User:Zizai/HDTV

From Wikipedia, the free encyclopedia

[edit] Computers, Video Files, and HDTV

by Howard Gilbert Jan 10, 2008

[cite: http://www.yale.edu/tp/HDTV.htm]


Today, every TV station broadcasts two signals on two different channels. One signal is the traditional analog TV signal that has been used since the first commercial TV transmission. The other is a digital signal in which the TV program is encoded as a stream of numbers, essentially as computer data. On Feb. 17. 2009 they will turn off the old analog signal. People with older TV sets will continue to get an analog signal on cable channels, and the Federal government is offering coupons that discount the cost of an external tuner that converts the new digital signal into an old analog format. Of course, existing game consoles and DVD players will continue to generate analog signals, so it will be quite some time before TV becomes an entirely digital medium.

Digital TV transmits a stream of data. For less than $100 you can add an adapter to your computer that will receive this stream, and programs to view or record the TV shows contained in it. Data is pretty much the same whether it comes from the Internet, a CD, a USB device, or a TV broadcast. Programs can read it, filter it, transform it, or write it to disk.

The bits and bytes have to be organized into units of meaning. They are called packets or records. Data on a hard drive is some number of 512 byte sectors. Data on a CD or DVD are some number of 2048 byte records. Digital TV transmits data in the unusual packet size of 188 bytes. Because each record has administrative fields to identify it, this small packet size is not particularly efficient. However, when a signal is broadcast over radio frequencies and is received by an antenna, there is always the possibility of interference from other sources or reflection when the signal bounces off a nearby building or hill and the echo arrives at the antenna with a slight delay. A larger packet size would mean that more data would be lost if something goes wrong. The smaller packet size is accompanied by error detection and recovery through the redundant retransmission of additional copies of information.

However, when the data is recorded to a hard drive or is distributed on a DVD or Blu-Ray disk, there is no reason to worry about interference. So if you watch Heroes live on NBC, it is transmitted in 188 byte units. However, if you buy the DVD or HD DVD disk, the same data will be formatted in more efficient 2048 byte records.
Digital TV is broadcast over the same frequencies as traditional TV, and it can be picked up with the same Radio Shack rooftop antenna or rabbit ears. The frequency that for decades we have associated with one analog TV "channel" can now carry a stream of 20 million bits or a little more than 2 megabytes per second. Most of it will be used to broadcast one High Definition and one old standard definition TV program. The rest could be used to broadcast more programming, or it could be used to send a stock ticker tape, traffic report, text news stories, even software updates. TV stations negotiated with the FCC so that they could put any kind of data they wanted onto the unused packets.
A cable TV system is immune to interference. It transmits digital data without the error correction and redundancy, so it can get twice as much data or 40 megabits per second on each of its 130 TV channels. Comcast recently demonstrated a system where it used 4 TV channels to create an Internet data feed of 160 megabits per second.

Verizon has a system called FiOS where they run a fiber optic cable to each house. This provides nearly unlimited bandwidth. AT&T has an alternate system called U-Verse that is a little cheaper. They run fiber to a big box in your neighborhood, but then they run traditional copper wire pairs from the box to your house. On those copper wires they run a heavy duty version of DSL with 20 megabits per second (20 Mb/s or 30 or 40 in some test systems). This 20 megabit connection then carries internet traffic, TV, and other content to your house.

The Comcast cable carries 135 channels simultaneously to your living room. You can hook the cable directly to a TV set and use the remote to decide which channel you want to see. However, U-verse only has 20 Mb/s of DSL bandwidth, which is just barely enough for one or two channels, and then only if they are compressed using the best compression technology currently available (advanced MPEG 4). So in the U-Verse system, you send a signal to the box in the neighborhood indicating what channel you want to watch, and it selects that signal from the fiber it is receiving and sends just that one program to you over your copper wires. Every house in the neighborhood gets its personal selection of program(s) to watch. While Comcast is a TV system that happens to carry data, U-Verse is an Internet protocol network that just happens to carry TV data.

With the old analog system, the quality of a TV picture depended on electronic components like the "comb filter", and circuits to detect and remove ghost images. You needed to spend a lot of money to get a really good set, and you needed a degree in electronic engineering to understand it.

With digital TV, the network or cable system transmits a sequence of numbers. For $100 or less you can get a device that receives these numbers and passes them on to the computer. If you can get the numbers, then you have 100% of the TV information. No matter how much money you spent, you cannot improve your reception because you have everything there is exactly as it was sent. From that point on, everything else is some kind of computer processing. You can buy a black box to do it for you in your living room, or you can download some software and do it yourself on your computer.

You can watch High Definition TV on your computer, or connect your laptop to the High Definition TV set on your wall. TV sets and computer monitors have become two different versions of the same thing. A TV set will be brighter but have a lower resolution because it is designed to be viewed from across the room instead of from two feet away. However, history conspired to create quite different rules for analog input cables. TV sets started black and white, then added color. So TV cables always have a black and white signal, and then additional color signals. PC video was standardized in 1987 instead of 1947, so it is based on three primary colors. All digital cables for TV and computers are based on versions of the same standard.

Why Digital? The first analog recording device was the Edison phonograph. Sound waves in the air vibrate a needle that makes an impression in a wax cylinder. Later the cylinder can be used to vibrate another needle that in turn vibrates a horn that recreates the original sound. Then Bell learned how to replace the wax cylinder with an electric wire to create the telephone. Others replaced the wire with radio waves to produce broadcast audio. Finally, Philo Farnsworth in 1927 used radio to modulate a rapidly moving beam of electrons striking a cathode ray tube to produce the first TV.

Each time an analog signal is received, recorded, and retransmitted, the process is imperfect and adds some noise or distortion. In the 1950's long distance phone system, a voice call became harder to understand as the signal was amplified again and again as it travelled from coast to coast or under the sea. A copy of a video or audio tape is not as good as the original.

Analog information can be digitized. To digitize sound, replace the Edison recording needle with a sensor. Let it vibrate as before, but instead of using it to create waves in a surface, have a computer record the position of the sensor as a sequence of numbers. Write these numbers anywhere a computer can store data. Later on, you can use the sequence of numbers to drive the position of a rod that drives the cones in a speaker system. With video, use a digital camera system to record numbers that represent the color of every dot in the picture.

The problem is that the numbers in the digital data take up a lot more room on any recording medium. A magnetic tape can be used to record sound or data. If you digitize the sound, the data takes up a lot more tape than you would have used to store an analog tape recording of the same sound. Two things fix this problem. Modern computer technology allows this data to be compressed. When the phone company began to digitize voice telephone calls, they needed 64,000 bits per second to transmit the data. Cell phones compress voice down to less than 10% of that original size. Technology also allows an awful lot of digital data to be transmitted over the same physical connection that used to carry old analog data. The two copper wires that carry one plain old telephone call can also (even at the same time) transmit 20 to 40 million bits per second of data. Fiber optic cable bumps that into the tens of billions of bits per second.

Unlike analog systems, digital data can be received and retransmitted precisely. There is no loss of data, and no errors are introduced. A digital phone call around the world sounds the same as a phone call across the street. If you record the data from a digital TV broadcast onto a computer disk, you make an exact copy. You can view the program the next data or next week, and the picture quality will be identical to what you would have seen viewing the broadcast live.

Generations Individuals can decide to replace any electronic device in our living room. The VHS player was replaced with a standard DVD player, which in turn is being replaced with some sort of HD player. Broadcast TV and radio, however, are controlled by the FCC. Those standards have to change slowly over time so that all the TV and radio stations can buy and test equipment and all consumers can prepare. The Feb. 2009 cut over to digital TV has been under way for a decade, and as a result it is already based on obsolete technology.

It may be hard to remember, but the first Pentium computer chips were not fast enough to play a DVD movie. You needed at least a Pentium II running at 233 Megahertz before it was possible for the CPU to decode MPEG 2 video. As the name suggests, there was an older MPEG 1 video compression that you could use on slower CPU chips, but it did not have very good quality. Well compression technology has advanced to use up all the CPU power on each new generation of processors. There are three standards in wide use today that are different levels of MPEG

MPEG 2 is used on DVDs and, because the FCC decided on the standard a decade ago, also in Digital broadcast TV. Many analog TV tuner cards for a PC have a hardware chip that digitizes and compresses TV programs to MPEG 2 in the tuner itself before sending the data to the PC. If the tuner card does not do its own compression, every CPU chip made in the last five years has enough power to do MPEG 2 compression of a digitized standard definition TV program in real time (as the program is being received over cable or antenna). For all these reasons, MPEG 2 is currently the most popular file format for recorded TV. There is a simple version of the new MPEG 4 compression standard that is better than MPEG 2 and only requires a little more processing power. Versions of this go under the name of "Divx" or provide the format for WMV files that Windows Media player uses. You can use this type of MPEG 4 to get a slightly smaller file size or slightly better quality than MPEG 2. However, this standard is not used on any commercially available movie disks, and only one computer device was ever produced that supported hardware compression of analog TV programs to this type of MPEG 4. It may be used inside a black box DVR or be an option in your computer software, but this format appears to have missed its window of opportunity to become important. There is an Advanced level of MPEG 4 that is used to compress audio and video for iPods and other portable players. It is also used for movies distributed on Blu-Ray or HD DVD disks. It is better than the simple version of MPEG 4, but requires a lot more processing of the data. At this time, a dual core 3 GHz Intel CPU chip can barely keep up with the CPU requirements for playing a Blu-Ray disk. To recompress a digital TV program broadcast in MPEG 2 to the advanced MPEG 4 format requires 6 or 8 hours of processing time for a one hour TV program. In the long run, MPEG 2 will remain because there are a lot of DVDs and because it is the FCC digital TV standared. Advanced MPEG 4 will remain because there are a lot of Blu-Ray disks and iPods and because it is the best technology currently available. Everything else will become a historical footnote. Within three or four years, PCs will have the processing power to do advanced MPEG 4 compression in real time.

Size or CPU Every consumer knows that if you are willing to pay more money you can get a product of higher quality. Well in video technology the cost is measured in bandwidth and file size on the one side, or else CPU processing on the other. If you use a higher standard and more CPU you can either shrink the file size or else improve the quality of the picture.

A standard DVD movie is recorded in MPEG 2 and uses 4-7 megabits of data per second. As the master version of the DVD is being created, computers or the technicians monitoring the process use a higher data rate when there is a lot of action on the screen, but then turn down the data rate when not much is changing on the screen. Broadcast network High Definition TV programs also use MPEG 2 and transmit data around 15 megabits per second. A broadcast digital TV channel provides slightly less than 20 megabits per second of data transmission, and stations have to save some of that to also broadcast a standard definition version of the same program for people with older sets. Since an HD screen has 4 times as many pixel dots, but the data transmission is only twice that of the fastest speeds on a DVD, the network broadcast TV shows have to use more aggressive compression on their pictures than movie studios use. When the players are in a huddle and not much is happening, the picture is sharp and detailed. When the ball is snapped and everyone is running around on the screen, the image of each player is less detailed than it was when they were standing still. However, you will only notice this if you take a snapshot of the screen, because your eye cannot really resolve detail on a fast moving object. A Blu-Ray or HD DVD movie uses advanced MPEG 4 and had data recorded at 30 megabits per second. It starts out with four times the data rate of a standard definition DVD, and then it jumps two generations in advanced data compression technology to produce an even better picture. Some people buy an HD TV set and then use it to watch standard definition programs. Before you feel superior, you should realize that if the only HD programming you have seen on your set is broadcast network shows like Football, then you still haven't seen what your TV can really do. You have to hook up a Blu-Ray or HD DVD player to get a data stream with the maximum possible resolution.

The problem for broadcast TV is that the 20 megabit per second capability of the TV frequency cannot be expanded. Broadcast picture quality has hit a brick wall that is not present in cable or fiber optic systems. Of course you can take the same episode of Ugly Betty and broadcast it at 15 Mb/s over the air and at 30 Mb/s (and maybe using advanced MPEG 4) over some other medium. The networks have to record the show at a better quality to eventually sell Blu-Ray disks, and they could offer the higher quality feed to some future generation of Cable TV system.

Why the PC? Because it's data. You can certainly buy consumer electronic devices with remote controls that don't look like a PC, but ultimately every device that processes digital TV signals is some kind of computer. You can buy a TV, a DVR, a Blu-Ray player, a DVD recorder, and so on, or for under $100 you can plug a Hauppauge HVR-950Q USB "stick" into your laptop or desktop computer that can receive digital or analog TV signals from broadcast or cable TV sources. Once you get the data into the PC, you record it to hard disk, play it back, or burn it to a DVD just like any other data file.

However, DVD and Blu-Ray movie disks come with an encryption system that is supposed to prevent copying. Now I don't want to encourage people to do illegal things, but it would require a certain amount of willful blindness not to notice that there are programs available to defeat these protections. I do not condone piracy, but the movie studio systems have made things unreasonably difficult to use.

The licensed Blu-Ray or HD DVD players will not play the movie over a digital connection to a monitor unless the connection supports HDCP. There is a problem if your monitor supports HDCP, and your video card supports HDCP, but they will not agree to use HDCP with each other. Then your option is to either fall back to an analog VGA connector or else to use one of the available software packages to make the player believe the disk is unprotected and therefore can be displayed on the monitor you have.

If you buy black box consumer products, then this article has little to tell you. You plug the power cord into the wall and press the on-off button. The rest you learn by reading the manual. However, once the data is on a computer hard drive then there are things you ought to know about devices and standards and file formats. For those who want to know how it works, this article will explain things.

Analog Painting Imagine you had to paint a billboard on the side of the road. To do this, a spray paint device moves left to right across the board spraying several different colors in a line one inch wide. You control valves that determine how much paint of each color is being sprayed on the current location. You can open each valve all the way, or shut off that one color entirely, or position the valve anywhere in the middle. When the spray head gets to the right end of the billboard, the paint shuts off, the spray head moves back to the left margin just under the line that was just painted, and the process starts again left to right painting the next line.

This is the way the TV screen generates a picture, except that the TV tube sprays electrons instead of paint. It is an analog system because the valve controlling the amount of paint moves smoothly from all the way off to all the way on and can generate any amount in between.

All that would be needed to produce a true TV-like system would be to connect the paint valves to electric regulators controlled remotely by a radio signal. Fifty years ago, an electronically controlled “valve” for electrons was available as a vacuum tube, and it was used to build the first TV sets. Today, computer chips make it possible to use other designs.

Digital Painting Computers operate on numbers. To digitize the previous example, replace the analog valve with a system that delivers paint in an amount represented by a number. Each spot on the billboard has a set of numbers that represent the exact color for that location.

The entire board requires a massive set of numbers. Fortunately, there are large sections with the same color or pattern, so the data can be aggressively compressed. Using what is now relatively old MPEG 2 technology, the picture from 4 to 10 TV channels can be squeezed into the frequency now used by a single analog channel.

In the 1960s vacuum tubes were replaced by transistors, and in the ‘70s Sony replaced the round dots on the TV screen with rectangular spots. Otherwise, analog systems have not changed much. Digital technology, however, changes rapidly with increased speed of microprocessors. With today’s faster CPUs, it is possible to get better quality or smaller files with MPEG 4.

Consumers may be willing to buy a new computer every three years, but they expect a TV or stereo to last for decades. TV stations are not willing to invest millions in new broadcast equipment every time someone comes up with a better compression algorithm. Therefore, public standards for consumer equipment change much less frequently than the format of computer media files.

Different Types of “Digital TV”? Generically, Digital TV is any system that transmits the TV picture as a series of numbers. There is a common set of core technologies, but they can be deployed in several different packages. You should not confused completely different system simply because both use the words “digital” and “TV” at the same time.

A “digital versatile disk” (DVD) is an updated version of the CD. The CD holds 700 megabytes of data. A manufactured DVD can hold up to 9000 megabytes of data to hold a movie in MPEG 2 format, plus several sound tracks and extras. A DVD movie is "Standard Definition" which means that it is designed for traditional TV sets. When displayed on a new HD TV, the DVD displays 480 lines of 720 dots per line. “Digital cable” transmits a stream of 40 million bits per second (using a technique called "QAM256") over each of the available channels from the 135 frequencies previously used for analog TV. This raw data stream can then be used to deliver two 15 Mb/s MPEG 2 broadcast HD channels or up to 10 standard definition 4 Mb/s MPEG 2 channels. Digital Cable Box: For over a decade, cable TV systems have offered a service they call "digital cable" that allows dozens of premium movie channels, plus pay per view, plus sports packages, to be received by a digital cable set top box. This stuff is all "encrypted" content, which means that it can only be received by the cable box or by devices with a special CableCard rented from the cable company. The cable company likes this service because each analog channel can carry up to 10 pay per view or premium channels bringing extra revenue. "Clear QAM": Next year the TV networks and stations will stop broadcasting analog signals. However, to simplify the conversion for people with old TV sets, cable companies will still be required to receive the new digital programs, convert them to old analog format, and continue to transmit them over channels 2, 3, 4, etc. They are also required to transmit the new HD network signal (at two programs per old channel frequency) and maybe the digital standard definition version of the program. This free-to-air TV programming cannot be encrypted, so it can be picked up by any TV set or PC tuner card that is designed to receive unencrypted QAM cable TV signals, or what is typically referred to as "clear QAM". Details will vary from one cable system to another, but in New Haven, CT (the author's home town) the Comcast system puts the CW and PBS on cable channel 89, NBC and ABC on 93, and Fox and CBS on 94. Then a grab bag of out of town (New York City), public access, home shopping, and CSPAN standard definition free programming is dumped on channels 122 and 123. Satellite TV systems (DirectTV) transmit a digital signal in a format very similar to that of digital cable, but it is all encrypted. Therefore, you need a separate satellite receiver device for every TV set or computer adapter card. Digital Broadcast TV transmits data at 20 Million bits per second on the same frequencies previously used for analog transmission. You pick this up with the same Radio Shack rooftop antenna (or rabbit ears) that you used for old broadcast TV. Typically they broadcast one High Definition program and then one standard definition copy of the same program. Blu-Ray and HD DVD are new versions of the DVD/CD media. Blu-Ray holds 25 Gigabytes of data per layer, while HD DVD holds 15 Gigabytes per layer. Both support HD movies typically compressed in some version of advanced MPEG 4. What does all this mean? Unless you understand the terms and read carefully, you won't know what you are getting and what is best.

The Source Start with a good, old fashioned rooftop TV antenna from Radio Shack. You will be able to pick up your current analog channels and new digital channels transmitted in 8VSB. Unfortunately, each TV station will broadcast both types of signal from their current location, so if you get a really weak signal now in analog you probably will get nothing usable in digital mode. You can split the signal and connect up any number of devices. None of the content is currently encrypted.

Start with Comcast (or any other brand) Cable TV. They broadcast analog TV on channels 2 to somewhere around 60. You can pick them up with an old fashioned TV and with any computer TV tuner, Tivo receiver, DVD recorder, or any other device. Then somewhere up in the higher channel numbers they will start sneaking in digital QAM channels. Standard Definition digital cable channels can be viewed on any TV set by connecting one of the outputs on the digital cable box to your TV set. HD digital cable can be viewed by connecting a HD TV set to the component or HDMI output on the back of the digital cable box. If you have a TV that supports QAM input, you can connect the cable directly to the TV and view free broadcast programs, including HD network content. You can split the signal and connect up any number of devices, but only the cable box will see the encrypted content.

Currently US satellite TV is all based on vendor specific conventions. The only way to receive the signal is with a proprietary box. The only way to record is to use their DVR. This is not particularly interesting because there is only one option.

The Set All high definition TV content is digital data. Today all high definition TV sets are flat panel or projection units. Those are two separate trends that just happen to have occurred at the same time. When HD TV first appeared as an option, the first generation of HD TV sets included big picture tube units. This should not be surprising since CRT computer monitors certainly had enough resolution to display HD pictures. Nevertheless, you must not assume that "digital TV" and "flat panel monitor" are the same thing. These are two different changes that just happened at the same time.

In the future, all TV sets and computer monitors will be digital devices that receive a stream of computer data over digital DVI, HDMI, or DisplayPort cables. However, at this moment all HD TV sets and all devices that connect to them also support one or both of the the two legacy analog connections. There is the 15 pin VGA (MD15) connector used on computer monitors since 1987 that carries separate analog Red, Green, and Blue signals for the three primary colors. Alternately there are "component TV" cables that carry a black and white signal ("Y") and two color components ("Pr" and "Pb").

So if you are trying to figure out how to hook things up, it makes sense to start at the ends and work toward the middle. The TV/monitor will have a bunch of different connections on its back. Some of them (the yellow component cable and S-Video) are old standard definition TV standards that you will probably only use to hook up an old VHS tape unit or early DVD drive. The important ones are the digital (DVI, HDMI, DisplayPort) and the analog (VGA, Y Pb Pr).

Receiver When TV started out there were only the VHF channels 2-12. Then higher numbered UHF channels were added from another frequency range. Cable TV had no reason to maintain this split, so it created 135 adjacent frequencies. That is why you have to tell a TV set if it is connected to an antenna or cable, so it knows which frequencies to scan. Digital TV reuses the exact same broadcast and cable frequencies, and just puts a different signal on them.

All TV sets and DVRs can receive an analog signal. Newer ones can also receive the digital signal, and better ones can receive "clear QAM" from the cable company. There are four types of PC adapter cards: those that receive only analog, those that receive only digital, hybrids that receive one analog and one digital signal from two different cables, and combos that receive either analog or digital signals from a single cable under program control.

However, to receive encrypted premium cable TV channels you need a "set top box" or CableCard. A digital HD set top box from your cable company will always be able to send an analog Y Pb Pr component signal to the TV set, and some can send digital DVI or HDMI signals.

The FCC required cable companies to provide a separate device called a CableCard that can do the premium channel decryption. Prices vary from place to place. Here Comcast will loan each subscriber one CableCard free of charge (if you have a device to plug it into) and then will rent you additional cable cards for less than the monthly rental of an entire set top box. CableCards can plug into a Tivo Series 3 DVR device or they can be plugged into the back of certain models of TV set.

Satellite TV always requires an external set to box tuner.

Movie studios do not want precise digital high definition pictures to exist in any electronic form where they can be recorded by an external device. Therefore, when a Sony Playstation 3, Blu-Ray or HD DVD player, or PC with a Blu-Ray or HD DVD drive is connected to a HD TV or computer monitor, the firmware or software requires that any digital connection between the screen and the box be encrypted using a technique called "HDCP". If the TV or monitor does not support HDCP, then the game console, player, or computer can connect to it using one of the two old analog connectors (VGA or component Y Pb Pr).

So what about buying a new TV? For the most part you buy a TV based on what it looks like. Is it bright? Are the colors sharp? That part is in the eye of the beholder. This article is about the stuff you don't see.

In chronological order of development, the technologies that will be discussed are:

Analog TV developed black-and-white standards in the 40's, which were then expanded to support color in the 50's. The VGA family of monitors, developed by IBM in 1987, which remains the standard analog (CRT) computer display. Digital video file formats (MPEG, QuickTime, Real, and MS) that developed over time and became particularly important with the success of DVDs. Digital TV (Enhanced and High Definition), which is just becoming widely available and is supposed to replace analog broadcast TV by the end of the decade. To make the points clearly, it is necessary to quote numbers like how many picture lines are displayed on a screen. There are several standards used around the world. This paper will quote the numbers for the US standard (called "NTSC"). Rather than confusing the reader with the numbers from all the other standards, it would make more sense to develop other versions of the paper targeted to other countries.

Show What We Can See Everyone knows there are sounds that a dog can hear but people can't. Radio, CD, and MP3 standards reproduce only the sounds we can hear. TV standards similarly concentrate on the picture elements we can see. To understand TV and computer monitors, it is important to first review the physics of light and the physiology of the human eye.

Color Light is carried by particles called photons. Each photon has a wavelength. The range of wavelengths that the eye can see is called the visible spectrum. Within this range, light with a longer wavelength and therefore lower frequency is red, while light with shorter wavelengths and higher frequency is blue.

Light passes through the lens of the eye and is focused on the retina. There are two types of receptors on the retina called "rods" and "cones". There are a large number of rods that provide night vision and peripheral vision, but they cannot distinguish colors. Cones are in the center of the eye at the point of direct focus. They provide more precise vision and color discrimination.

There are three different groups of cone cells. Each is sensitive to a range of light centered at a slightly different wavelength. Historically they have been called "Red", "Green", and "Blue" cones although doctors who have actually measured sensitivity note that the "Red" cones are actually most sensitive to orange photons, while "Blue" cones are most sensitive to purplish photons.

Photons enter the eye, are focused by the lens, hit the cones in the retina, and trigger a signal to the brain. Yellow light will trigger some of the Green cones and some of the Red cones. The brain gets these signals, compares the amount of signals from each type of cone, and decides that the light must be Yellow.

Only a laser produces photons that have a single pure wavelength. Ordinary light has a mixture of photons of different colors. A prism can split the light into its component colors, but evolution was not sophisticated enough to give our eye a spectrometer. Instead, our eye and brain depends on differences in the response of the different types of cones to approximate the color.

This means that the eye therefore cannot distinguish between a pure yellow light that triggers some Green rods and some Red rods and a light that contains a mixture of some pure green photons and some pure red photons. Both cases produce the same response in the two types of cones, the brain gets the same signal, and it "sees" the same color.

White light contains approximately the same amount of all colors. Black is what we see when there is no light at all. When you are just distinguishing black and white (actually shades of grey), the rods of the eye can also participate. Put up two charts. One has increasingly thin lines of black and white. One has the same lines in colors. The eye can distinguish much finder detail in black and white than it can differences in color.

The old analog TV standard recognizes this. It starts with a strong black and white picture, then a much weaker color signal. Modern DVD and other digital TV technology applies a much higher level of compression to the color than to the black and white part of the picture. This saves storage space, since we simply can't see the very fine color detail. You can think of this as an algorithm that changes the brightness of every dot on the screen, but changes the color only every two dots.

Refresh The eye is similarly imprecise about changes in the image over very short periods of time. The invention of the motion picture was made possible by the observation that when people see a continuous stream of still pictures they perceive it as continuous motion. Movies show 24 separate images every second. The TV did not have to show images faster than that, but 1/24th of a second turned out to be a difficult period of time for early electronic devices to support.

The US electric system of "alternating current" (AC) changes the direction of current 60 times a second. The early electronic equipment had a strong 60 cycle internal operation, and any attempt to use a competing rate produced problems. So the US TV signals adopted a standard where they would rewrite the content of the screen 60 times a second. Since the eye cannot distinguish changes that fast, pairs of two successive screen rewrites combine to form a single “frame” and the frames change 30 times a second.

In Europe, however, electricity changes direction 50 times a second. Since US and Europeans TVs operate on roughly the same range of radio frequencies, and each TV channel can carry the same amount of picture information, a European set that only has to display 25 frames per second can have 1/5 more detail on the screen than a US set that has to display 30 frames per second. So European TV has more lines and a sharper picture.

Today’s monitors can refresh at any rate and are not slaved to the frequency of the electrical system. Analog computer TV cards from companies like Hauppauge can receive both US and European TV traditional broadcast and cable signals with just a change in configuration. HDTV has the same screen resolution and refresh rate throughout the world. However, for reasons that remain controversial, the US decided to adopt a different “modulation” technique for digital broadcast over the air to antennas (8VSB) than is used by other countries (DVB-T), and currently you have to buy different digital TV cards to operate in the US than in other parts of the world.

Both TV sets and CRT computer monitors generate their picture with a stream of electrons directed at the back of the TV screen. The electrons hit a phosphorescent coating that then glows for a while generating the light. Different coatings fade at different rates. This is called “persistence”. If the glow fades away before the screen is rewritten, then the picture will flicker. Not everyone sees the flicker, but a small percentage of the population can detect it and finds it bothersome.

TV sets solve the problem by selecting phosphors with a persistence exactly tuned to the 50 or 60 cycle per second refresh rate. Computer monitors have a variable refresh rate. While they can be set to the same 60 Hz refresh rate of a TV set, CRT computer monitors operate better at rates of 72 to 85 Hz where no person can see a flicker. Flat panel LCD panels, however, operate on an entirely different flicker-free technique and are best operated at the old 60 Hz rate.

Interlace The eye is also flexible in the way it handles detail. There are only 3000 cones at the focus of the eye. This contrasts to the millions of pixels in in a $200 digital camera. To see the entire picture, the eye has to move around focusing for a fraction of a second on specific points of interest. If someone makes a movie with a handheld camcorder, the picture bounces around and is difficult to watch. The eye is always bouncing around much more rapidly, but the brain filters out the movement. We see a fixed image even though the eye pieces the image together by rapidly focusing on different points.

TV standards take advantage of this with a technique called "interlacing". The TV camera captures one picture with twice as many lines of detail as the TV set can display. In one cycle it transmits every other (say "odd numbered") line. Then in the next cycle it transmits the lines skipped before (the "even numbered") lines. On the TV screen the contents of every line flickers, first showing one set of lines and then the other. The flicker happens 60 times a second in the US (50 times a second in Europe). Since the eye is not particularly good at distinguishing changes smaller than 1/24 th of a second, we don't really see the flicker. The brain sees some data from one set of lines and some data from the other. It attributes the difference to something like eye movement and merges the information from the two sets of lines together. Therefore, we "see" a picture with information from twice as many lines as the TV set can actually display.

However, the "interlace" trick only worked on a CRT (tube type) TV set. LCD and Plasma flat panel TV sets do not flicker fast enough to make the interlace work. However, modern flat panel sets have a lot more lines and dots than the old CRT TV sets. So if a TV source is generating interlaced data, the electronics in the set top box, tuner, or TV set has to convert it and display it properly on a modern, higher resolution screen.

TV starts with the CRT The Cathode Ray Tube was invented more than 100 years ago. A vacuum tube has an electrode at the back, a focusing magnet ring, a metal mesh near the front, and a front surface coated with a phosphorescent material. The electrode is given a negative charge (with surplus electrons) while the front screen is given a positive charge (a shortage of electrons). Since there is no air in the tube, electrons want to jump from the negative electrode in the back toward the positively charged mesh at the front. Some hit the metal mesh and are absorbed. However, there are holes in the mesh and if an electron happens to pass through a hole it continues till it hits the phosphor coating on the inside of the front glass screen. The phosphor glows when it is hit by electrons.

The electrode is surrounded by a cover that only allows electrons to escape in a narrow beam. The magnet ring near the electrode can bend the beam path. In a TV, the beam starts in the upper left corner and sweeps left to right across the top of the screen. Then there is a pause while the magnet reorients and a second left to right sweep occurs just under the line formed by the previous sweep. The process continues until the beam has formed enough lines to fill the screen.

To get color, the front screen is coated with a pattern of tiny dots made of three different phosphor materials that glow red, green, and blue. Three different beams of electrons are directed at the screen. There is only one set of holes in the mesh, but since the three electrodes are positioned at different points in the back of the tube, when electrons from different starting points pass through the same hole they continue at slightly different angles and hit different points on the front screen. The phosphors are laid down so that the straight line from one electrode through all the holes hits only the dots of phosphor that glow red. The other two electrodes are lined up to hit all the green and all the blue dots.

Sometimes we talk about the three electrodes as the red, green, and blue "guns". They appear to "fire" electrons at the screen. This common language has two problems. First, the electrons actually "jump" toward the positively charged mesh rather than being shot from the electrode. More importantly, although the three electrodes are aligned to hit spots of the three different colors, there is nothing "red" about either the electrode itself or the electrons that it generates. Electrons don't have a color. The phosphor on the screen that they hit generates the color.

The light on the screen produces photons. They travel to your eye, are focused by the lens, and strike the cones at the back of the retina. The Red, Green, and Blue phosphors tend to trigger the Red, Green, and Blue cones. A mixture of the three types of light triggers some combination of signals that the brain senses as particular colors. With just three phosphors on the screen, the TV can generate the appearance of any color your eye can see because the eye itself only has three types of color receptors.

Originally the red, green, and blue phosphors were deposited on the back of the screen as round spots. However, if you cover a surface with a pattern of disjoint circles you leave a lot of area outside of any circle. Sony came up with the idea of a Trinitron screen, where the phosphor is deposited in vertical stripes and the spots are shaped as rectangles. Since more area is covered with phosphor, the picture is brighter.

The first mistake that computer people commonly make is to assume that the spots of color on the screen are directly related to "resolution". That would be the ideal situation, but an analog TV doesn't really have a clear "resolution" and a computer monitor supports many different resolutions.

A popular computer display resolution is 1024x768. When the adapter is set to this resolution, it wants to draw 1024 dots on each of 768 lines on the screen. In the adapter memory there are 1024 times 768 memory locations in which are stored the color desired for each of these picture elements or "pixels". The adapter has been told to refresh the computer screen at some rate. At a refresh rate of 70 Hz, the screen is rewritten 70 times a second. Every 1/70th of a second all 768 lines have to be written. Every 1/(70*768)th of a second a line has to be written (ignoring for the moment that there is some special overhead between screen refresh cycles to synchronize things). In that period of time reserved to draw one line, each electron beam has to scan the screen from left to right. The adapter divides the period of time that the beam will scan across the screen into 1024 time periods. During each time period it generates a voltage level for each of the three electrodes to generate the desired color for the currently active pixel. Then it shifts to the next memory location and changes the voltage level for the next pixel.

If everything was perfect, there would be 1024 holes in every line of the screen mesh, and 1024 red, green, and blue phosphor dots on every line. Furthermore, when the adapter was generating the voltage to the electrodes for one pixel, the magnetic ring would be directing the electrons to mostly go through a hole and hit a phosphor dot. Then the physical design of the screen would exactly match the logical resolution of the adapter. However, a computer monitor can be set to many different resolutions, and there is no particularly good way to adjust the timing to target the holes in the front mesh exactly. So instead, the adapter simply generates 1024 changes in the voltage level across the screen without regard to the number or location of holes or phosphor dots. If the beam generated by a pixel spans two phosphor dots, then the pixel "smears" across multiple dots on the screen. Alternately, one phosphor dot can get half its color from one pixel and half from another. Generally your eye doesn't see these tiny imperfections from normal viewing distances.

Plain Old Analog TV In the previous section, a CRT screen displayed information from a computer video adapter that started as a set of digital points. During the sweep of the beams across the screen, each digital value is converted to a voltage level for the period of time associated with one dot in the current resolution. Something of the same process occurs when a TV picture is generated from a digital source (DVD, digital cable, or PC TV Out).

However, conventional TV (broadcast to an antenna, VHS tape, or the cable signal on Channels 2-99) is an "analog" signal. A TV only has 240 lines. The analog broadcast signal is marked with a clear beginning and end to each line. However, within the line the signal is simply a wave form with components for intensity and color. This signal is decoded into a voltage level for the red, green, and blue electrodes. The signal could change continuously as the beam plays across the screen. When the beam happens to hit a hole in the grid and a phosphor on the screen, the current level of the signal contributes to the brightness of that dot of color on the screen.

The TV signal begins when a camera focuses the image on a sensor grid. A conventional TV camera records 480 lines (twice as many as the TV can display) and repeats the process 30 times a second. Today, a TV camera is digital, so it has a specific number of sensor points per line but back in the 1950's, the TV signal was scanned as a continuous wave in the camera. The camera transmitted this wave as an electrical signal across wires to the transmitter. The transmitter converted the electric wave to an electromagnetic wave broadcast through the air. The antenna at a house converted the electromagnetic wave back to an electric signal that the TV set amplified and sent to the electrodes at the back of the picture tube. Throughout the entire process there was no digital technology and it would have been misleading to say that the signal intended the TV set to have any particular resolution in terms of the number of dots per line. That noted, any individual TV set actually had some specific number of holes in the mesh and phosphor dots on each line.

The TV set had only 240 lines on the screen, and rewrites every one of them 60 times a second. The camera recorded 480 lines of picture 30 times a second. To get the appearance of more vertical resolution, the TV signal is "interlaced". Every other line is transmitted (240 lines of picture) in 1/60th of a second. Then a second pass transmits the 240 intermediate lines that we skipped in the previous pass. Every 1/30th of a second a complete frame of 480 lines is transmitted, but they are displayed on the screen as alternate sets of 240 lines.

A VHS VCR receives this analog broadcast signal and records it to tape. When you play the tape, you get back the same analog signal (plus any noise introduced by imperfections in the recording process).

The standards for broadcast analog TV (and therefore the conventional analog output from anything that feeds the TV: cable TV box, DVD player, VHS tape player, or PC) have not changed since the '50s. All the equipment has certainly changed. All modern TV cameras are digital. In the studio they record digital signals on digital tape. The information can be copied onto a hard disk, edited by computer software, and written back to tape. Old analog tape could not be edited as easily, and there was a loss of signal quality whenever you made a copy. Digital information is easily edited and every copy is perfect. The networks feed their programming to individual stations through big satellite dishes or fiber optic cable. This network feed is a digital signal in a form of MPEG 2 compression. In the end, however, the local TV station converts the digital signal back to a 1950 era analog signal to broadcast it to your TV set.

The analog signal transmits 240 lines. The start and end of each line is obvious from the signal. As each line is drawn, the analog signal continuously varies the amount of Red, Green, and Blue content to "draw" the line of picture on the screen. This signal can be digitized by a circuit that samples the amount of Red, Green, and Blue signal generated for a period of time that corresponds to one "dot", then stores the measurement as three numbers and continues on to measure the next dot. Every analog TV PC receiver card (and all Digital Video Recorder devices like Tivo) turn the analog signal into a stream of numbers. Typically these numbers are immediately compressed to a smaller, reasonable amount of data.

Any consumer device (TiVo, Replay) expects to get 240 lines of TV signal refreshed 60 times a second. Yes, this is an interlaced signal, but consumer devices do not expect to convert the recorded information to any other format and they do not expect to display the picture on anything except a TV. So they can simply record it as 240 lines 60 times a second.

However, a computer screen is not interlaced. To display the TV picture on a computer, the software has to "de-interlace" the original signal and create a picture with the original 480 lines refreshed 30 times a second. The digital file formats support a wide range of number of lines, number of dots per line, either interlaced or non-interlaced (also called "progressive" scan).

If a conventional TV set quotes a resolution, it reflects the number of horizontal phosphor dots per line. The vertical number of lines is fixed at 240. Horizontal TV "resolution" is different from computer display resolution. A computer display adapter is digital and the screen image is generated as dots or "pixels" of color in the computer memory. These dots are then transmitted to the display. A TV signal is a continuous analog signal, essentially a "wave" of colors painted right to left across the screen. There are no discrete dots in the TV signal, but the colors change continuously across the screen. Every time the electron beam sweeps over a set of phosphors on the screen, it generates a dot of picture. More horizontal dots will display more of the fine details in the original wave, provided that those details are present in the analog signal. A higher resolution conventional TV might display a better picture from a DVD player.

The US standard for TV signals is called "NTSC". The US power grid generates alternating current at 60 cycles per second, and back when TV's were invented there was no better timing reference. NTSC represents a combination of lines and refresh rates that made sense given the vacuum tube electronics of the 40's if you had to do your operations at a rate of 60 per second.

Each TV channel has a fixed amount of bandwidth. In other countries the electric power cycles 50 times a second and their TV pictures change at that rate. Given a fixed bandwidth, if you transmit fewer screen images per second you can add a few extra lines to each image. So traditional European analog TV standards feature more vertical lines in each picture, but change the screen image less frequently.

Obviously we can do much better in the modern world of computers, where internal timers measure speed in billionths of a second and microchips provide enormous processing power. The real problem is that this technology is continuing to improve. Any standard that is comfortable today will be obsolete in a few years. The new digital TV standards take this into consideration by selecting values that are just a little beyond the reach of today's mass market devices. The industry can spend the next decade letting the mass market catch up and completing the conversion.

To position it against new digital TV standard, the Plain Old Analog TV standard is referred to as "480i". It has 480 lines of image, but they are interlaced so that only 240 lines show on the screen at a time. By flashing alternate views of every pair of lines, the TV presents what appears to be a sharper picture than 240 lines, but not as clear a picture as if the set had 480 lines of picture to display all at once.

Modern Flat Panel Display Resolution is a number of dots per line and lines per screen that the source computer, player, game console, or set top box is generating. Each LCD or plasma flat panel display has a physical native number of dots per line and lines per screen. You get the best picture if the source resolution matches the physical screen. For any other values, the screen has to shrink or stretch the picture.

Computers and old TV sets traditionally had square looking monitors with what is called the 4 to 3 aspect ratio of width to height. However, the popularity of movie playback and HD TV is increasing the number of computer monitors and TV sets with a more theatrical wide-screen aspect ratio of 16 to 9.

Originally there were a small number of standard resolutions with sell known names. PC adapters and their device drivers supported these specific resolutions and often did not support anything else. For historical reasons it is useful to mention the most popular. Conventional 4:3 resolutions include 640x480 (480 lines of 640 dots per line) which was the original VGA, 800x600 (SVGA), 1024x768 (XGA), and 1280x1024. Wide aspect ratio computer monitors use the same set of numbers, but they typically match the next higher dots-wide number with the next lower dots-high number. So a handheld wide aspect ratio device will take the 800 from 800x600 and match it to the 480 of 640x480 to get 800x480, while a 19" $160 desktop computer monitor ends up with 1280x768.

HDTV has two specific resolutions. One is 1280x720 (called "720p") and the other is 1920x1080 (called "1080p"). You will find LCD TV sets with these resolutions that can double as a computer monitor, but screens manufactured to be a computer monitor will typically have a slightly larger number of rows. As shown above, the computer standard is 1280x768, so there are 48 more lines than the TV standard of 720.

However, modern flat panel monitors come in many shapes and sizes, and their native resolution is often slightly different from any of the standards. You can find panels that are 1920x1200, 1680x1050, and 1440x900. There are 30" computer monitors with resolutions of 2560x1600. At boot time the monitor tells the video card what its native resolution is, and current video cards and device drivers can accept and adopt to any resolution.

Monitors can accept analog signals in the old computer VGA format or in the old TV component signal (Y Pb Pr). They produce a sharper picture, however, if connected by a digital connection such as a DVI, HDMI, or DisplayPort cable.

Analog Formats and Connectors Everything in a computer is based on separate values for Red, Green, and Blue (RGB). The adapter maintains a separate byte of data for each color, the adapter then generates a separate voltage level for each color one one of the pairs of pins on the plug that connects the computer to the monitor. Inside the display adapter, and for that matter inside the TV set itself, three electrodes generate a flow of electrons for each of the three colors.

So it may be surprising that TV sets operate on an entirely different system. If TV started out as a color picture, then it probably would work the same as computers. However, the first TV sets produced a black-and-white image. Black-and-white TV was available for more than a decade before color signals were added to existing TV channels. Unfortunately, the black-and-white part already took up most of the channel, so the color was squeezed into a smaller part of the spectrum.

The black-and-white part of the signal is called luminance and is represented in the standard by the letter "Y". This signal is not equal parts of red, green, and blue. The human eye sees the three colors differently, and the black-and-white signal has been weighted to reflect what looks best to the eye.

In high school geometry everyone learns that conventional "Cartesian" coordinates (x, y, z) can be translated to "Polar" coordinates. In 3D, the Polar coordinates are represented by the vector length (magnitude or absolute value) and two angles from the origin direction. In the same way, any combination of red, green, and blue can be represented by the black-and-white "Y" signal (like magnitude or absolute value) and two other values that represent color differences (like angles) from a pure black and white origin.

There are three colors. In the TV system, there is one Y value and two chrominance values. You can translate from either system to the other. In the modern world of computers, we add and subtract numbers to do the translation. However, the old TV systems did not have computer chips. The Y value and the two chrominance values were analog waves picked up through the antenna and amplified through the vacuum tubes. Two waves can be "added" together by combining them directly. One can be subtracted from the other by flipping its value (positive to negative) and then combining it with the other wave. The three waves can be combined together through various filters, and the output fed to the red, green, and blue electrodes in the back of the TV tube.

Better: Component (3 jacks) A High Definition TV or Progressive Scan DVD will support three RCA jacks that carry a “component” signal. The three jacks are colored red, green, and blue, but the signal they carry is not RGB like a computer display. The green jack is the black and white intensity signal labeled “Y”. The other two cables are labeled Pb and Pr and they carry color information.

With three separate wires to carry the three separate signals, the receiver does not have to apply any filter to split a signal. There is, therefore, no currently important limit on the frequency of the signal or the resolution of the screen. Component cables can be used for anything from Enhanced Definition up to High Definition TV.

Good: S-Video (1 round jack, two pair of wires) [Standard Definition only] Almost all DVDs, SVHS tape devices, and moderately priced conventional TV sets have an S-Video connector. This is also called a Y/C connector because it has two pair of wires. One carries the Y black-and-white signal. The other carries the two color signals combined together using standard wave multiplexing technology.

Because the two color components have been combined, they have to be separated by a filter on the receiver. This is fairly simple to do on the plain old standard resolution TV. As the resolution increases, the frequency of the signal increases. The conventional technology for filtering the C signal into two elements does not work correctly at higher frequencies. That is why the three component connectors are required for HDTV.

OK: Composite (1 yellow jack) [Standard Definition only] The oldest and most widely supported standard combines all three video signals as a compound wave form in a single "composite" signal. Typically this is represented as a single yellow RCA jack, typically combined with the red and white stereo audio jacks. If you go to Radio Shack or Circuit City, one of these three-connector audio/video cables is typically called a "dubbing cable". You get a few of them with every VCR, DVD player, or computer sound card. Since the device wears out or becomes obsolete long before the cable develops a problem, you probably have a drawer full of dubbing cables at home.

Comb Filter When signals are combined together on a wire, the receiver has to separate them out again. The device for splitting a composite signal (or a TV channel received on cable or from an antenna) into a black-and-white Y signal and the color signals is called a "comb filter". The quality of the comb filter is an important measure of the quality of most TV sets. However, if you don't actually do the tuning inside the TV set, then the comb filter function may be performed by another device. For example, if a channel is actually received by the VCR, and the signal is then transmitted to the TV over component or S-Video cable, then the VCR comb filter is doing the work and the TV comb filter is not involved.

Remember, the color signal was squeezed into the small amount of spectrum left free in the TV channel after the big black-and-white signal took up most of the space. To make this squeeze work, the color signal has less resolution than the luminance. Each color value may apply to two "dots" generated by changes in the black-and-white part of the picture.

You might think this changes on a DVD or digital TV picture, where more bandwidth is available and there is no history to screw things up. On the other hand, DVD and DTV images are compressed, and engineers looking for the best compression algorithms find that people tolerate more aggressive compression of color than luminance. Even then, there is a lot more color information in a DVD than in broadcast or cable TV. That is why you want to use S-Video or component cables to connect the DVD, SVHS tape, cable set top box, and other modern equipment to a modern TV.

Worst: Cable TV F-connector The least desirable option is to connect devices with a piece of TV coax cable (technically, an RG6 cable with an "F" connector at each end). To use this connection, one device (the cable converter or VHS recorder) generates the signal as TV channel 2, 3, or 4 and the receiver tunes the channel as if it were received from a broadcast antenna. This is also called a Radio Frequency or "RF" connection.

The problem is that TV radio frequencies were assigned before color TV. There is a big frequency range for the black-and-white signal, a smaller frequency range for the color, and then a range for the audio. A TV tuner can pick up the signals, produce the Y/C division, and then split the color signal into two elements. However, there is some error and loss of quality in the process.

The simplest connection between a cable box or VCR and a TV set is a single RF cable. All the video and audio are carried on the one wire. This also produces the worst picture. Any of the three previous alternatives are preferable.

Choose the highest quality available connection from the first available connector in the following order (best to worst):

component, Y Pb Pr (three RCA connectors) S-Video, Y/C (two wire pairs in a round "DIN" connector that looks like a mouse or keyboard connector) composite, one coax wire typically packaged as a yellow RCA connector, frequently combined with two stereo audio wires RF (cable TV wire) Cables The biggest opportunity for a rip off is the high cost premium cables. Over the one meter connection between the TV and an external tuner box, there is little measurable difference between the cables that came free with the box and some monster $60 cables sold separately. Use standard cables initially and upgrade to something better only if there is a problem. When it doubt, see if you can borrow some premium cables from some friend who already got suckered into buying them. Do a “blind” test where someone switches the cables and someone else judges the picture without knowing which type of cable is currently being used.

Digital Cables (DVI, HDMI, DisplayPort) Internally, the computer generates three numbers for the colors of each dot on the screen. Internally the flat panel monitor or TV has to generate three numbers for each dot on the screen. Now you can use one of the analog cables (the component Y Pb Pr cables or the 15 pin VGA cable) to convert each number into an amount of voltage difference sent over the cable in a short time window that corresponds to each dot. Or you can use a digital cable to simply send the numbers as a stream of data just like the USB, firewire, Ethernet, or other computer cables.

Digital cables produce a sharper picture because the numbers can be transmitted precisely. An analog connection is less sharp because some information is lost converting the numbers to analog voltage and then converting them back to numbers.

Although we normally think of the DVI as a digital interface, it normally also contains a set of analog Red, Green, and Blue signals. Video cards typically come with a DVI to VGA plug adapter, so that an old VGA analog only monitor can be connected to the card when a digital connection won't work. At power up, the card tests the digital and analog lines and uses whichever one is connected to something.

HDMI is a smaller plug that supports only digital data. You can convert the digital part of DVI into an HDMI plug and visa versa, but HDMI cannot be adapted to analog. HDMI also contains wires for digital audio.

DVI is a computer standard, although some TV sets and cable TV set top boxes have a DVI plug. HDMI is a TV and consumer electronics standard, although some PC video cards support it.

The next video connector standard will be DisplayPort. It can support a higher resolution in a 3 foot cable, or it can support the current 1080p highest resolution TV standard in a a longer 15 foot cable. The first DisplayPort devices appear in early 2008 and more will roll out over time.

Video cards (and mainboards with integrated video) can support the HDCP encryption mechanism. If your computer and your monitor both support HDCP, then it doesn't matter if the plugs and cables that connect them are the clunky DVI or the newer, smaller HDMI.

In order to license necessary technology from the movie companies, software packages that play Blu-Ray and HD DVD disks will not display their picture on a monitor through an unencrypted DVI or HDMI connector. If the software will not play disks for this reason, then either you have to upgrade your system to enable HDCP encryption or else you have to switch from the DVI/HDMI digital cable to the old Y Pr Pb or VGA analog cables.

File Formats Good: MPEG 2 (DVD) MPEG 2 comes in two flavors. The type of MPEG 2 file written on a DVD disk is called Program Stream and stores data in 2048 byte chunks (which is the native size of one block of data stored on a DVD). The type of MPEG 2 broadcast by TV stations or transmitted on a cable TV system is called Transport Stream (TS) and formats data in 188 byte chunks. These smaller units are needed to minimize loss and simplify recovery when there is interference with a signal broadcast and received by an antenna, but they are inefficient when data is stored on a disk. So while the TS format may be initially used when a broadcast or cable program is first recorded, after you edit it to remove commercials you will probably convert it to Program Stream for long term storage.

When Microsoft began to offer its Media Center option, it required a hardware encoding chip on any TV tuner that could take over all the work of compressing a digitized version of the standard definition analog TV program. This is probably a good idea in any event, because when the encoding is done on the card there is so little demand for CPU that a single computer can easily record 4, 6, or more TV programs all at the same time. Recording standard definition analog programs as an MPEG 2 stream was also attractive because there are several programs that can convert an MPEG 2 recorded program directly into the disk file structure used for video DVD disks. The recorded program can then be burned directly onto a DVD-R for long term storage.

Both the FCC digital TV standard and the DVD movie file format combine both picture and sound. In fact, they support several alternate sound tracks typically in different languages. So MPEG 2 provides two comprehensive file format standards.

Better: MPEG 4, Divx and Windows Media (WMV) The simple versions of MPEG 4 represent the highest compression or best quality that can be recorded in real time on a current computer. Unfortunately, while lots of devices have hardware MPEG 2 encoding on the board, and while simple MPEG-4 encoding chips have been available for several years, only one vendor produced a PC product that provided hardware MPEG 4 encoding.

Therefore, Divx and Windows Media tends to be an option when the tuner card does not do any compression and the data is compressed instead using the CPU. That is fine if you are only recording one program at a time. Given the current mix of hardware, you will have trouble trying to do MPEG 4 compression using the CPU for two or more programs at the same time.

MPEG 4 is simply a video compression standard. There is no generally accepted file format. Microsoft has the WMV format, Divx has its format, or you can create AVI files. Every software package has its own way of adding the audio part, and players do not always accept every version.

Best: MPEG 4 Advanced, H.264 The best video encoding standard is the advanced version of MPEG 4 described in "Part 10" of the standard. The problem is that current computer hardware cannot compress video using this standard in real time. Generally you have to capture the program using one of the other standards, and then overnight you can recompress it to advanced MPEG 4 at a slower pace. Typically you will need to run your dual core CPU at 100% for six hours to recompress one hour of HD video.

However, computer chips always get faster, and the particular operations required to compress video are one of the areas to which chip vendors pay particular attention. Some of the processing can get turned over to the unused capability of the video adapter Graphics Processing Unit. This may reduce the amount of time required for overnight processing, but at this time it would be speculative to predict that future systems will be able to do real time advanced MPEG 4 encoding of a program as it is received by the TV tuner.

Blu-Ray and HD DVD disks have standard formats for advanced MPEG 4, but they have not had the influence over file formats that the DVD had. It is too soon to determine if a file standard will emerge.

CODEC Software that encodes and COmpresses and DECompresses a multimedia file format is called a CODEC. Nothing prevents a proprietary media player from embedding its own CODEC in its own code so it can only be used by the one application. However, Windows has a system where the CODECs can be registered to the operating system and then shared by all applications. Well written packages will be nice and register their CODECs to the system as a whole.

Windows comes with a standard starter set of CODECs. Since Windows Media Player 9 is a free download, anyone can easily install the latest set of Microsoft CODECs, including support for MPEG 4.

The first CODEC that people notice is missing is the MPEG 2 support needed to play a DVD. There are patent license fees associated with the MPEG 2 standard, so a CODEC can only be legally obtained by purchasing a product.

The simplest way to get MPEG 2 support is to purchase a commercial DVD player product, like WinDVD or PowerDVD. MPEG 2 support is often bundled with hardware devices like ATI video adapters or Hauppauge PVR cards. Generally, however, these CODECs will not run if they discover that the corresponding hardware device is no longer installed in the computer. You can also buy an MPEG 2 decoder with the PureVideo support from Nvidia for its video cards. There are also a bunch of non-standard CODECs that can be installed in any Windows system. They claim to do a better job with a sharper picture, smaller file size, or lower CPU demand.

AVI Thanks to DVD, there is a pretty well established format for MPEG 2 video files. All the other standards have a format for the data streams, but not an actual file format.

Microsoft created the AVI file. Basically, AVI is a standard header that tells Windows what CODEC to use for the video, and which to use for the audio, and where the streams of data are. AVI is the file format to use for all the forms of video compression that don't have their own universally understood file formats.

Standard File Extensions and Formats A Windows Media video file has an extension of WMV.

An MPEG 2 file on a DVD has an extension of VOB. MPEG 2 files recorded directly on your computer from a TV tuner typically have an extension of MPG. This is the Program Stream format with larger (2K) records.

An MPEG 2 file recorded from digital TV (broadcast, cable, or satellite) is in Transport Stream format with smaller records. It typically has an extension of TP or TS.

Any program that plays DVD video will also play MPG files, but you often have to upgrade to the latest release before they will also play HDTV TP/TS files.

Transport Stream as the name suggests is optimized for cable or antenna broadcast. If you record a long TV program in TP/TS format and then want to jump around in it (to show the key pass play in the football game) you will discover that your player software is slow to respond. To jump backwards or forwards, or to fast forward at high speed, you are better off converting the Transport Stream file into a Program Stream file (convert the TP or TS file to an MPG file).

Divx or Nero Recode produce their own MPEG 4 files with non-standard extensions. The Plextor USB TV receivers can compress programs in real time to yet a third version of MPEG 4. Unfortunately, although all of these files are MPEG 4, none of these programs will play files created in the other's file format. This is is a problem that needs to be solved.

Storage CDs are too small for serious TV. Today a DVD costs around $.20 and holds 7 times as much data as a CD that costs only slightly less.

The problem is that a DVD holds 4.7 gigabytes of storage. That is way more than an hour of standard definition TV. However, HD TV requires 4 to 7 gigabytes per hour even after you remove all the commercials (the actual file size depends on the format the network uses for HD). You can fit some shows on a DVD, and split others across two DVDs.

However, you should also consider the cost of disk. A 500 gigabyte hard disk costs $100. That means that one gigabyte of hard disk costs about the same as 4.7 gigabytes of DVDR. DVD is still cheaper for long term storage of things you really want to keep, but if you have a reasonable amount of disk you can keep things on disk long enough to ask the hard questions: how many times will I watch this, how long do I want to keep it, etc. After a month or two, you may not feel as attached to some series that started out so promisingly.

DVD A commercially manufactured music or data CD or a movie DVD contains 2K blocks of data encoded as pits in the metal surface of the disk. The CD was invented 20 years ago. Advances in technology permit improvements on that standard. First, the pits on the disk surface that carry the signal can be made smaller, allowing more data to be written in one spiral around the disk surface and allowing the adjacent spiral tracks to be packed tighter. The usable disk surface is extended, and the data is written in a slightly more compact encoding. Combining all these features, the surface of a DVD can hold 4.7 gigabytes of data compared to the 700 megabytes of a CD.

A commercially produced DVD can do even better. In addition to the base data embedded in pits in the metallic base, a second set of pits is added in the semitransparent plastic coating. The DVD sensor can focus either on the pits of the back layer or of the coating layer. This nearly doubles the amount of data on the complete disk and makes it possible to store almost 9 gigabytes of data.

A two hour movie can barely fit on one surface of a disk. Add several sound tracks, including a director’s commentary, plus trailers, bloopers, and other extras, and most movies require the capacity of a two layer disk.

Writable disks (CD-R, CD-RW, DVD-R, DVD+R, DVD-RW, or DVD+RW) have a reflective back surface covered by a plastic coating embedded with dye spots. Initially the spots are transparent. A laser operating at high power generates enough heat to change the chemical composition of the dye material and turn it dark. Later, a laser running at much lower power can focus on the surface. Where the dye was not changed, the light still reflects back. Where it was changed, the dark color absorbs the light. This technique only operates on a single layer, so current writable DVD’s have only 4.7 gigabytes capacity.

A writable disk is not featureless. It is manufactured with guidance information that enables the drive to position all the 2K blocks and the spiral tracks. The DVD-R(W) and DVD+R(W) standards accomplish this slightly differently. Each has advantages and disadvantages. Fortunately, the difference is more a matter of reprogramming the firmware in the drive rather than changing the hardware. Today most new DVD writers support both families of formatting.

Although some companies produce CD-R disks labeled “music”, the CD has the same physical low level structure whether it contains audio or computer data. The difference is in the directory structure written onto the disk by the program that burns the data. Similarly, a DVD video disk is simply a DVD data disk with some directories and files with special names.

The video files on a commercially produced movie DVD are encoded by an encryption system called “CSS”. For the movie to play, it must be decrypted by hardware in the DVD player or software on the computer. It is illegal to create software to bypass the CSS protection, but such software is widely available from internet sites outside the US.

A movie DVD compresses video in the MPEG 2 format. Ten years ago this was leading edge technology. To play a movie DVD on a computer required a 333 MHz Pentium II CPU. DVD players were expensive. In those days, DVD players came from Japan or Korea. Today CD and DVD drives and players come from China. Prices have been falling month by month. By Christmas 2003, DVD players were selling for $30, and a DVD writer for a computer costs $100.

All current DVD movies are designed to play on traditional TV sets. They may have a very sharp picture, and some may be designed to stretch the picture horizontally for a “widescreen” display, but the files are still encoded to display a picture with 480 lines of resolution 30 times a second by alternating two sets of 240 lines 60 times a second. New DVD players have a “progressive scan” option to intelligently combine adjacent groups of 240 lines (deinterlace) to provide a 480 line picture on TV sets with enhanced resolution.

High Definition DVD We are now seeing the launch of two competing systems that substantially expand the amount of data that can be stored on the CD/DVD disk form. HD DVD and Blu-Ray represent slightly different versions of the same basic technology. Exactly what encoding standards and features will be on each format has been the subject of negotiation right up to the wire.

The main difference is the manufacturing technology. The HD DVD is basically the same as a current DVD, only the pits are smaller so they have to be read by a blue laser. Generally speaking, you can take a current DVD production factory and turn it into an HD DVD factory at very low cost.

Blu-Ray, however, requires a new coating on the disks that is much thinner than the plastic coating on a DVD or CD. This is a new material and requires new equipment. To protect the disk surface, this thinner coating must be harder and more resistant to scratches.

There are certainly technical details that people can argue about for hours, but a quick summary is that Blu-Ray is more advanced technology that, today, is much more expensive. HD DVD is less advanced, but it starts at a much lower cost.

Which wins depends on how two predictable technology trends interact. Every previous CD or DVD technology has started out expensive and then become affordable over time. At some point Blu-Ray will become affordable. However, at the same time compression is reducing the amount of space required to hold HD content. So the technical advantages of Blu-Ray over HD DVD may become less important.

Channels and Modulation Digital TV is carried over the same broadcast and cable TV channels as plain old analog TV. This means that the radio frequency spectrum is divided into 6 MHz units. When the first TV stations came on line, a block of frequencies was assigned to channels 2-12 (VHF). Later a block of higher channel numbers (UHF) became available. UHF channels originally ranged up to 83, but channels above 69 are no longer assigned.

When cable TV systems appeared, they were not constrained by this historical accident. Channels 2-12 have the same frequencies on cable as in broadcast, but the higher channels were assigned successive 6 MHz blocks up to something around channel 183.

Within the assigned frequency range, a signal is generated by some form of "modulation". The simplest technique is the Amplitude Modulation of the AM radio band. The strength of the signal goes up when the sound gets louder. AM, however, is sensitive to static, interference, and is hard to tune.

When there is no sound, AM generates no signal. The next technique was Frequency Modulation or FM. In FM there is a background carrier wave that is always present. When there is silence, the carrier signal is present but it is doing nothing. The signal is generated by imposing a variation over this base carrier. FM is insensitive to lightning and interference from appliances. The carrier also allows the receiver to automatically tune in to the station once it gets close enough to the base frequency.

AM holds the frequency constant but changes the intensity of the signal. FM holds the intensity constant but changes the frequency. Essentially the two different techniques are completely orthogonal to each other. If you think in AM terms, there is no signal on an FM channel and visa versa.

It is possible to add more signal to a frequency by coming up with new way to modulate the new signal that is independent and therefore invisible to whatever techniques were used to generate the old signal. In the 1950s a radio signal could only be processed by analog components and filters. Today the same digital technology that makes computers and cell phones possible also provides digital signal processing chips that can make fine distinctions that would never be possible with analog components.

In FM radio, there is a big primary signal that provides monaural sound. A second weaker signal is then imposed over it that represents the difference between the left and right signal. By applying this second signal to the first, and then applying the exact opposite of this second signal to the first, the original signal can be converted to stereo. In the first generation of "home theater", an additional signal could be provided to split the left and right signals into front speaker and back speaker components.

When an FM station is far away, the first thing you lose is the stereo. However, if you switch back to “mono” mode, you may still be able to tune in the stronger monaural signal.

A conventional analog TV stations broadcasts separate video and audio signals. The audio part is the same as FM radio. The video part has a large signal component for the black and white (intensity) part of the picture. There is then a secondary component for color. The color part has a third level of signal that splits it to provide enough signals to drive the Red, Green, and Blue electrons of the TV set. The analog video signal has synchronization points to mark the start of a screen refresh (60 times a second), and the break between each of the 240 lines on the screen.

When an analog TV station is far away, the color fades out. Some black and while signal remains, but as the distance increases it is eventually lost in the snow.

Digital TV uses the same 6 MHz range of frequencies, but with different modulation schemes. Actually there are two modulation systems in the US:

8VSB is the modulation scheme for over the air digital TV broadcast in North America (ATSC). QAM is the modulation scheme used for digital cable TV and direct satellite broadcast. Terrestrial broadcast TV has to be able to recover when the signal bounces off nearby buildings or hills. With error correction, it squeezes a little less than 20 megabits per second of signal out of the 6 MHz TV channel. Cable TV channels are transmitted in a protected environment, so it doesn't need the same level of error recovery. The high density QAM modulation they use can get twice as much signal, or a bit less than 40 megabits per second out of the same 6 MHz range.

Packet Format No matter which type of modulation is used, the output of the tuner is a stream of bits that form a sequence of 188 byte packets. If the tuner is connected to an antenna, the stream is 20 million bits per second, while the clear QAM of a cable TV system yields 40 million bits per second.

Each packet as a Packet ID or PID value. When you tune to a new stream of bits on a new radio frequency, you first have to wait for a packet with a PID of 0.

The PID 0 packet contains a definition of the logical TV program streams contained on that physical channel. A broadcast over the air TV station typically transmits one HD program stream and one standard definition stream for old TV sets. On a cable channel, there can be up to 10 different standard definition streams.

Each logical TV program has its own control PID. If you select one and then wait for the next packet with that PID number, it will contain information about the PID of the video component of that program and the PIDs of each audio component (because TV shows can be broadcast with both English and Spanish audio tracks).

In addition to this minimal control information, there are other optional control packets that identify a logical channel number and sometimes the call letters of the station. Cable TV also transmits information about the program including its title and some episode information. However, the availability and use of this data is not assured.

So if you want to view or record a particular TV program, you tune to the physical channel frequency, wait for the PID 0 packet, select a program stream, wait for its first control packet, and then start to process all the packets that contain the desired video and audio PID numbers (typically ignoring everything else).

Scaling A standard definition program or DVD movie has 480 lines of 720 dots. A HD program can have 720 lines of 128 dots or 1080 lines of 1920 dots. The problem is that your actual monitor cannot have three different sizes. It has some particular number of lines and dots per line. So when it is asked to display a picture that uses a higher or smaller resolution, it has to transform the picture it gets into an image that fills its screen.

Now you can just do the transform with simple arithmetic. However, there are more sophisticated tricks to take the image you have and make it look better and sharper than it really is. This logic tends to be embedded in the TV set, although if you use a monitor connected to a PC then it may be embedded in the PC program use use to watch TV.

While there are all sorts of possibilities, a few special cases are more important than all the others. HD TV makers pay particular attention to maximizing the display of an old DVD disk on their screens. DVD players may take over and do the scaling. It all depends on how the TV is connected to the player. If the TV is connected with an old component or S-Video cable and if the TV indicates that it is receiving an 480i signal, then any scaling improvement is up to the TV itself. However, if the player is connected by component (Y Pb Pr) cables or HDMI and the TV indicates that it is getting a 720p or 1080p signal, then the scaling is being done by the DVD player itself.

Computer Analog TV Adapters Now that broadcast TV will cut off its analog signal in Feb. 2009, it is getting a little late to buy a TV tuner that can only receive an analog signal. However, there will still be analog signal on cable TV, and for the next few years some analog tuning may be necessary. In New Haven the Sci Fi Channel is analog 52 and Comcast has not indicated plans to offer a digital version of the feed.

Digital TV is data that is already compressed and can be recorded directly onto disk. Analog TV, however, must be digitized by your receiver card. Many TV tuner devices also have an MPEG 2 encoding chip that can compress the stream on the fly before sending it to the computer. When the computer gets this data, it is just like the digital TV stream. Without built-in compression, the CPU has to use a lot of processing power compressing the numbers it gets from the TV tuner to produce an MPEG 2, WMV, or MPEG 4 file.

The author has had best results with the Hauppauge family of cards. The Hauppauge PVR 150 will record a single TV channel, but if you want to record more than one thing at a time you should get the PVR 500 card with two independent TV tuners. This also requires only one TV input and one PCI slot. There are other vendors with PCI Express adapters that should probably work as well.

The Plextor PX-TV402U is an external analog TV tuner with built in compression, that connects to your computer over a USB cable. It has hardware MPEG 4 compression. This produces a better picture and a smaller file size than every other device available, all of which have only hardware MPEG 2 compression.

Computer Digital TV Adapters There used to be only a handful of digital TV tuner cards from specialized vendors. As we move closer to the analog TV cutoff, all the companies that used to make analog tuners are introducing digital products.

There are a few digital-only products with specialized features. However, in the price range around $100 there are a large number of hybrid and combo cards.

A hybrid card has one analog tuner and one digital tuner. Each is connected to its own antenna jack. You can process one analog and one digital program at the same time.

A combo card has one tuner that can do either analog or digital depending on how it is programmed. It has one antenna jack.

A Separate Network Appliance The most interesting digital-only device is from a small company called Silicondust Engineering. They build a box called HDHomeRun that receives the digital TV signal and transmits it to your computer over your home Ethernet network. All the other TV tuners are devices that connect to the PC through USB or by being plugged into the PCI bus. Since they are real devices, they need real device drivers. However, writing a device driver is a complex programming task and the company doesn't always have great programmers.

The HDHomeRun box internally has the tuner devices and an embedded operating system. It handles all the hardware level programming inside its own box. Then it transmits the HD TV data as ordinary Internet traffic.

Your computer receives data from the HDHomeRun just as it would read any other data from the Internet. Hundreds of programs on your computer read data though the internet. This is very simple programming that almost anyone can do. As a test, the author spent a few hours and wrote a 20 line program from scratch that received the TV program from the HDHomeRun and recorded it as file on disk. When something is that simple, it is much, much easier to get it right, and the vendor is much quicker to adapt it to new situations.

Currently, HDHomeRun is the only device that can feed QAM unencrypted network HD programs from a cable TV system to Windows Media Center, or the two standalone Windows programs Sage TV and Beyond TV (it also supports the open source Linux community, but that is not for ordinary people). There is a trick. If you can understand the trick, then you can decide if you want to use this box.

Consider the ABC station in New Haven. For decades, it has been WTNH Channel 8. The FCC assigned every TV station a second frequency for its digital signal. WTNH got the second channel 10 over which to broadcast the digital version of what it has always transmitted in analog form over channel 8. A digital broadcast channel can transmit three or more programs. Typically the first one is the HD program and the other two are used to broadcast community announcements, school closings, and a continuous weather radar picture. So the HD ABC program is really 10-1 (subprogram 1 on physical channel 10) although if you are used to thinking of WTNH Channel 8 you might prefer to still think of it as 8-1 even though the digital transmission has nothing to to with physical channel 8. If you go to TV Guide, or any Internet version of TV listings like TitanTV.com or Zap2it.com you will see listings for digital WTNHDT on channel "10.1" or some such designation.

Now Comcast receives the digital WTNH broadcast and sends it in QAM form over its cable system. However, all the channels up to 89 were already in use for analog or other digital traffic. So Comcast puts the WTNH HD traffic on the first subprogram of channel 93. Comcast doesn't tell you this, and I doubt they have anyone you can call on the phone and ask. However, just like every other TV/VCR device made in the last couple of decades, HDHomeRun (and Dvico and all the PCI based tuner cards) have a channel scan capability. Start it up and 10 minutes later it will have mapped out all the stations on the cable.

Microsoft designed Media Center to support analog broadcast, cable, and digital broadcast channels. They didn't add digital cable to the design of the program. Right now the only HD programming that Media Center knows about is broadcast digital, which means that the program and its program guide and user interface expect to find WTNH ABC programming on physical channel 10 (which is associated in some guides with the old channel 8). Microsoft isn't going to change this any time soon.

So HDHomeRun provides Media Center (and Sage TV, Beyond TV, and a handful of other programs) with a version of the standard Microsoft TV driver that lies. Media Center displays the guide, and you decide you want to view or record channel 8-1 (or 10-1 depending on how the guide is set up). The HDHomeRun program looks for a simple text file you have stored on your C: drive. It has one line for each TV station. The line for WTNH reads:

8-1 10 qam256:93 1 So when Media Center or any other program tries to tune to the first digital subprogram on broadcast channel 10, the HDHomeRun driver sends a command over the home network to tell the HDHomeRun box to tune to QAM channel 93-1 instead. Your PC programming sees the subset of the digital cable that contains free network broadcast HD programming, but thinks that it is connected to a really good antenna on your roof.

Once you view HD you don’t want to watch or record in standard definition any more.

Media Center, Beyond TV, or Sage If you have one TV adapter connected to one source, then you can probably schedule all your viewing and recording manually.

Media Center is built into Windows Vista Home and Ultimate. Beyond TV is made by Snapstream. Sage TV is another version of the same thing. The author uses Beyond TV, but assumes the others are equivalent.

In Beyond TV, a user selects one or more TV layouts during installation. For example, I begin by selecting USA, my zip code, and my cable service. That defines all the Comcast digital channels. To support the HDHomeRun, I add a second dummy layout consisting of the over the air digital broadcast network channels that I would be receiving (however weak) if I put an antenna on my roof.

I then carve the Comcast Digital layout into pieces. The channels less than 99 are Analog and can be received by the Hauppauge card or any other analog tuner. The channels above 100 can only be received through the cable box and the one device (the USB Plextor) connected to the SVideo out from the cable box.

Now I define the input devices.

The HDHomeRun is actually connected to the cable, and the remapping file lets it pretend to be receiving broadcast digital network HD programming on the channels where those programs would be received through a rooftop antenna. The Plextor can receive all channels in the Comcast digital lineup, but because it uses the cable set top box the computer has to use an InfraRed device to pretend to be the remote control and switch channels on the cable box. The Hauppauge video card can receive channels 2-99 in the Analog lineup. Now I launch Beyond TV. It loads TV listings for the next 10 days over the internet. With the mouse and keyboard I can scroll between channels (up and down) and between time periods (right and left). After all the usual cable channels, there are six extra channels representing the HD network broadcast channels received through the HDHomeRun. I can select individual programs, record all new programs in a particular series, or "record movie within 30 days". It assigns devices to record each selection, and if it runs out of devices in a particular time period it will try to move some recordings to alternate showings of the same program. This is vastly simpler than manually scheduling each device using the programs that came with each device.