DV Format
- Digital now
- Why digital?
- Digital video - wherer did it come from
- The DV advantage
- The physical side of digital video
- DV details - video standards
- DV - image, sound, tape
- Digital vs analog
- DV in use
- DV & DVD
- DV & computers
- You've made it
1. Digital now.
Television, which began life in the 1930s, and which grew to become the major focal point of family life in the second half of this century, gave birth to the Video Age in the 1970s, with the invention and mass appeal of the consumer video cassette recorder (VCR). Our view of culture and society has been completely transformed because of this relatively simple and accessible medium. The VCR and the camcorder have created new definitions for the way we see ourselves and added a variety of new terms and expressions to our language.
Now the Video Age is going through a major revolution with the arrival of digital video. This momentous development will bring new opportunities, creativity and a wealth of new technology to our fingertips. Video as we know it will be replaced with a new focus, as the advantages of digital vide are universally recognized.
But revolution by its very nature is about upheaval and confusion, as people have to learn new ideas and a new language. There is both information and misinformation to wade through, and relatively simple principles have been made unnecessarily complicated, intimidating, or (worse still) boring. It is our aim to bring this subject to you in a clear and concise way, so you can easily understand the potential of digital video.
2. Why Digital?
Simply put, video has taken us a remarkable distance in a relatively short period of time. We have grown used to the ability to connect a fairly inexpensive VCR to our TV, and to record programs on an inexpensive video tape. It is this low cost/high benefit situation that will keep what is called analog video in use for many years to come. Now analog technology has been surpassed.
The reason for this is that recording video digitally delivers remarkably better picture quality, sharper images and better color reproduction. And isn't that kind of improvement what we're always looking for? On top of that, digital copies of digital videos are unrecognizable from the original, which makes editing and image manipulation -- even at the level of the average camcorder user -- so much easier and with higher quality than that delivered by analog video technology.
Before we go any further, it is essential that we understand the difference between analog and digital video, otherwise the digital revolution will pass us by.
With analog video, light and sound are captured and recorded as electrical signals, transmitted as waves that can be represented by the up and down movement of a line. These signals look like mountain peaks and valleys, with variations in the height of the mountain and the depths of the valley, and variations in the distances between peaks and between valleys. With light, those variations are the differences between dark and very bright, as well as colors; and with audio, the differences are between no audible sound and very loud sound. Another way of looking at these waves is to imagine them as waves on the ocean -- infinitely variable -- going from dead calm to large waves. There is just as much variation in the electrical signals captured on analog.
The problem with analog recording is that it is hurt by interference which can reduce the quality of the electrical signal and make the recorded picture quality far worse than what was captured by the camera or VCR. Going back to our mountain analogy, the interference can change the height of the mountain tops and the depths of the valleys -- make them seem higher or lower -- changing the actual recording so that it no longer accurately represents the true image.
Analog video is also affected by timing errors, so what should be a straight vertical line, such as a telephone pole, will play back wavy instead.
Digital recordings don't have to deal with the wide variations found in analog recording. Digital recording is binary, with its electrical signals consisting of just two values, "on" or "off" ('1' or '0') -- there's a signal there, or there isn't. Should there be interference, while it may alter the strength of the "on" of "off" signal, the circuitry of the digital equipment can still tell whether the signal is "on" or "off" -- that's all it has to do. In a language of 1s and 0s, a message can be translated clearly. This makes digital recording almost immune to signal problems, and results in the highest quality picture and audio. This is a major advantage over analog. Digital is the language of computers. Computers easily store and transfer binary signals, from machine to machine, disk to disk, hard drive to floppy disk -- without distortion. It is exactly the same with digital video.
GO TO THE TOP
3. Digital Video - Where did it come from?
The digital video revolution is now in full swing. The boom in personal computers has created a voracious appetite for all things digital, but there have been other factors involved in this revolution:
1.technological advances -- giving manufacturers the ability to make digital video equipment;
2.more efficient manufacturing -- delivering more affordable products;
3.business and consumer demand.
Research on digital video and development of digital video products began many years ago, and by the late 1980s was progressing well on several fronts. In 1994, a standard was created for a recording format, and work on consumer digital video by several manufacturers was streamed into a single effort -- shortformed here to DVC (Digital Video Cassette). This immediately brought the efforts of more than 50 companies into focus, leading to the introduction of the first consumer DVC format products in late 1995.
While digital video originated as a professional technology, and continues to increase in importance in that sector, the latter stages of the 1990s will see its expansion into all corners of the prosumer and consumer arenas.
GO TO THE TOP
4. The DV advantage
4.1 Let's compare digital video to analog video
The image from a digital video product is a significantly better picture than that available from an analog video product. Digital video has approximately twice the horizontal resolution that can be produced by a standard VHS video cassette recorder. The resolution of a DV standard image is about 25% better than that from an S-VHS or Hi-8 camcorder or deck. While resolution is dependent on a products components and circuitry, standard VHS and 8 mm video are capable of delivering about 250 lines of horizontal resolution, with S-VHS and Hi-8 at about 400-420 lines. The DV format is capable of delivering over 500 lines on horizontal resolution (of course, actual performance will depend on the individual camcorder model).
Another way to look at digital video's image superiority is to note that an NTSC digital video signal contains three times the data of its analog counterpart, a PAL digital video signal contains six times the data of its analog counterpart . Digital video will deliver the absolute best consumer video quality. Interestingly, DVC has almost the same resolution as analog Betacam, a very popular professional video format -- an amazing quality jump for consumer equipment.
4.2 Color rendition advantage
Horizontal resolution is not all that goes into making a superior image; color resolution (or rendition) is also very important. Color rendition refers to the ability to accurately reproduce colors, without smear or blur. Analog can have trouble with color blur and color noise, but digital video does not. In a video image, color smear or blur is when, for example, the red of a woman's lipstick seems to smear beyond her lips, while color noise is indicated by random sparkles in the picture.
Because there is neither color blur nor noise, digital video delivers a far more life-like video image on the screen. This will be especially evident in images shot on a camcorder, and with images played on large screen TVs. What you will see is much sharper subject edges and clearer color reproduction.
4.3 Audio Comparison
DVC & Analog Digital technology has already made its impact in the area of sound with the acceptance of CDs (Compact Discs) in the mass market. People's demands, when listening to music or watching television are higher than ever. Their expectations for sound quality with Digital Video are equally as high.
The DVC standard ensures that audio quality matches that of an audio CD or, as you might guess, digital audio tape (DAT), since the audio portion of the digital video signal is itself recorded digitally.
Both CDs and DAT are superior to analog audio tape, since both CDs and DAT record digitally. As explained previously, analog recordings work with electronic signal variations (remember the mountain peaks and valleys analogy?). Not all the variations get recorded, and those that do are often subjected to interference, sometimes minor, sometimes major. Digital recordings, because they have to handle only "on" and "off" electrical signals, produce significantly superior sounding audio.
Most analog video tape recorders use analog sound. Digital video tape records digital sound, for significantly better audio quality than analog video. The digital video standard includes Pulse Code Modulation (PCM) audio recording, with two recording modes, either 16-bit stereo for highest quality, or two 12-bit stereo channels (total of four channels). Both modes deliver more than the normal, natural range of sound audible to human beings, ranging from quiet to the full output of a symphony orchestra, without distortion and without noise.
4.4 Compatibility and Copying
Digital video equipment is backward compatible, in that you can transfer a video from digital to analog equipment. That means that if you have a DVC recording you can copy it to a traditional VCR. All digital video equipment has analog video outputs (either S-video and/or composite video) found on current analog equipment, so you can play a digital video on a regular TV, or transfer a digital video onto an analog VCR. Most digital camcorders also have a digital output connector called IEEE 1394.
A DV cassette of either standard or mini size will not fit into any non-DV format's equipment. Cassettes used in other formats will not fit in DV equipment.
If you record using a digital video camcorder, you can then copy the video onto a VHS or Super-VHS VCR, 8 mm or Hi-8 camcorder, or any of the broadcast formats, such as 3/4-inch or Betacam, if they have the appropriate input connectors. Keep in mind, though, that because the digital signal must be converted to an analog signal, and an analog signal cannot contain as much information as a digital signal, there will be some loss in signal quality in the newly-made analog copy. Any further copies made from this analog copy will show the same quality loss as with normal analog copying. This copy can be made using S-video connectors or RCA-type connectors (some new DVC camcorders use a mini plug instead of RCA connectors). No special connections are needed to make this kind of copy.
Anyone who has copied an analog video tape (VHS or 8 mm, for example) will know that a copy (called second generation) made from the original will have worse picture quality than the original. Make a copy (called third generation) from the copy, and picture quality suffers even more. Each generation away from the original creates a progressively worse copy. Eventually, the image breaks up. How many generations away from the original this takes depends on the quality of the original and the equipment used.
The digital video signal is a robust signal, free of the problems encountered by an analog signal when it is copied. Using digital connecting cables (IEEE 1394), you can dub multiple generations without signal or quality loss. If a digital video signal is being transferred, say from a DVC camcorder to a DVC VCR, the signal does not go through any conversion process, it flows directly from tape to tape as a digital signal. The appropriate comparison is with copying computer files, since the data storage is digital in both instances. The same applies if the signal is transferred from tape to computer; the signal does not go through any conversion process, therefore there is no signal loss.
4.5 Copyright Digital video camcorders (at present) do not have an analog line in. That would let you record onto DVC from a VHS VCR or 8 mm camcorder. The reason is copyright. If you could copy to DVC, then you would be able to make any number of perfect digital copies from the copy, which has the potential for copyright violation.
DVC VCRs with an analog line-in connection are being manufactured with a copy prevention system that interacts with the copy guard systems in pre-recorded videos (such as those you can rent from your local movie shop). Videos you have made will not have this prevention system, so there will be no problems making copies from your home-made VHS or 8 mm movies onto DVC VCR.
Pre-recorded DVC tapes also will have a copyright protection scheme to prevent unauthorized copying.
4.6 Editing Overview
The digital video standard specifies certain system requirements that make editing digital video much easier than editing analog video. For example, digital video recording equipment must record specific data on the tape,
including a time code, an index of the start and stop points of each recording, the date and time, and photo print information. And, of course, DVC's greatest advantage in editing is that the DVC copies of DVC originals are exact copies with no loss.
4.7 Computer Connection
Perhaps the most intriguing aspect of digital video is the ability to transfer the signal to computer. Transferring an analog video signal to computer requires the translation of that signal into a digital form so it can be read by the computer. Depending on the quality of the equipment used this leads to varying degrees of quality loss. When the digital data on the computer is sent back to the analog VCR for recording, it must be converted back into an analog signal, causing further signal loss and a further reduction in quality. The final analog tape contains a video signal that is significantly poorer in quality than the original analog video.
This problem doesn't exist with digital video. Signal quality does not deteriorate regardless of how many times it is moved between tape and computer, even if the video is edited or manipulated (for example: special effects). Getting the digital video signal onto computer requires the use of the proper computer equipment and cables, designed specifically to handle digital video (IEEE 1394).
If you transfer the digital video on your computer to digital video tape, there will be no quality loss, but if the transfer is to an analog VCR, there will be losses, simply because the analog tape cannot handle the large quantities of data held in the digital signal.
GO TO THE TOP
5. The Physical Side of Digital Video
5.1 Cassette Design (smaller is better)
At the heart of the digital video format is a new cassette. In fact, there are two cassettes sizes. Currently, the first allows for recording up to four and a half hours (4h30min) of video, while the latter will hold a maximum of one hour of video. The cassette that is the main focus of our attention here is the second one, and it is called the mini-DV cassette (mini DVC). As its name implies, it is small -- 2.6 x 1.9 x 0.5 inches (66 x 48 x 12.2 mm). In comparison, an 8 mm cassette is 3.7 x 2.5 x 0.6 inches (95 x 62.5 x 15 mm). A VHS cassette is 7.4 x 4.1 x .98 inches (188 x 104 x 25 mm). The mini-DV cassette takes up less than half the overall space (43.4%) of an 8 mm cassette. The mini-DV cassette was designed for use in smaller, portable equipment such as camcorders, and the DV standard cassette for digital video VCRs and broadcast equipment.
The tape inside the cassette for VHS is 1/2-inch (12.7 mm) wide, 8 mm tape, as its name implies, is 8 mm (or slightly less than 1/3-inch) wide, while DV tape is 6.35 mm (1/4-inch) wide.
The tape cassette housing itself protects the tape until it is pulled into the record/play mechanism. The DV cassette uses a reel lock system to prevent tape sagging and other tape damage. The tape is therefore not only protected within a robust shell, but the tape also is wound neatly within the cassette for best alignment during both recording and playback. The locking tape door opens to expose the tape only when the cassette is in the machine, minimizing the entry of dust.
5.2 Cassette Labeling
The label on each cassette will identify it as a digital video tape, whether it is a mini or standard DVC tape, and show the recording time in minutes. So, a 60 minute mini-DV cassette will read: DV M 60. A two hour standard cassette will read: DV 120. This labeling is the same whether the tape is for NTSC or PAL recording.
As with analog tapes, professional quality digital video tapes will carry a different label. Only use tapes recommended by the camcorder or VCR manufacturer.
5.3 Record Prevention
Like other video formats, the standard and mini-DV cassettes have record-prevention tabs. These tabs are similar to those found in the 8 mm and Hi-8 formats -- the tabs are movable -- unlike the break-off tabs of the VHS cassette. In contrast to an 8 mm cassette, however, a DV cassette shows an open hole to prevent recording, while a closed hole allows recording.
5.4 ID Holes
Just as an 8 mm cassette has certain identification holes on its bottom surface, a DV cassette also has ID holes on the cassette bottom. These holes indicate tape thickness, cassette grade, and tape grade. The camera or VCR, using this information, will adjust its circuits to obtain a near perfect match between the tape and the recording/playback system.
5.5 MIC (Memory in Cassette)
One unique feature of the DV cassette is that you can buy cassettes containing a memory chip. This feature is abbreviated to MIC (memory in cassette). Cassettes containing the chip are more expensive, but allow greater convenience. Whether a cassette contains a chip or has ID holes will depend on the individual tape manufacturers and their marketing strategies. Cassettes with a chip will be marked to indicate this. The chip can record information such as a table of contents, date search, and photo search. There also is some room for future expansion of capabilities. The chips in professional cassettes have more memory than those in consumer tapes.
The data that will be written on the memory chip will depend on the equipment manufacturer, and may include information about the camcorder settings or lens settings, for example.
5.6 The DV Tape Itself
The digital video standard uses an entirely new tape structure and formulation. Although there are some similarities with certain Hi-8 tapes, digital video tape is superior, made to last longer, and better able to stand up to repeated use without failure. Because of the way the internal mechanisms of digital video equipment work with the tape, digital video tape is stronger than Hi-8 tapes, and much stronger than VHS tapes.
DVC tape is an advanced form of metal evaporated (ME) tape. While ME-type Hi-8 tape is the best in that format, the ME tape used in digital video is superior. The magnetic layer is double coated to give higher output and less noise. A hard layer of carbon is placed over the magnetic portion of the tape to give maximum protection and prevent the tape from wearing out over long periods of use. It's the same idea as applying black topping to a driveway; the black topping helps protect the underlying surface, helps prevent damage to it. As well, a new type of back coating is applied to the tape, reducing friction, providing more stable tape movement, and reducing jitter.
Digital video could not work without DVC ME tape. It allows the recording of a large amount of data, with higher output and lower noise, and protects the data with a protective coating and friction reduction. Digital video would be impractical without this new tape formulation.
Comparing tape widths: VHS tape is 1/2-inch (12.7 mm) wide, 8 mm tape is 8 mm (slightly less than 1/3-inch) wide, and DVC tape is 6.35 mm (1/4-inch) wide.
5.7 Handling precautions
The precautions for handling digital video tape are not that much different from any tape product (video or audio). The main ones are:
Day to day
1.Do not touch the tape
2.Always put the cassette in its case
3.Do not subject the tape to shock or impact (dropping it)
4.Do not expose the tape to strong magnetic fields
5.Do not leave in a car (because of the heat, cold, vibration)
6.Do not use the tape if it has gotten wet or has had anything spilled on it hours before using
8.Do not store the cassette in hot, humid or dusty locations
9.Do not leave the cassette in the DV recorder
10.Never disassemble the cassette
Long term storage
Follow the Day to Day rules plus
1.Always store the tape vertically, tape rewound, and the tape in its case
2.If you are storing for a long time, occasionally fast forward and rewind the tape.
GO TO THE TOP
6. Down to Digital Video Details
Let's rev things up and give you the full picture, the secret of digital video recording. Yes, the people who figured out how to handle digital video brought a lot of brainpower to bear on the problem of getting the signal on the tape, combining the sciences of mathematics, engineering and physics to get it to work. The joy of it is, explaining it in ordinary terms is fairly easy, now that they've got it all figured out.
Before we can begin our journey, some basics about video theory have to be covered. Remember that digital video is built on present day video technology. It will be easier to understand DVC if you understand how video works.
6.1 Video Primer Research into creating what we now know as televisionbegan in the early years of this century, but it wasn't until the 1940s that competing and incompatible broadcast technologies were standardized, and TV's remarkable growth began. This is the basis for all video display systems, including computer monitors and consumer digital still cameras.
There is an interesting similarity between the way television works and the way a movie is projected in a theater. When you watch a movie, you are actually seeing 24 separate pictures flash on the screen in front of you every second. But they appear so fast on the movie screen, your brain cannot process them as individual pictures. Instead, they come together as a single image, showing movement. If you have seen movies from the early days of the cinema, you will notice "flicker." The image on the screen flickers because movies were shot at 16 pictures per second, and your eye and brain can process these as just barely noticeable individual pictures shown on the screen.
A television picture is made up of 30 complete pictures per second in North America, 25 per second in Europe. In both movies and television (video), each individual picture is called a frame.
The picture you see on a television set is actually being projected on the back of the screen you're watching, but there isn't a movie projector inside your TV. An electronic device shoots an electron beam at the back of the TV screen, beginning at the top left corner (as you face the screen), continuing in a left-to-right motion all the way down the screen. Let's use the analogy of a artist setting up a canvas and having a spray can of paint. Pointing the can at the top of the canvas and spraying from right to left (remember, we're painting the inside of the screen, so everything is backwards), the artist stops spraying at the end of the line (the edge of the canvas). The artist then moves the spray can back and sprays another line under the first one, doing this again and again until the entire canvas has been "painted" with paint, line by line. Basically, that's how television images get "painted" on the screen -- but at a speed much faster than possible with the spray can of paint.
The TV's electron beam also varies in strength, depending on the brightness of the picture it is trying to "paint." The stronger the beam, the more light is seen on the screen. In essence, that's how a black and white TV works. A color TV uses three electron guns, one each for red, blue and green -- the primary colors. Each gun paints its own color. When all three guns paint the same area, you get white -- it's an additive color system.
When the TV standards were being adopted many years ago, everyone had their own idea about which standard to adopt. Some countries decided they would establish their own individual television systems. For that reason, there are now three major TV standards in the world, with several variations of those. And they are incompatible. In North America, Japan and some other countries, the television paints 525 lines (called scan lines) for each picture. This standard is called NTSC (National Television Standards Committee). The system used in England and much of Europe & Asia is called PAL (Phase Alternating Line), with 625 scan lines and 25 frames per second. France went its own way and developed SECAM (Se'quentiel Couleur 'a Memorie), also used in some other countries.
There is one important point that has to made here: In NTSC, of the 525 scan lines, only 480 lines are picture information, and in PAL's 625 scan lines, only 576 lines are picture information. The other lines are used for internal timing.
The NTSC system is also called the 525/60 system, PAL and SECAM are called a 625/50 system, the first number from the number of scan lines, the second . . . well, read on.
Remember the flicker of early movies? The same thing happened with television in its development days, but with TV, it was the top half of the picture that flickered. Even though the electron beam that was painting the picture did it very quickly, it just wasn't fast enough. By the time the beam was painting the bottom of the screen, the top half of the screen was starting to fade before it could be replaced by the next frame. The solution was to cut the picture in horizontal strips and number them (1 to 480 for NTSC and 1 to 576 for PAL), projecting the odd numbered strips first, then projecting the even numbered ones, doing this for every frame. It happens so fast, we can't see it. Each half a frame is called a field, or, putting it another way, two fields are interleaved to make a frame. This system is called interlaced scanning, and is used in every television system in the world. In computer monitors, and in the future High Definition TV standard, there is also non-interlacing (which is what TV started out as), made possible by almost 80 years of advances in electronics.
So, to get back to 525/60 and 625/50, the 60 and 50 refers to the number of fields per second (two fields make a frame and there are 30 frames per second in NTSC TV and 25 frames per second in PAL and SECAM).
6.2 Why will NTSC and PAL continue?
There are just too many television sets and VCRs in use worldwide for NTSC and PAL to be abandoned. As a result, digital video camcorders will be made in both NTSC and PAL forms. SECAM camcorders are not usually made, since it is easy to convert PAL to SECAM. This should all change if the television sets come with a digital input (IEEE 1394). The television set will do the conversion in a digital world. A good example of how this is happening, is being able to show computer information on a television set (web style television).
GO TO THE TOP
7. Journey from the Lens and Microphone to the Cassette
7.1 The Lens and the CCD - converting light to electricity
Now we can start the actual journey from the lens to the digital video cassette. What you are going to read here is an overview; some of the systems have been simplified for the non-engineer. As well, each manufacturer of DVC will have slightly different ways (circuitry) of getting the information to the cassette, but they all must follow this system.
(Note: There is more than one thing going on inside a DVC camcorder at any one time. There have to be adjustments made to the white balance, autoexposure, shutter and focus control systems. These systems will differ from manufacturer to manufacturer and are not covered by this book.)
What we see, and what the camcorder sees, is not digital. Remember, digital (binary) is either on or off. The light waves hit the camcorder's image sensor, called a charge coupled device, more commonly known as a CCD. The light waves are converted into electricity. The electrical signal varies significantly with the levels of light and darkness. This variation is what gives rise to the term "analog."
The electrical signal now goes through an extensive conversion process within the camcorder's internal circuitry, before being recorded on tape. Remarkably, despite everything that must be done (read on), it is almost instantaneous, such is the power of the computer circuits within the digital video camcorder.
7.2 Conversion to Y, R-Y, B-Y
Circuits inside the camcorder change the electrical signal coming from the CCD into black & white (Y) and color (C) information -- Y/C.
Next comes a circuit that changes the Y/C into Y, R-Y and B-Y -- the signal is now the black & white information, red (R) minus the black & white information, and blue (B) minus the black & white information. The black & white signal is considered to be the most important part, because it contains the detail and resolution information. This is the same system as used for broadcast TV, and minimizes color interference.
Now it's time for the analog to digital (A/D) conversion, which is in three stages: sampling, quantization and binary notation.
7.3 Sampling
In sampling, the signal is divided into millions of sections. The black & white component is sampled 13.5 million times a second (13.5 Mhz). That means it is divided into 13.5 million sections. The color signals are sampled 3.375 million times a second (3.375 Mhz).
A useful analogy is to imagine you have set a film camera and a tripod beside a street. If you took a picture once a minute, the sampling rate would be once per minute. If a car drove by at a time other than when you took a picture, the picture would not show the car. If you increased your sampling rate to once per second -- taking one picture a second -- you would capture a fair recording of the activity on that road. If you increased the sampling rate even more, to 1,000 samples a second, you'd get an even more accurate recording of the activity on that road.
Therefore, a sampling rate of 13.5 million times a second is designed to give an accurate rendition of the activity of the black & white image. The black & white component is sampled four times more than the R-Y and the B-Y signals. This is called 4:1:1 video -- the black & white sampling ratio to the R-Y sampling rate to the B-Y sampling rate. Expressed in another way, the black & white signal is sampled 13.5 Mhz. If you divide 13.5 by 4 you get 3.375 Mhz, which is the number of times per second the colors are sampled. In PAL format, the R-Y and B-Y are sampled differently because of the design of the PAL system. The frequency of the sampling is 6.675 Mhz, but the effective sampling rate in a frame is the same as NTSC, it just gets there in a different way.
(There is also 4:2:2 video, which is used in broadcast. Because of all the stages a broadcast signal has to go through, including transmission, the color components are sampled at double the rate of 4:1:1 video, producing a higher quality original signal, better able to withstand the rigors of broadcasting.)
7.4 Quantization
The quantization process puts numbers to the sampling. Going back to our original analogy of mountains and valleys, the quantization process gives us numbers that tell us just how high the peak is, and how low the valley is, along with everything in between.
How many possible shades of gray, or colors, are needed to accurately translate to video what our eyes can see in the original image? Our eyes see an almost infinite number of shades. This would be impossible for a video system to handle. As with most computers and digital still cameras, 256 shades is the number required. So our black & white signal has a total possibility of 256 shades, as does the R-Y and the B-Y components. In computer terms, this is called 8-bit quantization (although 256 shades is equally accurate).
The video signal is now represented by three strings of numbers, one for Y, one for R-Y and one for B-Y. The Y component has numbers appearing 13.5 million times a second, with the other two 3.375 million times a second each. Each time a number appears it has a value between 0 and 255 (in a computer "0" is a number, so there are a total 256 numbers).
7.5 Binary Numbers
The binary stage of this analog-to-digital conversion changes the quantization numbers into the string of zeros and ones (or "off" and "on" electrical pulses) that are the digital signal.
The numbers that have come out of the quantization process are decimal numbers and computers cannot handle them, being able to use only a "0" and a "1" ("on" or "off" in electrical terms). So anything involving computers (and there's a "computer" inside every digital video product) has to be converted into binary notation.
Every decimal number has a binary equivalent. For example, the number "2" is 00000010, the number "9" is 00001001, and the number "17" is 00010001.
7.6 Compression
The process so far has created an incredible quantity of digital information (data). Each second of video is the equivalent of 162 million bits per second (Mbits) of data, with each bit either a '0' or '1.' The signal has to be compressed, otherwise you'd need a cassette the size of a truck to hold 60 minutes video.
Video compression means the removal of data using mathematical formulas. The digital video standard is no exception, employing several mathematical functions to undertake this process. These are Discrete Cosine Transform (DCT), Adaptive Quantization (AQ), and Variable-Length Coding (VLC). It is not necessary to know the mathematics behind these operations, but it is important to know what the effects of these functions are.
7.7 Flanking Area Removal
The video signal captured by the camcorder actually contains more data than is shown on the screen. This extra material is called the flanking area, and it contains none of the picture. It is a spare area around the picture used for timing in the original television standard. In the digital video compression process, this extra material is discarded immediately because it is not needed, so that the compression schemes only have to work on 124 Mbits of video data per second, instead of the original 162 Mbits.
7.8 DCT Blocking & Shuffling
The video data stream is now divided into small portions, measuring 8 x 8 pixels (picture elements). This divides the video frame into very small individual blocks or boxes (8,100 for NTSC, 9,720 for PAL). These blocks are grouped together in 5 columns, called macro blocks.
The macro blocks are now shuffled, or moved around. The reason DVC does this is to even out the data in the video image so that there is an even flow through the rest of the compression circuits.
Imagine there is a single lane of traffic with 10 cars on it. If the road opens into five lanes side-by-side and all the traffic (data) remains in the middle lane, traffic flow will be impeded, held up. But if we number each car, and spread them over the five lanes, it doesn't matter what lane we put them on. The cars can now move much quicker, and when the five lanes reduce back down to one lane, it is very easy to ensure the cars will continue in the right order.
The video image is shuffled to a set of rules (so it's the same every time). The data goes through the rest of the compression process and then the data is deshuffled back to its original position before it is recorded on tape.
7.9 DCT
DCT is a very complex mathematical process. It does not compress the image, it changes the data into a form that can be easily be compressed. The following is a simple analogy that conveys the essence of this process. Each individual DCT block is processed separately. All the information about the top left pixel is stored complete. Then for the next pixel, only the difference between it and the base pixel (upper left hand corner) is stored. For instance, if the video image is a blue sky, the difference between the first pixel and the next one would be '0'. This continues for all the pixels in the block, with only the differences being stored. If there was some detail in the lower half of the block, then bottom pixels might be different from the base pixel by '-1' or '+3,' for example. (Technically, DCT stores the DC coefficient of the block in the base pixel, and the other pixels are stored as AC coefficients.) After this process is finished, the values of easch pixel in the DCT block are weighted by using specific values. These values do not depend on the image, they are present in the DVC system.
7.10 AQ & VLC
This is a very important stage. So far all that has been done is to represent all the information in a different way; no data has been removed. Adaptive Quantization (AQ) and Variable Length Coding (VLC) take hold of the data and do the actual data compression.
The compression ratio of DVC is 5:1 (the video information is squeezed to 1/5 its original size). The circuit first "guesses" at how much compression it needs to use for each DCT block. Areas with not much detail don't need very much compression and the areas that have a lot of detail need more compression. This is an adaptive process that changes with the video images.
The compression circuit tries to maximize the number of '0s' in the video data. Instead of storing a string of zeros (ie '000000000000000') the DVC stores a phrase that says, in effect, "15x0," so the data takes up less room.
Of course this won't happen very often, so the compression circuit tries to round off all the values from '-1' to '+1' down to '0.' If the circuit can get a 5:1 compression ratio out of this, then that is what it uses to compress the image. If the data stream is still too large, the circuit will try to round off all the values from '+2' to '-2' down to '0.' This process continues until the proper amount of data compression is reached.
Needless to say, this is an oversimplification, because this process is actually very complicated.
So, by the end of this compression process, only about 25 Mbits per second are going to be recorded on tape.
7.11 Deshuffling
The original shuffling process is reversed -- data is deshuffled (put back in the correct order) -- before being recorded on the tape, otherwise the recorded image could not be viewed in fast forward. The image appearing on the screen would look like a jigsaw puzzle with the pieces in the wrong place.
7.12 Error Correction
In the error correction system, extra data is added to the video data. This new piece of information lets the playback system double check that it has "seen" the data properly. If an error does occur, the playback system will "see" this has happened and will correct it. This is similar to the type of error correction used in computer modems to check that data has been transmitted and received properly.
7.13 Playback
To play back an image recorded on tape, this entire process is put in reverse (including generating a new flanking area), ending up in an image that can be viewed on a TV screen. This is only possible because the DVC recording process follows rigid rules, which can accurately translate the data back to a video signal.
7.14 Audio recording Overview
When sound is captured by a microphone on a digital video camcorder, it is converted into a digital signal. This includes both sampling and quantization, which, as with video data described previously, convert the analog signal into numbers. Interestingly, the audio data is not compressed, since there is not as much of it as video data. The audio is then shuffled, error correction is added, and then recorded on the DVC tape.
Audio signals can be sampled at 48 KHz (48 thousand per second), 44.1 KHz or 32 KHz and can either be 16 bit or 12 bit quantization. Of note, some digital video camcorders will offer just 16-bit recording, others just 12-bit recording, and some will be switchable between the two modes.
The higher the sampling and quantization rates, the better the data represents the original sound.
7.15 16-bit sound
The term "16-bit" refers to the amount of data recorded and the range of that data (16-bit converted to decimal numbers means that there are 65,536 different numbers that can represent any sample). In DVC, 16-bit sound is the highest quality sound, with the most data for the truest and fullest range of sound.
The sampling rate can be 48 KHz (48 thousand per second), 44.1 KHz or 32 KHz. The choice is up to the manufacturer of the DVC camcorder or VCR. DAT uses 48 KHz sampling and CDs use 44.1 KHZ sampling.
The DV format standard allocates a fixed amount of space on the tape for the sound data, and 16-bit sound fills the entire space -- it has more data and needs all the space.
7.16 12-bit sound
With 12-bit, the sound quality is just slightly lower, because the amount of data gathered through the sampling and quantization procedures is lower than with 16-bit (4,096 variations for 12-bit compared to 65,536 for 16-bit). The 12 bit sound is sampled at 32 KHz. Because there is less data, it does not take up all the space available to audio on the tape. Because of this there are two two-channel channels.
The benefits of having 12-bit recording are obvious to those who demand more of their equipment: This will let you use the microphone mounted on the camcorder plus up to two remote mikes, all recording on separate channels, at the same time. At the editing stage, selections can be made for the desired mix for the finished video. Or, if only two channels are used, the original audio can be left on the tape while new audio (music and narration, for example) can be added without erasing the original sound during the editing process.
GO TO THE TOP
8. The recording/playback system
There's nothing wildly different about digital video's record/play mechanism when compared with, say, a VHS system or an 8 mm system. As with the analog formats that came before it, digital video tape is moved past a rotating cylinder (drum) which contains the recording/playback heads. That cylinder is angled so that when the heads place the data on the tape, they do so in tiny rows that are on an angle to the direction of travel of the tape. These rows are the data tracks, and the process of placing the tracks on the tape at an angle is called helical scanning.
But that's not to say that there aren't some mighty important differences between digital video and analog video recording -- because there are.
First and foremost, because so much data must be transferred to the tape, the record/play drum rotates at a speed of 9,000 revolutions per minute (rpm). In comparison, the record/play drum in either an 8 mm or VHS system rotates at what seems a more leisurely 1,800/1,500 rpm (NTSC/PAL). The drum itself is about half the size of an 8 mm cylinder and about one-third the size of a full-size VHS drum.
Tape moves through a DV system at a speed different from that of other formats, although it is not as wildly different as that of the drum rotation speed. Tape in a DVC machine travels at a rate of almost 18.9 mm/second. In comparison, tape in an 8 mm machine travels at about 14.4 mm/sec. For VHS, the number is 33.35 mm/sec.
We've noted that the recording mechanism places tiny rows of data on the tape. The width of each one of those tracks on a DVC machine is 10 um (microns) compared to the 20.5 um of an 8 mm system or the 58 um of a VHS system. Now you begin to see there is something amazing at work here.
A DVC machine will place 10 tracks (one track for each 48 video scan lines) for one frame of NTSC video, and 12 tracks for PAL, in a space on the tape that's the width of a human hair. Remember, with digital, all that has to be placed on the tape is the equivalent of a "0" or a "1" -- in electrical terms, off or on -- which is why the track can be so small compared to an analog signal.
Four types of data are recorded onto the digital video tape: ITI data, audio data, video data, subcode data. Separating each type of data is a gap containing no data, but the gaps do contain signals used for tracking. As noted previously, when editing, audio and video data can be overwritten. The gaps prevent data destined for one data area being written in another.
Here's more information about what is recorded in each track:
1. Insert and track information (ITI) -- basically an index to where everything is in this track. The track information confirms the track width specifications, and that the track mode is SP or LP. The insert information is map of where the video, audio and subcode data are recorded.
2. Audio -- audio data is placed here, along with auxiliary sound data, such as the recording time and the recording mode.
3. Video -- video data goes here, along with auxiliary video information. The auxiliary data is optional and can include the recording time and date, whether the video was shot in wide mode, and the input source (i.e. which TV channel). As well, it can include camera information and other data, as the equipment manufacturer decides.
4. Subcode -- used for the time code, index ID (keeping track of where various scenes start), and photo mode ID (necessary for still picture recordings/playback). It also can be used for text data, such as superimposed text.
5. Preamble -- this is recorded at the start of each type of data, and it guarantees that the machine can lock onto the signal.
6. Postamble -- prevents any erased data from remaining on the tape when previously recorded material is overwritten.
7. Gaps -- these occur between each type of data and guarantees the data will not overlap.
Auxiliary data and subcodes also are not compressed, and are added to the data prior to recording.
GO TO THE TOP
-
9. Digital Video and Professional Use
There are many different video cassette formats currently in use. There are different cassette formats within the broadcast and professional spheres (Betacam and 3/4-inch being just two), with VHS and 8 mm for home use. All these formats use cassettes of different sizes.
In digital video, there are two cassette sizes: a larger cassette called 'DV standard' and a smaller one called 'Mini-DV.' Unfortunately, despite cassette size standards, the professional world already appears to have abandoned recording format standardization within the digital video area, coming up with different recording formats using the DV cassettes. Consider this: cassettes recorded in one professional digital format cannot be read by the DVC machine of another manufacturer, and in some cases cannot even be read by different models from the same manufacturer! Very confusing. DVCAM is different from DVCPRO, which is different again from DVCPRO 4:2:2, even though they all use the DV cassette. One of the main differences between the formats is tape speed, although there also may be differences in data rate, audio and other recorded information.
GO TO THE TOP
10. DV & DVD comparison
How does the Digital Video Cassette (DVC) compare to what was first called the Digital Video Disk, but which is now known as the Digital Versatile Disk (DVD), introduced to the consumer market in early '97, which also has improved picture quality over analog video formats?
DVD uses quite a different technology. Obviously, it's a disk, and like a music CD or computer CD-ROM, it offers superior random access to information stored anywhere on the disk. But a disk cannot hold nearly as much information as tape, which is why DVD uses a much heavier compression scheme than DVC. On top of that, tape is far superior for time-shifting than a disk, which is very limited in what can be recorded on it. The data that is stored on the disk is placed there physically; the actual physical surface of the disk is altered. The data is read by a laser beam.
While DVD was designed originally for playing back prerecorded movies (manufacturers are working on a standard for recording on DVDs), DVC was designed as a portable record/playback medium. This becomes evident when still frames are viewed. Because of the compression schemes used, DVC has the best still frame images, while a DVD still frame will not be as clear. DVC and DVD data compression systems are explained in the technical appendix at the end of this book.
11. DVC & Computers
It's easier to move large quantities of water through a fire hose than it is through a drinking straw. That analogy will help you understand at least part of the idea behind the subject of connectivity, specifically as it relates to digital video.
Keeping with this water analogy for a moment, imagine if you had to put out a fire, and the water had to move through several different pieces of equipment. Some pieces use a garden hose to carry the water, others use a drinking straw, and yet others use a firehose. Wherever you hit a straw, the flow of water would slow right down. Imagine how much water would be getting through to the fire hose at the end. Not much.
Moving into the electronic world, not all connections are alike. Some are able to pass small quantities of information, while others can handle much larger quantities of data.
A new standard has arrived on the scene, and it allows manufacturers to standardize on the connections and cabling (interface), regardless of the type of equipment, and move remarkably large quantities of data between many different kinds of products. This is especially important with digital video, which is data intensive. Digital video equipment must handle very large quantities of data, far more than analog equipment.
This international standard is known officially as the IEEE 1394 High Performance Serial Bus, but is most commonly referred to as IEEE 1394. IEEE is the Institute of Electrical and Electronic Engineers, an organization which has helped create many international standards, of which 1394 is just one. For example, IEEE 1284 is a parallel printer cable standard.
IEEE 1394 will let you connect a VCR, TV, computer, DVD player, computer, printer, camcorder, CD player, and audio amplifier (and who knows what else in the future) that have the proper connections -- and transmit data between them all. Huge quantities of data. Very quickly.
11.1 Why another interface?
Over time, any number of connectors and cables have found their way into common use, either because manufacturers decided to hop on the bandwagon and use something that already was being used, or a group of manufacturers got together and decided to create a new standard.
However, over time, as technology advances, a connector and cabling may prove to be a bottleneck -- although everyone uses it, it can't keep up with the changes in demands placed upon it. It is at this point that manufacturers either grasp for the already existing device that's on the parts shelf, or create a new one. If too many manufacturers go their own way -- act independently -- then you have a problem for the end user who cannot interconnect equipment from different makers.
Simply stated, technology has now reached a point where massive quantities of data have to be moved between devices, yet the connectors in common use cannot do the job. Take the common RS232 connector, for example. It can move 20 Kb per second. The Universal Serial Bus will hit a top transfer speed of 12 Mbits per second. IEEE 1394 will move 100 to 400 Mbits per second, and likely more with further development.
Looking at these speeds from another perspective, when working with large data transfers, computer software companies suggest you should use a SCSI hard drive, because of its rapid data flow. Yet SCSI will handle just 5 Mbytes. Even Ultra SCSI will move a maximum of 40 Mbytes.
In the case of IEEE 1394, an international standards organization has developed, in collaboration with representatives of several industries, a new standard electronic connection device and cabling, to meet both current and projected demands. One of those demands is the ability to move the massive quantities of data used in multimedia applications, and digital video in particular. That standard has gained acceptance among computer, video and audio manufacturers.
11.2 Origins and development
You may have heard IEEE 1394 referred to as FireWire (TM). This is not quite correct.
IEEE 1394 is based on Apple's FireWire (TM) bus and connection system, but it is not the same. Apple is now one of the many companies supporting this new standard.
IEEE 1394 was accepted as an industry standard in late 1995 by the Institute of Electrical and Electronics Engineers (IEEE), after more than a year of effort by a group of computer and consumer electronic manufacturers to speed up its adoption.
More than 50 companies which make broadcast and consumer video equipment, and which have adopted the digital video (DVC) format, also incorporated IEEE 1394 as the standard digital audio/video interface for all DVC equipment.
11.3 Using IEEE 1394
Who will use IEEE 1394? The answer appears to be: everyone. One of the reasons for this is because IEEE 1394 cables and connectors are simple and low cost, so they can be used just about everywhere electronic equipment can be connected. Some have noted that IEEE 1394 products may be less expensive than the cables and connectors they replace.
At the consumer level, work is underway to have IEEE 1394 connections for all manner of electronic devices in the home, ranging from TVs, camcorders and VCRs, to audio components and video games, along with computers and their peripherals. Home security systems and electronic music instruments (MIDI) are another possibility.
Digital video camcorders found their first homes with the prosumer, those who want higher quality without the need for full broadcast capabilities and cost. The companies supporting the digital video standard included IEEE 1394 as part of the DVC standard. Video editing and audio mixing/editing equipment are naturals for IEEE 1394 interfaces.
Almost from the setting of the standard, the major broadcast and professional equipment manufacturers announced that their digital video equipment would incorporate IEEE 1394 connections. Audio equipment manufacturers will follow suit.
11.4 Making the 1394 Connection
As previously noted, the IEEE 1394 cable and connector are simple and inexpensive to make. The cable uses standard wiring, although the configuration is unique. There are two types of cables: one contains two power lines and two data lines, the other just two data lines. The data lines are shielded, and the entire cable is shielded. Nothing really out of the ordinary.
What's fun is that the connector is derived from Nintendo's GameBoy (TM) connector, which means it has been field tested by countless thousands of people, and found to be very durable. The cables are easy to use, even when you are making connections when you can't see the plug (in the back of a machine, for example). Just as there are two different cable types, there are two different connector types, one with four pins, the other with six pins. There will be adaptors allowing the two to work together.
11.5 Making 1394 work
  
IEEE 1394 creates an environment where as many as 63 devices can be connected without using anything other than the cable and connectors. A piece of equipment does not have to be turned on for the next product in the line to take advantage of the connection. And, products can be plugged into the line, or unplugged, even when the bus is in full operation, without causing any problems, and with automatic recognition of the connection or disconnection of that item. This is known as hot plugging.
The standard is a combination of hardware and software that allows for the movement of 100 to 200 Mbits (megabits per second) of data. Future versions, capable of transmitting 400 Mbits and 1.2 Gbits (gigabits per second), are simply waiting for the availability of special integrated circuits (chips) which can handle that data load.
IEEE 1394 will handle, for example, two simultaneous channels of broadcast quality, full-motion (30 frames per second) video and CD-quality stereo audio.
To describe how multimedia applications such as full-motion video are handled, imagine a post office sorting system. The IEEE 1394 device that is sending the video reserves a certain amount of space on the conveyor belt for "boxes" of data equal to 512 bytes. The number of "boxes" will be determined by how much video is to be sent. While the "boxes" are placed on the conveyor belt at equal intervals, the space between each "box," although very small, is still large enough to contain occasional instructions such as "stop" or "play," for example. Each "box," as it is placed on the conveyor belt, is labeled with the address to which it is going. Depending on the speed of the sending and receiving devices, more than one video signal can be sent along the conveyor belt.
The receiving device sends a note back to the sender saying that the package was received. If the parcel arrives damaged, error correction is started. This is done for every parcel of data.
12. You've Made It!
As was said in the start of the book it was our aim to bring the subject of digital video to you in a clear and concise way. We hope that we have succeeded.
The digital video revolution has just begun - it promises to be an exciting time. The better you understand digital video and are able to enthusiastically convey its potential to your audience, the more exciting it will be for you.
Enjoy the Digital Video Revolution!
GO TO THE TOP
Excerpted from: http://www.canon.com
This article was acquired on the "fair use" basis.
We encourage You to visit the source website for more information on this topic.
|
|
|
