giving greater control over the luminance and each separated color BUT it uses a magical interpolation method to derive the color green. RGB requires more bandwidth to do the same thing without using interpolation or compression to derive the signal – Red is red, Blue is blue and Green is green.

Is that it? Well, if it was 1999 then I would say yes. But the world has now become digital and so our journey continues.

The Digital Signal

If the above information wasn’t complex enough, now I have to go and use the “D” word. A digital signal can refer to any number of signal types that are encoded into a bitstream and sent down a cable. This can be a USB cable, a Firewire cable, an HDMI cable, an HDSDI cable or even a DVI cable. There are probably more . . . but my brain hurts already. Instead of focusing on the individual types of cables and applications, which will come in a later post, I will just describe the theory of sending a digital signal down a cable.

Now that we have newer and more powerful high definition cameras, we have the ability to encode our signal digitally and send that encoded information in the form of a bit stream – 1’s and 0’s. Since our world is not a digital world (unless we’re actually in the Matrix and I just happened to take the wrong pill) a digital signal is derived from sampling the things around us. This is the future of video and (to some extent film) so make sure that you understand this principle.

Sampling is the reduction of a continuous signal to a discrete signal.  A common example is the conversion of a sound wave (a continuous-time signal) to a sequence of samples (a discrete-time signal).For audio this means sampling the sound waves that are given off by noises. (1)

All things in nature occur in waves.  My favorite waves are the ones you can surf . . . but that doesn’t apply here.  I’m talking about sound waves and wavelengths of light.  In order to make them digital we have to sample these signals, like you did in 9th grade algebra, and turn those “analog” waves into digital 1’s and 0’s.  Scientists use complex equations to do this.  I don’t remember any of them.  You just need to understand that waves (light waves and sound waves) are “sampled” to create a digital representation.  I’ll get to that in another post as well.

Now that we have our digital signals through the miracle of sampling,  we can send these signals to a monitor or recorder and . . . Here’s the great part . . . because the signal is a bit stream, again 1’s and 0’s, there is no chance of quality loss.

In the component YPbPr and RGB signals above, the video signal will degrade over distance. With digital, there is no loss. That’s the beauty of digital signals . . . they’re either there or they’re not.  Keep in mind that there are distance limits on digital signals too,  but you will lose your signal long before you will degrade your image.

This means that you will always have the highest quality signal possible when sending from one device to another.

This is the preferred method of recording because it is like making a clone of the original information. However, because the information is now data we have entered into the world of compression . . . and the complexity of video has now only just begun . . . but that, kids, is a whole different story.

What did we learn today?

If you’re brain doesn’t hurt that much and you’re still hungry to learn more about the subtle nuances of video technology . . . please visit the following sites for more information. If you had enough, then by all means end your journey here. Just remember this little tidbit of knowledge:

1. Digital first for the best quality in recording and monitoring
2. Component second for good quality in recording and monitoring
3. Composite third as a last resort for recording but still a good way to monitor your signal

Now go out there and impress someone with your knowledge!