♣No More Full-Power Analog Steakhouses After June 2009♣

"For every complex problem, there is a solution that is simple, neat, and wrong."  -H.L. Mencken

...Super-Massive Black holes inhale entire galaxies. Gamma-ray Bursts release more energy in the blink of an eye than our sun will produce in the next billion years, wiping out countless alien civilizations. Supernova explosions scatter elements, like cosmic pollen on the stellar wind. And yet, an even more dramatic event swept across the US in June 2009→

971 full-power TV stations halted all broadcasting in analog. Forever.

At a time of the station's choosing on 12 June 2009, Federal Law mandated that all full-power television stations cease broadcasting in the NTSC ANALOG over-the-air format that they used as far back as the '40's, and that they instead broadcast using ATSC DIGITAL (or not broadcast at all, as was the case of 35 of the 971 stations). No more NTSC analog, just ATSC digital. And as 12 June 2009 passed, only digital TV broadcasts eminated from all full-power US TV stations.

“No more analog for me, Mom.” ↔ Michael Fitzsimmons, “Peggy Sue Got Married”, 1986.

Whether or not a given program is broadcast in High Definition, all programs now are broadcast using Dolby sound. The sound is broadcast with anywhere from 1 channel (Mono) up to 6 discrete channels (called "Dolby Digital 5.1"), and it has nothing to do with High Definition. And that's our very first HD flyover.

NOW... This is a critical concept that you must understand 100%. (Even before we really explain "High Definition".)→

Neither Congress nor the FCC required any TV station to broadcast in "High Definition". Congress only required that all full-power TV stations cease broadcasting in NTSC analog on or before 12 June 2009; and if full-power TV stations still wanted to broadcast, they had to broadcast using ATSC Digital standards only.

High Definition and Digital are NOT the same, any more than steaks at Applebee's and at Peter Luger's are the same. Digital TV stations MAY broadcast High Definition pictures, but only if they want to; in fact, in the US, High Definition pictures must be broadcast using ATSC Digital. But whether any given Digital station broadcasts a given program in High Definition (or in Standard Definition, the "opposite"), that is the station's call.

Again → DIGITAL ≠ HIGH DEFINITION→

Nor do you need a TV set that's capable of displaying High Definition pictures in order to watch Digital broadcasts.

In fact, the FCC has not even established any standards for High Definition.

♣An Inconvenient Truth♣

“The Chief Cause Of Problems Is Solutions.”   -Sevareid's Rule

ATSC Digital (which still usually uses 8-VSB modulation) doesn't work terribly well. Digital signals must blast in over-the-air at full strength, or they don’t show up on your TV at all. Digital signals (especially those transmitting on channels 2-13) don't travel as well as analog signals did; ATSC digital is inherently weaker than NTSC analog→

“Using a large rooftop antenna about 50 miles from the transmitters and their antennas in Los Angeles, we did a 'shoot-out' on station KCET, channel 28.

“Using an RF splitter, we put one TV on analog; the other split went to a Channel Master CM-7000 DTV converter box.

“Side-by-side, the digital picture and sound were superior. Then the UHF signal on channel 28 began to fade. The analog TV displayed some snow; the digital picture remained unchanged.

“BUT... When the signal faded a bit more, the digital signal lost it big time, with ribbons of color and stuttering sound (pixelation). Our friends at CBS refer to this as falling off the digital cliff. The analog TV was still watchable, just with a bit more snow. While analog signals fizzle out over long distances, digital signals tend to become completely incomprehensible at a certain falling-off point.

Conclusion→ “The Digital TV experiment was not yet ready for prime time. It was in development for over 10 years, and it just does not work well in the real world.”

—Bill

Thank you, Bill.

“With digital, you get a great picture or you get nothing,” said Mark Wigfield, a spokesperson for the US Federal Communications Commission. (In 2004, a more "robust" form of digital modulation, E8-VSB, was approved by the FCC, but virtually no digital TVs (and no converter boxes) presently (2010) can receive E8-VSB modulation.)

While the traditional analog broadcasts (since the '40s) would become snowy as signal strength decreased, digital transmissions often completely fail with even a slight drop in signal strength. This means that many viewers who experienced a bit of static with some weak analog broadcasts are not receiving some digital signals at all.

Note that UHF is now channels 14-51; 51-83 (less four channels for public safety uses) are going to mobile phone companies. Note also that digital stations have TWO channels→ A real channel that they actually are broadcasting on, and a "virtual" channel that shows up on the tuner of the TV set, a channel that you think you're watching. Sometimes the two are the same; sometimes they're not. For example, in this area, you think that you tune to VHF channel 2 to receive station WMAR; however, through the wonders of digital TV, you really are tuning to UHF channel 38.

VHF broadcasting uses less electricity at the transmitter than UHF, and VHF equipment is less costly; but VHF electromagnetic waves don't penetrate buildings as well as UHF, according to Russ Abernathy, director of television and technology at WKNO (PBS, Memphis, broadcasting on real UHF channel 29).

Said Joel Kelsey, a policy analyst at Consumers Union, publishers of Consumer Reports, “These are people who have taken action early, done everything the government told them to do, and now they find themselves ... with less service than they had before. This is clearly an issue for a lot of consumers — many more than the Federal Government had anticipated.”

The Ugly Side Of ATSC Digital→ Digital isn't like it's late Cousin Analog... Digital has an ugly side. If you corrupt any of the millions of bits within a digital TV frame, the whole image will be corrupted. You'll see a perfect picture, or you'll get nothing at all; or worse, you'll get the dreaded pixelation of both picture and sound.

Cousin Analog was a different beast. As long as you received some signal, you could get some a picture. As the signal strength degraded, so did the image; but there was a wide range of quality, and analog TV didn't have to be perfect to be watchable. Occasional snow or lines didn't really make the image unwatchable. And they didn't affect the sound.

If you understand how digital TV works, you knew the day was coming. The public was sold a whitewashed tale of a digital TV world where you'd get more stations in crystal clear perfection. The truth is that a large percentage of the population has lost channels. The fact that, now that full-power TV stations all have switched to digital-only, folks may not receive all the TV stations they received in analog has, of course, been swept under a carpet of propaganda and minimized.

That's great news if you're a cable or satellite company. Cable providers have been big supporters of the digital TV transition for a long time.

THE BOX→

All converter boxes are NOT the same. Most boxes subsidized by the NTIA (the US government) are "bare bones" so that they can be sold cheaply, and it is your box that actually has to receive the digital transmission. If you're having reception problems, try a different box.

Think of the transition from analog to digital TV as analagous to the change from vinyl records to CDs... or from videotape casettes to DVDs... or from film cameras to digital cameras. It's a more "interesting" way of transmitting TV programs from broadcast stations.

But, just as you can't play your CDs on a turntable, you can't watch a digital broadcast with an analog TV set. At a minimum, you'll need to add the now famous "converter box" to your analog TV... a relatively inexpensive ($40-$90) gismo (parts plus assembly = about $25) that "down-converts" the new digital signals reaching your antenna back to the old analog signal, before passing it along to your analog TV.

In other words, with a converter box, you and your TV are exactly where you were before digital-only came along; except you had to get a $40 subsidy coupon and usually add a few dollars to it; and you may lose some stations you watched prior to digital-only.

These converter boxes still are being sold at many retailers; e.g., Best Buy. The boxes themselves, from what Dr. Steve says he can see inside, have roughly $10 worth of parts in them... add the labor (perhaps 20 minutes) to put the boxes together... and you have a box that costs $25 to make... and that is selling above the price of the $40 subsidy coupon. Wal-Mart price→$50; Best Buy price→$60. Whatever... if you can't bamboozle the customer into buying an expensive new HDTV set, you still can make a nifty profit by selling millions of converter boxes. Radio Shack did.

The price of converter boxes will slowly decline after application for coupons ends on 31 July 2009. There really is not a lot of stuff inside the boxes.

A question we frequently receive→ How come folks didn't need a converter box when TV stations started broadcasting in color in 1953? Answer→ When color was added to TV in late 1953, it was "compatible" color. B&W sets could receive color broadcasts; and color sets could receive B&W broadcasts. However, even though digital uses the same 6 MHz channels as the old analog broadcasts, ATSC digital does things to the TV signal from the get-go, things that must be reversed in the TV receiver; converter boxes and digital sets perform these reversals; analog sets don't.

In a nut-shell, digital requires more than the 6 MHz channels available; to get it to fit, many compression techniques were applied. These must be reversed in a digital receiver or a converter box; they cannot be reversed in an analog TV set.

Are we saying that the folks who designed color were smarter than the folks who designed ATSC digital? Color is compatible primarily because of the work (invented and patented in 1938) of a brilliant French engineer, Georges Valensi. Let us simply say that there was no equivalent of Georges Valensi around when ATSC digital was strung together. Nor was there any incentive for compatibility of digital and analog.

WILMINGTON- THEY KNEW→

If you lived in the Wilmington, NC, area in September 2008, five of your local TV stations "jumped the gun" and began broadcasting only in digital as of high-noon on 8 September 2008. Kind of a "Let's see if this whole idea really holds any water".

This region, selected by the FCC, was highly affluent, among North Carolina's wealthiest... and affluent areas have less dependence on over-the-air TV reception with antennas than do lower-income areas, where there often are language barriers and minimal Internet access and lots of rabbit-ears.

And since 66% of households in the North Carolina test area watched TV stations that were outside the test area, and since not even all of the TV stations in the test area ceased broadcasting in analog, and since the test area is coastal and flat and has little correlation to digital TV in hilly areas... the test really just was a video game that the FCC was playing to look good. But even with all the disclaimers, the FCC and Congress were shocked by the results.

How Did The North Carolina Test Go?→

When the analog signal from Wilmington's NBC affiliate WECT was shut off, many viewers outside the "official digital market" for WECT, viewers who had been able to receive WECT in analog, were not able to receive the digital broadcast from WECT. When WECT told callers there was nothing that anyone could do to restore their reception, the former WECT viewers blamed the Government.

SO... One factor that emerged from the Wilmington test was a shrinkage of out-of-market coverage. The old analog signals simply reached farther than the new digital broadcasts. The analog signal of Wilmington's WECT reached as far north as Raleigh, NC, and as far south as Myrtle Beach, SC. After WECT went digital-only, some viewers who were cut off were able to get the NBC network from other affiliates closer to their homes, but others could no longer receive an NBC affiliate station at all.

It was not clear what the FCC or broadcasters could do for these viewers, short of recommending that they buy a bigger antenna or a more directional antenna (perhaps on a rotor). FCC Chairman Martin subsequently told members of the Senate Science and Transportation Committee that a possible solution would be for TV broadcasters to erect special "repeater" antennas (and install additional digital transmission equipment) to expand their reach. This distributed broadcasting would require precision in transmission and reception that bordered on (or even exceeded) the technological limits of 2010 TV technology.

Distributed Transmitters→ Actually, with digital broadcast equipment, multiple transmitters are not quite as absurd as they seem on the surface. Now that the ATSC has approved the A/110 Synchronization Standard, the standard for Distributed Transmissions, many stations are wondering if distributed transmission (often called DTx) using multiple smaller transmitters might work better than a single high-power transmitter. DTx systems use two or more transmitters, each transmitting exactly the same signal at a very specific time. Reception, however, is not trivial; it works best where the viewership is divided into two or more areas... perhaps by some feature like a mountain range.

Digital TV's Cliff Effect was another problem revealed by the Wilmington test, affecting about 15 percent of Wilmington residents who still used analog TVs. Even with subsidized converter boxes to convert digital signals to analog, about a third of these also needed new antennas to get a signal strong enough for their converter boxes.

A Note On Antenna Rotors→ Don't try to use a simple rotor with a digital TV or a converter box.

The rotor will drive you crazy trying to tune in the best direction, because by the time the digital picture appears and locks in, you already have spun the antenna past the optimal point. It’s called the Digital Delay, and it’s not fun.

After June 2009→ Some time after June 2009, many broadcasters are expected to adjust their digital transmissions by→

• Making them stronger, increasing their effective radiated power; and/or...

• Placing their transmitter antennas on higher towers.

So folks with analog sets and converter boxes may want to wait a while after the June 2009 analog cutoff before investing in an expensive rooftop antenna. The digital transformation was expected to result in 89 percent of TV stations reaching more (but possibly different) viewers, while 11 percent were expected to reach fewer viewers, according to the FCC.

But many people with analog TVs who had installed digital converter boxes felt like they were getting the short end of the stick after June 2009. (And many were.)

Congress assumed that over-the-air TV service would get better with digital, not disappear, when it ordered broadcasters to shut down analog signals. Instead, millions of Americans who relied on free, over-the-air TV lost one or more channels after the digital switch in June 2009.

We would advise waiting on purchasing any expensive new antennas or other equipment until you see how things are for 2010. Call your local TV stations and ask them what their plans are for improving their transmission. To take but one example→ WIS, channel 10, is an NBC-affiliated TV station located in Columbia, SC. The station operated for about a week after 12 June 2009 using a temporary, reduced-power transmitter. During that time, WIS converted its old analog transmitter to a full-power digital transmitter. The conversion was completed about 19 June 2009. Their new, full-power transmitter on channel 10 expanded the WIS coverage area and improved reception for many viewers. Others still required higher, more powerful antennas.

Nota  Bene→ Wilmington, NC, 2009→ According to Dan Ullmer, Chief Engineer at WECT, the NBC affiliate in Wilmington, NC→ “Digital technology is far more complex than analog television, so we see more technical issues. These include lip sync (sound and picture not working together), pixelation (bars of tiny colored blocks), audio dropouts, and black screens. The source of the problem is not always easy to track down and can come from the program originator, the network, satellite distribution, servers, station processing equipment, digital microwave links to the transmitter, transmitter issues, cable operator equipment, satellite operator equipment, or issues with the viewers equipment and/or antenna.”

What About UK?→ The UK plans to end all analog over-the-air TV by 2012, replacing it with digital. The switchover began in 2008. The Border area has already started to make the switchover and planned to be fully digital by 2009. Unlike the US, the UK is NOT switching virtually everyone on the same date. (Makes some sense, ey?) Here are the specific dates→

Border started in November 2008 and finished in 2009.
West Country started in April 2009 and finished in September 2009.
Granada switched in 2009.
Wales started in August 2009 and finishes in 2010.
STV North switches in 2010.
STV Central switches between 2010 and 2011.
West switches between 2010 and 2011.
Central, Yorkshire and Anglia switch in 2011.
Meridian switches between 2011 and 2012.
London switches in 2012.
Tyne Tees and Ulster switch in 2012.
Channel Islands switch by 2012.

US vs UK→ A Contrast→ As each section of UK converts to digital, eligible folks (e.g., the elderly) contribute $40 (USD); in return, a government technician comes to your home. No need for advertising, playing with converter boxes and/or expensive antennas. (Ok, true, folks in the UK (generally) pay an annual license fee to operate their TVs.) This help is voluntary; if you want it, it's there, but it's not forced on you.

What Do Eligible (Over 7 Million) British Citizens Get?→

• The UK will provide eligible folks with easy-to-use equipment that suits their needs.

• The UK will help with installing equipment in folks' homes.

• The UK will fit a new dish or antenna (where they can) if it is needed to make the new equipment work.

• The UK will give folks an easy-to-understand demonstration of how everything works.

• And the UK will assure that there will be someone folks can call for help while they're getting used to things.

Canadian satellite subscribers and digital cable subscribers (those who use a set-top digital cable box) will not be affected, but analog cable subscribers (who plug the cable from the wall outlet directly into their TV set) will eventually need to get a digital cable box. (After 2013, or when 85 per cent of its customers are digital, Canadian cable operators will no longer be obliged to carry an analog signal.) So your Canadian cable bill might well go up at that time. Of course, Canadians can always buy a new TV set with an ATSC digital tuner and continue receiving American (and Canadian) over-the-air digital broadcasts.

♣Our First Exposure To Digital♣

"The Future Is Endowed With Essential Unpredictability, And This Is The Only Prediction We Can Make."   -Paul Valery

How We Got Here→ High Definition broadcasting was introduced in Japan using MUSE, an analog system with compression. In the US in the 1980's, both CBS and RCA developed analog HDTV systems that were compatible with NTSC analog sets.

But the TV industry made the decision to leapfrog over analog and develop digital television. One reason was to stimulate television manufacturing in the US. A Grand Alliance was formed within the television industry to agree upon a set of standards and to promote HDTV sets. The TV industry then lobbied Congress and the FCC for legislation and rules that would foster HDTV, supposedly for the benefit of broadcasters and set manufacturers.

NTSC vs ATSC→ While roaming the highways and byways of High Definition TV, you'll often see these abbreviations→ NTSC and ATSC; no big deal, really. In a nutshell, each of these two abbreviations stands for a committee and for the standards developed by that committee→

1. NTSC is the set of standards for analog TV. The black and white NTSC Standards were based on a format finalized in March 1941 by the National Television System Committee... NTSC. These standards were approved by the FCC in May 1941. The committee met again in the early 1950's to develop an extended set of standards for compatible analog color TV, approved by the FCC in 1953.) The NTSC system was the one withdrawn from broadcast use by full-power TV stations in June 2009.

2. ATSC is the set of standards for digital TV over-the-air broadcasts. (Impress your dingo→ ATSC stands for "Advanced Television Systems Committee".) The ATSC digital standards replaced the NTSC analog standards. The ATSC Committee was created in 1983 by Congress. A lot of the ATSC committee became interwoven with "The Grand Alliance", a pompous name for a consortium of companies and research labs that developed the ATSC standards (and numerous related patents). ATSC is the only over-the-air TV system in the US for full-power broadcasts after June 2009.

   The primary ATSC standard in use by US digital TV stations is the "A/53" television standard, but additonal ATSC standards continue to grow in an attempt to fill the holes in the digital system.
   
   

   ATSC Digital TV's Myriad Of Standards And Patents

Digital TV (DTV)→ A system for broadcasting and receiving picture and sound and data using digital signals (0's and 1's), as opposed to analog signals (continuously varying waveform signals) used by traditional TV since the 1940's. With digital TV, the picture and sound are compressed early in the transmission process. Upon reception, the signal must be de-compressed by a new and different type of TV set, or by a standard NTSC analog receiver with a converter box costing $40-$90.

Introduced in the late 1990's, complete with bugs, and then with patches attempting to fix the bugs, digital TV transmissions allow broadcasters, if they wish, to provide TV with higher resolution pictures. The optional higher resolution pictures are called High Definition or HD; broadcast HD pictures are made up of sixty 720- or thirty 1080-horizontal lines per second, compared to thirty 480-horizontal lines per second on the screen for an old NTSC analog picture; HD also contains more picture data on each of its 720 or 1080 lines.

Sound→ All ATSC digital is broadcast with Dolby sound, with from 1-6 sound channels. Digital TV also can transmit multiple programs simultaneously on a single physical (6 MHz) TV channel, called multicasting.

Some have likened multicasting to distributing several additional TV broadcast licenses to every full-power TV station, at no charge to the broadcaster. In Other Words→ Although the digital transition has meant cash outlays for stations (digital equipment, tower modifications, etc), it opens up major additional revenue "opportunities" through multicasting. (Every sub-channel that's multicast is another stream of pay-fer commercials.)

Since June 2009, all full-power TV stations have ceased broadcasting in analog. Digital TV alone rules the full-power TV spectrum, channels 2-51 (the so-called "core" TV spectrum). Former TV channels 52-67 will go to the phone companies paying the most $$$ (these frequencies are gold); and former TV channels 63, 64, 68, and 69 will be used (someday, presumably) by emergency service providers such as fire fighters.)

One Comment→ All newer digital TVs (HD and non-HD) also include NTSC analog tuners, so that they can receive both old analog and new digital over-the-air broadcasts. This is not the least expensive solution, but it definitely covers whatever TV stations and the FCC can throw at you. If you're flush and employed and confident about 2010, consider a digital TV (a TV with an ATSC tuner); it does not have to be an HDTV set. Just digital.

♣TERMINOLOGY→ HD And HDTV And Dolby And Such♣

"Not everything that can be counted counts, and not everything that counts can be counted."   -Albert Einstein

A Second Fly Over→ What's HD (A Little Deeper)?→ HD over-the-air is 6 of the 18 possible kinds of digital television pictures specified in the ATSC standards; these top six have a clearer and sharper resolution and a better-looking picture than the other 12 digital broadcast formats. The digital sound is the same quality in all 18 picture resolutions; but only the top six picture resolutions are prime dry-aged NY strip high definition.

HD has a higher resolution video signal; an HD camera at the TV studio generates 1.5 billion bits per second. This video is compressed before being mixed with Dolby sound (also compressed); compressed video and sound are mixed with miscellaneous data and broadcast over-the-air.

High Definition is broadcast using the same transmitter as Standard Definition. The FCC requires that all full-power TV stations broadcast in ATSC digital. But no TV station is required to broadcast a picture of HD quality; it is up to the station whether or not to broadcast in High Definition; the FCC only requires that the station broadcast over-the-air in digital, following the ATSC digital standards.

HDTV→ HDTV is a TV set (receiver) that can display HD broadcasts or the Blu-ray Disc DVDs in all their HD splendor and detail. An HDTV set also can display Standard Definition (SD). For use with cable, an HDTV set needs a set-top digital box rented from the cable provider (or the HDTV set needs a built-in QAM tuner) in order to receive HD over digital cable. An HDTV set has one to two million tiny square picture elements called "pixels" on its screen.

Example→ To get the best TV picture, you'll need an HDTV set and an HD signal as input. An HD signal may come over-the-air (OTA) to your antenna, or it may come from digital cable, satellite, or FiOS; or it may come from a Blu-ray Disc DVD player.

Clear QAM→ Non-encoded cable programs sent without charge are called clear QAM, as in... "Local affiliates of the major networks are often broadcast via clear QAM (free), and often in High Definition, by cable providers".

What's Dolby?→ Whether the video is high-definition (HD) or standard-definition (SD), all US digital television uses Dolby audio. The Dolby format of any TV broadcast varies from one audio channel (mono) to six-channel Surround Sound (5.1). If you have Blue-ray Disc, you may listen to 7.1 (eight-channel) Dolby. Dolby has sound of CD quality. Dolby 5.1 can be connected to six speakers located around the room; or you can always rely on the speakers in the HDTV set. After compression, Dolby 5.1 is carried at the TV station and over-the-air on a bit stream at 384,000 bits/sec.

In Other Words→ All High Definition broadcasts in the US must come from Digital TV transmitters. But all digital TV broadcasts are not necessarily High Definition; some are HD, and some are not HD.

If the HD broadcast is local, like the 6 PM news, the local studio must incorporate expensive new HD cameras and such. If the HD broadcast comes from the network, like a football game on CBS, it's up to each local station whether or not to broadcast over the air in HD. Most local stations that carry network programming (called "Network Affiliates") broadcast their network's HD programming in HD. It makes sense.

♣Pixels And Lines♣

"It Is Undesirable To Believe A Proposition When There Is No Ground Whatever For Supposing It True."   -Bertrand Russell

Now at last it's time to start digging into some High Definition technology and engineering... simple stuff like Pixels and Rows and Lines.

First, Rows and Lines are the same thing. (See what simple stuff this is.)

• Now... What Is a Pixel? A Pixel is short for "A Picture Element". Like the images on a computer monitor, the images on an HDTV set are composed of many tiny squares. If you look closely at an HDTV set, especially a 50" set or larger, you'll see the tiny square pixels. The brightness and color of each pixel on the screen varies rapidly, changing 30, 60, 120, or even 240 times per second, presenting .the illusion of a motion picture.

• What Is a Frame? A Frame is one complete picture. The number of complete pictures captured by a TV camera and displayed on a TV receiver's screen in one second is called the Frame Rate. On today's (2010) HDTV sets, the frame rate is typically 60 or 120 or even 240 per second.

• What Is a Field? Field is a term that crops up in Interlaced imaging in a studio TV camera; or when "interlaced" video reaches a TV set. An interlaced picture is taken by a TV camera as→ First, all the odd-numbered lines in the image that's focused in the camera are scanned; that's the first field. Assuming HD format 1080i30, that takes 1/60^th of a second. Then, all the even-numbered lines focused in the camera are scanned; that's the second field, and it also takes 1/60^th of a second to scan.

  Each of these "half-frames" (odd, then even) is called a Field. When a 1080i30 HD broadcast reaches your HDTV set, the first Field, the one with the odd-numbered lines, is not displayed on the screen; instead it's stored in the HDTV set's memory. Then the Field with the even-numbered lines arrives; this second field is then "woven" together with the first field that we stored in the set's memory; and voilà, after our weaving we now have a complete Frame, a complete picture, which is displayed progressively from top to bottom (line 1, 2, 3,... 1079, 1080) on your flat-screen HDTV set. Simple stuff this definitely be.

  1080i30→ Field #1 + Field #2 = A Complete Frame

• What Is Flicker?→ Flicker is not something that happens in the world of broadcasting or in our TV set. Flicker is a biological thing; it's not an engineering thing. Flicker exists in the visual system of the viewer. If the frame rate (complete pictures per second) is too slow, some folks will see an annoying flicker. But the TV set itself is not flickering. Again, flicker is not a bug in the broadcast, nor in the HDTV set.

On the screen of an HDTV set, there will be anywhere from just under one million square pixels (on "720p" sets) to just over two million square pixels (on "1080p" sets). Older analog TV sets have rectangular pixels, not square ones, like HDTV sets do.

• 1080p sets have 1920 pixels per row × 1080 rows.

1080p has the potential for displaying more of the detail in High Definition signals, compared to 720p or 1080i30. This difference in sharp detail is most noticeable→

1. If your set's screen is 50" diagonal or bigger.

2. If you watch the TV from within a few feet of the screen.

3. If you're watching an HD signal.

Otherwise, with quality sets, the difference in detail (or sharpness) between 720p and 1080p often is subtle. Take a "test-drive" before purchasing an HDTV set; 720p is less expensive than 1080p, and no station can broadcast over-the-air in 1080p. Yes, Blue-ray Discs can display 1080p.

Each pixel on an HDTV receiver's screen is made up of three "sub-pixels", one red, one green, and one blue. By combining these three colors in various proportions, a single pixel appears to take on most of the colors that we can see.

A  BRIEF  JOURNEY  INTO  VISION→

Three pigments is by no means a maximum for all lifeforms. Up to 10% of women and lots of birds are tetrachromats, with four pigments; in 2-3% of women, the fourth cone pigment lies in the orange range, allowing these women to see about 100 million colors. (Genetics dictates that all human tetrachromats must be females.) Many birds can see into the ultraviolet region of the spectrum. And there is a species of shrimp that has 20 different pigments... and thus requires 20 primary colors.

OK... Back to HD→ How many rows of square pixels are there on an HDTV set's screen? "720p" tells us there are 720 rows; "1080p" tells us there are 1080 rows. The "p" in 720p and 1080p tells us that each row of pixels (row 1, row 2, etc.) is energized (lit up) Progressively. In other words, progressive means that the display of lines progresses from the top to the bottom of the screen in sequential order in a single pass.

HD programs are broadcast over-the-air in one of two ATSC digital formats→ Either 720p or 1080i. 720p has fewer rows of pixels than does 1080i, but it has the advantages of "p"... the advantage of progressive scanning→ 720p is better able to display rapid motion; e.g., a football game. Why? Because in a "p" (progressive) system, the Complete picture is captured by the TV camera at once. Every line is scanned sequentially in the TV camera, from top to bottom. Line 1, line 2, line 3... until the bottom of the picture is reached. And with 720p, 60 complete frames are captured by the camera every second.

The Bottom Line→ Broadcasts at 1080i must blur the picture slightly to prevent "twitter"; twitter is where something appears on a horizontal line in only one of the two fields making up an "i" frame. 1080i transmissions Do have slightly better spatial resolution than 720p broadcasts do. But 720p transmissions have better temporal (motion) resolution.

1080i produces a sharper picture when the image is stationary or moving slowly, because it has more scanning lines. 720p, on the other hand, excels at reproducing rapidly moving objects without blurring, since its full-frame progressive scanning never chops a frame in two. And 1080p is the best of both worlds→ both good spatial and good temporal resolution. But ATSC digital cannot broadcast 1080p over the air in the 6 MHz that the FCC has allowed for digital channels. (6 Mhz channels are the same for digital channels as they were for analog channels.) But you can get 1080p60 from Blu-ray Discs, which is higher resolution than any HD that is broadcast over the air.

♣Deeper And Deeper Into Video Formats We Go♣

“They say that 90% of TV is junk. But 90% of everything is junk.”   —Gene Roddenberry

Now let's bore deeper into "Video Formats" like 480i and 720p and 1080i and 1080p→

Resolution Confusion→ Few items are used so erroneously by so many folks as video formats. More people are confused by video formats than by HDL and LDL cholesterol scores. Over the years, your dog Wolf has learned that when something appears to be as confusing as video formats, they are very simple in actuality... so don't let the sales folks in the Big Box Stores try to confuse you on the subject of video formats.

First→ Video formats like 720p and 1080i and 1080p are used in two different ways when one is chatting about HD and non-HD broadcasts→

1. Video Formats are used to describe the resolution at which images are picked up by the TV camera at the TV studio and broadcast from the TV antenna atop the TV tower.

2. Video Formats also are used to describe the resolution at which TV receivers display their pictures.

These two uses can be quite different. And Format Codes can apply to both analog and digital broadcasts, and to analog and digital TV sets.

We specify the Format Code with which a TV Picture is being transmitted, and the way that a TV Picture is being displayed on our TV receivers, by answering three questions→

1. How many rows of pixels are there?

2. Are the rows of pixels interlaced (one of two fields at a time), or are they progressive (the whole frame at once)?

3. How many complete pictures (frames) are broadcast or displayed each second?

Time For an Example→ Let's see what the Format Code is for our good old NTSC analog broadcast system. This is a nice example because, unlike the new digital formats, the old analog TV signal was transmitted and displayed in just one format code; and the picture was always transmitted with the same format that it was displayed on the old TV set's screen.

First→ We know that 480 lines or rows of video are broadcast for each analog picture (or "frame"); so our format code would begin with the number 480 (the resolution).

Second→ To answer the question→ Is each frame interlaced or progressive, we simply ask if all 480 lines are transmitted at one time from the TV camera, top to bottom... 1, 2, 3, ..., 479, 480. Because that is one way to do it, send all 480 lines progressively from top to bottom... send them all at once. If that's how our 480 lines are sent, progressively, then our second value is a "p". "p" stands for "progressive"... and so we'd now have 480p for our Format Code.

But there also is another way that we can send the 480 lines that eventually will make up the picture on our analog TV set; we also can send them "interlaced". Interlaced means that half of the lines (240 lines), every other line, the odd lines, are sent from the TV camera first... 1, 3, 5, 7, ..., 477, 479. And if we don't have a flat-screen TV, those 240 lines are displayed immediately, no storing them. And then, 1/60th of a second later in this example, the even lines generated from scanning the image in the TV camera are transmitted... 2, 4, 6, 8, ..., 478, 480 and displayed immediately. If this is the case, we'd now have 480i for our format code.

It turns out that, to reduce flicker in the old NTSC TV sets when NTSC standards were being developed, the old analog NTSC system sent those 480 lines interlaced... first the odd scanning lines trace across the image in the analog TV cameras; then "black" is transmitted for an instant; and then the even scanning lines from the TV camera trace across the image in an "interlaced" fashion... so we put an "i" after the 480... meaning that frames (complete pictures) are broadcast with 480 lines, but first the odd lines are sent, and then the even lines are sent. We specify 480 lines with "interlaced" transmission and display as 480i.

And Third→ We want to specify how many frames (complete pictures) are sent per second. It's important to understand that "480i" also can be a "family" of resolutions. Each "member" of the family refreshes the whole picture on the screen (the frame) a different number of times per second.

In the US, under the old NTSC analog 480i standard, 30 complete pictures were sent from the TV station to our TV sets each second; i.e., 30 frames per second. And so we now can specify our complete Format Code as 480i30. And if we are sure that there is no room for ambiguity, and that everyone will know that 30 complete frames are being sent and displayed each second, we can use a "short-cut" and simply specify the Format Code in our example as 480i.

Let's pause here for just a moment to go over this, because it's very important. A Format Code written as 480i can mean two different things→

1. It can mean that we're certain that everyone knows that the frame rate associated with 480i is 30, and so we're writing 480i as a "short-cut" for 480i30.

2. Or 480i can stand for a "family" of Format Codes... 480i30, 480i29.97, 480i24... in other words, lots of Format Codes that are the same, except for the frame rate.

Fast forward now to the present and to HD broadcasts. There are two formats used for HD broadcasts in the US under the new digital ATSC standards. The first is 720p60. (720 lines (rows) are in each complete picture; p = progressive (not interlaced), the whole picture is sent from top to bottom at once; and 60 complete frames (pictures) are sent each second to our HDTV sets.) Because this Format Code 720p60 is so common today, it is often simply written as 720p, and we assume that 60 complete frames will be sent from the TV transmitter to our HDTV set every second.

The second format that is used for HD broadcasts in the US is 1080i30. 1080 rows in each picture; i = interlaced, the 1080 lines are sent as 540 odd lines and then 540 even lines; and 30 complete frames are broadcast every second. Because this resolution is also so common today, it is often written simply as 1080i, and we assume that 30 complete pictures will be sent from the TV transmitter to our HDTV sets every second. ¿Está claro? Great.

FORMAT-CODES

Currently (2010), there are six picture formats that the ATSC Standards have labeled HD; but only two of these six video formats are used in the US for HD over-the-air (or cable or FiOS or satellite) broadcasts→ 1080i30 and 720p60.

All US HDTV sets will work fine receiving either of the two HD formats, 720p or 1080i. ABC sends out its HD programming at 720p (720p60), CBS sends out its broadcasts at 1080i (1080i30), no problems. They both look great on good HDTV sets.

The most common HDTV sets today (1080p, called "full HD") can display 1080p60. 1080p60 is not one of the standard ATSC 18 video formats, and it is not presently (2010) broadcast over-the-air nor via cable nor satellite nor FiOS.

(On 1 August 2008, DISH Network Satellite TV made the movie "I Am Legend" available "On Demand" at 1080p60. This upgrade in resolution wasn't available for all subscribers, but it was a no-charge addition for anyone who owned an HD DVR capable of playing MPEG-4 video.)

Store Vs Home→ Self-styled "purists" warn that sets never look their best in a store; the lighting is bad, and store employees rarely set them up and adjust them properly. But this is exactly why you should look at them in a store. Most people use their TV’s right out of the box, with all the default settings. So comparing the store displays is a great way to see what is likely the worst for each set. If the set looks good in the store, it should look even better in your home.

Store Cable Vs Home Antenna→ In the store, the HDTVs are connected to cable. So how can you tell what HD will look like at home, if you'll be using an antenna?

You can't, not in the store. But any HDTV set that has a high rating for its picture and a built in ATSC digital tuner can give you an excellent picture at home through an antenna, better than cable. Your HD picture at home will depend on the terrain from your home to the TV broadcasting tower and your distance from the broadcasting antenna. In other words, the signal strength of the station at your antenna. If the signal strength is strong, you will get an excellent HD picture. Better than cable.

A Critical Point→ While HD is only transmitted over-the-air at 720p and 1080i, your HDTV set always will convert the transmission format to the set's "native" format. Thus, if you are watching HD at 720p on ESPN HD or ABC using a 1080p HDTV set, your HDTV set will "upconvert" 720p→1080p. If this upconvert is done well (which is one reason that some sets cost more than others), the picture that is displayed on your 1080p set will be excellent.

Also 1080p60 HD video signals are available from the newer generation Blu-ray Disc DVD players. But broadcasting 1080p60 over-the-air would require a channel bandwidth greater than the 6 MHz available in the US for any analog or digital channel. And 8 MHz TV channels (as much of Europe uses) would reduce our capacity in the US for texting and twittering. And so we in the US didn't convert over to the super-beautiful, more-lines-per-image European system.

Note→ Both analog and digital transmissions in the US are broadcast using 6 MHz (6,000,000 cycles per second) radio frequency channels... e.g., Channel 6 is transmitted using 82 MHz to 88 MHz, whether NTSC analog (color) or ATSC digital (color)... everything is broadcast in color today, even old black and white movies.

The  Bottom  Line→ Should you buy a 720p or a 1080p→ We'd consider a high quality 720p; it's cheaper, and few folks can see the difference in resolution between the two formats on high-quality sets, until they reach 50". If you have a large screen, or a large projection TV, or Blu-ray Discs; or if you actually believe you can tell the difference in resolution between 720p and 1080p, then spend the extra money and purchase the 1080p. Also, if you are planning to display pictures from your PC on your HDTV, get the 1080p.

Nota Bene→ 1080p, when used to describe HDTV sets, is a short-cut terminology for 1080p60, often called "Full HD". However, in 2010, 1080p is increasingly used to mean 1080p120 (and occasionally 1080p240).

THE  BOTTOM  LINE

♣There Is ATSC DIGITAL And There Is CABLE DIGITAL♣

"All Things, Good And Bad, Must End. Nothing Is Forever."   -DAwn McGatney

There are two types of cable TV for our purposes, the original analog cable, and the newer digital cable→

• With analog cable, a single program can be placed into each 6 Mhz cable slot. Often, if the TV set is "cable ready", the cable wire plugs directly into the set; usually, there is a switch that allows alternating between over-the-air tuning and tuning to cable channels. Receiving analog cable is simple... just a matter of tuning to the proper frequencies that correspond to the analog cable channels. There is no need for any special demodulation.

  With analog cable, if you don't have a "cable ready" TV set, you can get a tuner box from your cable company; generally, this analog cable box converts the cable frequencies to channels 3, or 4 on your TV set (whichever channel is unused in your area). You simply connect the cable box to your TV set's antenna screws (or coaxial cable input), and set the box to the analog cable channel that you want to receive.

• Digital cable is very different. Several programs fit into one 6 MHz cable slot. And there is compression.

  Digital cable is QAM modulated ("Quadrature Amplitude Modulated") at the head end (at the cable company), either 256-QAM or 64-QAM (more on QAM coming up). QAM is the modulation system by which digital cable channels are encoded and transmitted over fiber-optic and coaxial metal cable.

  QAM tuners are analagous to ATSC tuners which receive over-the-air digital broadcasts; and many new "digital cable ready" HDTV sets have tuners for both ATSC (Over-the-air) and QAM (digital cable). A TV set with a QAM tuner also de-modulates (and decodes) digital cable without the need for renting a set-top box from the cable company.

  QAM cable can carry nearly twice the number of bits/second as ATSCdigital over-the-air modulation; but since QAM requires a cleaner signal path (that is, a path with fewer potential errors), QAM it is a good match for digital cable.

HD can be sent over only digital cable. If you are one of these 27 million subscribers (2008) who have totally analog cable, you will not receive any HD from your analog cable. None. Never. Ever.

Cable programming begins at what's called the HEAD END. The HEAD END is the location where the cable system accumulates ITS programming... from satellites, through microwave towers, from local broadcasts over-the-air, from its own studios, from local broadcasts over fiber-optic cable, from Neptune, you name it. There are about 8,000 HEAD ENDs in the US.

The HEAD END first assigns a "virtual" channel number to each program that it makes available to the community. The HEAD END then "shoves" each program onto fiber optic (or metal coaxial) "trunk cables", which carry the programming to each neighborhood that is serviced by this cable provider.

Today, the majority of digital cable programming is ENCRYPTED at the HEAD END, especially "premium" channels like HBO. (You have to pay to play. And to play, you have to have the proper "key" to decode the encryption.)

So there are three kinds of cable→

1. 100% Analog Cable.

2. A Hybrid Analog/Digital Mixed Cable System.

3. 100% Digital Cable.

      § § § § § § § § § § § §

1. Analog Cable→ Analog cable systems can carry programs only in the 480i30 format, the lowest video quality format.

   Another interesting thing about analog cable is that one program (like the TV Guide Channel) totally fills one physical 6 MHz RF slot; there is no compression, no stuffing of 5 programs into one physical RF channel. Analog cable channels tend to have channel numbers below 100. (There just aren't that many physical cable channels.) And what you see is what you get... you tune to cable channel 66, you GET physical RF cable channel 66 (located on the RF spectrum at 475.25 - 481.25 MHz).

   Finally, analog cable is susceptible to interference. So in summary, Analog Cable→ No HD, one program per 6 MHz physical cable channel, and possible interference.
   

   ---

   
   

2. Hybrid Analog/Digital Cable→
   
   

   Typical Allocation Of A 750 MHz Hybrid Cable System

   
   

   ---

   
   

3. Digital Cable→. Unlike analog cable, Digital Cable CAN carry HD (720p and 1080i) programs to your HDTV set; and it also can carry programs in Standard Definition (SD = 480i) and Enhanced Definition (ED = 480p).

   Since Digital Cable is digital, and digital things can be represented by bit strings, we would expect to be able to measure the capacity of digital cable in bits per second. Digital cable, using 256-QAM modulation in a 6 MHz physical RF channel, can carry up to 38.4 million bits per second. This, however, represents potent compression by the cable provider; this is nearly double the maximum bit rate that an over-the-air TV station can broadcast (19.39 Mb/s).

   LOSSY→ The more digital cable compresses a cable program, the worse it will look. TV compression is definitely a lossy process. (A digital process is called "lossy" if, when it comes time to reverse it, you don't end up with all the information that you started with originally; this "loss" of information can be subtle, or it can be a major degradation that makes watching a TV program nearly impossible and very unpleasant.)

   Digital cable provides a higher quality picture than analog cable. There is a dramatic improvement in color resolution.

   Analog NTSC TV shows only about 120 of its 480 lines in color; digital cable shows about 270 of its lines with color. But again→ Digital compression "softens" the quality of a picture, particularly on digital cable channels that are severely compressed.

   So in summary→ Digital cable can transmit HD, it can carry a high bit rate, it can be tightly compressed, too much compression looks bad, and there is more color detail with digital cable.

A CRITICAL POINT→ If the FCC has counted correctly, and there are 20-40 million households that presently have 100% analog cable, and these households will not be able to receive programs in High Definition.

And if your cable provider offers Both analog and digital (hybrid), but you don't subscribe to digital cable, you will not receive HD programs which the cable company has encoded; if your HDTV set has a built-in QAM Tuner, as most newer sets do, you Will be able to receive Some HD via digital cable at no cost, usually the HD programming which is locally broadcast unencoded... like Letterman on CBS, carried by your local CBS affiliate station in HD.

In Other Words→ All HDTV sets have ATSC tuners (so they can receive digital signals over-the-air); some HDTV sets also have QAM tuners. Back when all TVs were analog, the "Cable Ready" TVs didn't need QAM tuners; that's why you just plugged the cable into the back of the TV and tuned in the 6 MHz analog cable channel of interest... no QAM receivers (tuners) or set-top boxes were required. Same for the current HDTV sets with QAM tuners... just plug your cable into the back of the set and receive un-encoded digital cable.

Now... TV manufacturers rarely discuss it or print anything in their instruction manuals about it... but cable providers relay to any customer whose set has a QAM tuner all digital (including HD) transmissions that are not encrypted. So if your QAM tuner is part of an HDTV set, you'll get some HD free. No set-top boxes are required. How can this miracle occur?

Answer→ If cable providers broadcast local programming, then they also must carry local broadcasts of High Definition programming, unencrypted ("in the clear"), programming that does not require their customers to use rented equipment, per FCC Sec. 76.630 and CFR Title 47, §76.901(a). Cable providers comply with these laws by broadcasting HD over QAM (cable digital) channels. The law does not require cable providers to advertise the availability of HD, and cable representatives often will wrongly insist that a converter box is needed to view ANY HD channels.

The cable channel numbers seen on your digital cable set-top box and in your program guide are VIRTUAL numbers. In the old days of 2-digit analog cable channels, the channel you were viewing actually corresponded to an "RF channel". When you went to channel 75, your cable-ready TV or set-top box was actually tuning to the radio frequency for cable channel number 75 (528 - 534 MHz), and displaying the single "service" that resided there (if any).

But with the advent of digital cable, and hundreds of three digit channels, that all went out the window. Now the cable company can stuff up to a dozen "services" into a single cable RF channel. These services basically pile up in that channel, and now your set-top box or your new HDTV set pulls them apart into individual programs.

The virtual numbering plan in the digital cable system is proprietary, and requires the set-top box to decode. So they may have one or more services at RF channel 11. But HDTVs with QAM tuners don't know about virtual service numbers, they just knows that they found it at RF channel 11. So the TV set gives it service number "11-0", and if it finds any others piled up there at RF channel 11, it'll call those 11-1, 11-2, etc. After all, HDTVs are expecting to find digital multicasting within a 6 MHz RF channel anyway; they are trained to sniff around, looking for multiple programs (services) in one physical RF channel.

Many cable companies shift their line-ups frequently, so that QAM-sters must rescan to find the new arrangement. But it's not uncommon to find 9 or 10 free HD channels, though you may find one on channel 81-11, for example (the 12th "service" on major cable channel 81). And you are not required to pay extra. But if you want a non-broadcast HD channel (like Discovery HD) which is almost always encoded, then you will need a set-top box (or at least a "CableCARD").

In Other Words→ The QAM tuner integrated into your HDTV set allows the free reception of unscrambled digital programming sent out by cable providers, usually local broadcast stations; however, most digital channels are scrambled, because the providers consider them to be extra-cost options and not part of the "basic cable" package.

The CableCARD→ Since July 2007, the FCC has required cable providers to separate security hardware (and software) in their set-top boxes from the hardware (and software) that does the receiving and tuning and recording. All of the security issues (the part that makes you pay $$$ each month, or you get nada) are now on a card that slips into the box (or into the back of your TV)... the "CableCARD".

The FCC is working on the premise that you can buy the box (or buy a TV that incorporates the box), take it with you if you move, and just get a new card from your new cable company; the card is about the size of a credit card (but a bit more chunky). Comcast in 2009 did not charge for the first CableCARD in most markets, but they required that their technicians install the card in your TV.

Back to Comcast→ Since the signals that we get from Comcast are sent to our homes in "cable digital", they must first go through a "set-top box" or a DVR, which is rented from Comcast, for decoding any scrambled cable digital channels.

Anyway, when we select some cable channel on the Comcast set-top box, the box converts "CABLE DIGITAL" into either→

1. Analog, for older NTSC TV sets (usually received on channel 3, or connected directly into a "cable ready" analog set), or...

2. Into ATSC digital for the newer ATSC digital sets.

SO... If you have an OLD ANALOG TV, the set-top box will convert "CABLE DIGITAL" to ANALOG, and you can then watch most cable channels on your analog TV. You can even watch some programs on your analog TV set that are broadcast in ATSC digital (but not HD, for that you have to pay extra).

♣Satellite♣

"When A Debater's Point Is Not Impressive, He Brings Forth Many Arguments."   -The Talmud

Clarification→ Digital Cable and Digital Satellite do not mean that all programs are received in High Definition. Some digital programs may be in HD, other digital programs may not.

Satellite TV, more correctly called Direct Broadcast Satellite or DBS, since the 1990's has been digital (and compressed), allowing more channels per satellite. Digital satellite provides a better picture and better sound (less snow and crackle). But like Cable Digital, Satellite Digital also is different from ATSC (over-the-air) digital.

As of 2009, DirecTV had 18,008,000 customers, and DISH had 13,800,000, for a total of about 32 million subscribers.

Because High Definition broadcasts from satellites to dishes require more "bandwidth" on the satellite broadcast spectrum than non-HD, satellite providers are compressing HD transmissions more, now using a newer technology than the MPEG-2 compression specified by ATSC for over-the-air broadcasts (and used by many cable providers). The newer technology is MPEG-4. MPEG-4 must be decoded using new set-top boxes.

In addition, to handle HD, DirecTV is employing a newer transmission protocol called DVB-S2 from the SPACEWAY-1 and SPACEWAY-2 satellites, allowing DirecTV to squeeze more HD programming into its satellite signals than previously feasible using the older "DSS" transmission protocol.

DirecTV→ DirecTV Satellite transmits HD video in a format that is not one of the 18 ATSC video standards. It is a combination of ATSC 720p60 and 1080i30 formats. The ATSC standard specifies 1920 pixels per line for 1080i30, but DirecTV reduces this number by 1/3 and broadcasts 1280 pixels per line, similar to 720p60. (Satellite can broadcast in any format it likes; ATSC applies only to over-the-air broadcasts.

The practice of reducing the original resolution of an HD signal between the broadcast facility and the home is called down-sampling.

Moral→ If you're an HD perfectionist, you may want to view local HD broadcasts using a rooftop antenna to avoid the possibility of downsampling and/or additional compression after leaving the broadcast station.

♣FiOS♣

"Here We Are In This Wholly Fantastic Universe With Scarcely A Clue
As To Whether Our Existense Has Any Real Significance."   -E.F.Schumacher

FiOS stands for Fiber-Optic Service. (A little thought will reveal why Verizon didn't name it "FOS".) Verizon also has pointed out that in Irish, "fios" means "knowledge". Unlike cable, FiOS is a "Fiber To The Premises" (FTTP) service. Said Verizon, “Verizon FiOS is the latest in fiber-optic technology. It delivers laser-generated pulses of light, riding on hair-thin strands of glass fibers, all the way to your front door. When FiOS meets your HDTV set, you get TV at blazing-fast speeds.”

In Other Words→ With FiOS, stuff is not just carried to (and from) your neighborhood via fiber, as with ordinary cable. In the FiOS world, the fiber actually reaches your home; this is the key to the speed and bandwidth of FiOS. Verizon uses FiOS to bring 100% digital television (with tons of HD) directly into homes.

Verizon has predicted that, after it finishes its tweaking, you'll have 100 million bits per second flowing to (and from) your home on fiber. That's three times the THEORETICAL limit of ordinary cable. It's fast enough to let you download entire movies in a few seconds. FiOS can run much faster than cable or satellite. FiOS is fast enough to bring just about anything into (or out of) your home, stuff that you'll want for the Network of The Future. Every premium channel is available on FiOS in HD without "lossy" compression.

When you subscribe to FiOS TV, you abandon cable. FiOS HD is not compressed; no "lossy" compression. It's HD television picture is superior to both cable and satellite. In early 2008, Verizon was giving away a small HDTV set with a FiOS subscription; they told us that they wanted folks to see just what FiOS HD looked like.

Some FiOS Problems→

• Verizon may have to dig up your lawn to get the fiber to your house, but it's like plastic surgery; the scar usually heals quickly. However, the installation inside your home isn't trivial; it can take several hours.

• FiOS presently is available only in limited areas. Some locales will never get FiOS.

FiOS allows Verizon to deliver hundreds of channels of TV (including lots of HD), just like Comcast has been doing, but without the need for compression that degrades the picture. FiOS IS FAR FASTER than cable; "ON DEMAND" takes perhaps three seconds to kick in with FiOS; cable takes quite a bit longer than that. And the FiOS TV picture is clearer and brighter and more consistent than cable, according to folks who have tried FiOS and cable.

Verizon also has introduced a digital video recorder (DVR) that can show its stuff on ALL The TVs in your home... not just the TV that the DVR box is sitting on. You even can transfer pictures and music from your PC to your TV using FiOS.

The Bottom Line→ Compared to DirecTV and Dish Network (Digital Satellite), and to cable providers such as Comcast, FiOS TV provides better picture quality. FiOS should continue to provide more selection, whereas satellite providers may have maxed out their bandwidth capacity. In fact, the satellite providers (along with many cable providers) have increased compression from MPEG-2 to MPEG-4. In order to provide more channels, they have no choice but to increase compression, which results in reduced picture quality.

As of late July 2009, Verizon's FiOS TV had 2.5 million subscribers. By December 2009, FiOS offered 632 channels, 125 of these channels with uncompressed (higher quality) HD.

♣Digital TV Reception♣

"Television has done much for psychiatry by spreading information
about it, as well as contributing to the need for it."  -Alfred Hitchcock

• Your HDTV set is going to do one or two things upon receiving a digital signal. If it is receiving an interlaced broadcast from any source, 480i30 or 1080i30, it is going to have to de-interlace these into 480p30 or 1080p30 for your flat screen HDTV set. Higher quality HDTV sets do this de-interlacing better, and their pictures look better.

• Next, your HDTV set will convert the input signal's format to the set's native format. If you are watching an HD program on NBC (e.g., Leno), it is being broadcast in 1080i30. If your HDTV set is a 720p60, it must convert 1080i30 to 720p60. Higher quality HDTV sets do the conversion better. And they look better.

• The additional compression that cable and satellite apply degrade their HD pictures. Many cable head-ends get their local station feeds via a fiber cable, some via microwave, or occasionally directly over-the-air.

• Digital over-the-air is finicky. You need just the proper antenna aimed at just the right tower, a tower's that's not too far away, with just the right lead-in cable to your HDTV set in order for over-the-air to equal FiOS; otherwise, FiOS wins. You may have a little snow using your rooftop antenna with an analog set; after that station switched to digital last June, you will probably get nothing; or worse, a picture with pixelation, the sound cutting in and out.
  
  

  ---

  
  

  Idea→ Broadcast stations need to consider building analog translator stations to fill in the gaps left by digital; these would be lower-power analog stations, and thus they would not place an inordinate burden on the broadcast station.

• Digital TV was conceived over a great many years, by a great many companies and individuals, often with competing interests and patents. There is little about digital that could not be re-done better. Digital TV uses error correction on top of more error correction, patches on top of patches. Only in April 2009 did the ATSC finalize another set of compatible standards that require more hardware for the broadcast station and digital receiver (ATSC-M/H), and which allow ATSC digital to be received by handheld and mobile digital receivers. Why were these new standards not incorporated in the original ATSC standards? Ask them.

  Yes, digital looks better than analog... when things are perfect. If you are using a roof-top antenna, be prepared for a new, more expensive antenna, installed by a profesional, especially if you are not in a flat area. Mountainous is the worst for digital. Digital TV today is roughly where analog color was in 1954... recall the grass turning pink and so on? FiOS + a great HDTV set will look great; compressed cable and satellite or a roof-top antenna in a hilly area... sorry.

  Says WBNG-TV (Binghamton, NY) Chief Engineer Chris Ball, “The digital signal doesn't bend and may not reach lower lying areas as well as the old analog signal did.” He suggests installing your antenna on the roof for better results.

  If you still aren't getting a signal, Ball suggests moving the antenna to another area of your home. “Either the antenna isn't large enough, or it's not pointed in the right direction; and when folks point it, they try to do it the same way they did with analog, just move it and watch the picture come in. What you have to do with digital is move it a little bit, let it sit and pick up the signal, and go forward from there,” says Engineer Chris Ball.

• Installing An Antenna→

  • A quality outdoor antenna is required. The antenna should be at least three feet above your roof and not in the attic underneath metal roofing. Also, if your present roof-top antenna is more than 15 years old, it will need to be replaced.

  • The lead-in wire connecting the antenna to the TV or converter box should be the round “coaxial” cable, NOT the old style “twin-lead” or flat wire, which will result in signal loss and poor picture quality. Like antennas, cables periodically must be replaced every 10-15 years.

  • Connections between the antenna and lead-in wire and to the TV or box must be secure, and "impedance matching transformers" used at the antenna (and when needed at the TV or box). In other words, if the cable is "75 ohm" and the input to the box is "300 ohm", you'll need a small, inexpensive matching transformer. Ideally, leave no excess coil of lead-in wire on the roof or on the floor by the TV or box. The longer the cable, the more signal loss.

  • The antenna must be pointed at the station(s) you wish to receive. TV antennas have a “front” and “back” side, and the front must be pointed at the station(s). The smaller elements, the directors (or the open end of a “V” shaped antenna) are the front. The longer elements behind the cable connection are the reflectors. All other factors being equal, the more elements, the more powerful the antenna; but the more elements, the more directional the antenna, and the more precisely it must be aimed at the desired TV tower(s). TV antennas will work poorly, if at all, when not properly aimed.

• There is no such thing as a "digital antenna"; i.e., an antenna with special features that works especially well receiving the new ATSC digital programs. All antennas are just a hunk of metal, tuned to receive a certain frequency, or range of frequencies. Whether it is receiving an FM, an AM, a Digital TV, or a Single SideBand signal, an antenna is just a hunk of metal that will receive the frequencies it was designed for, regardless of the modulation, analog or digital. If you pay $100 for an indoor digital TV antenna, you're being ripped.

♣Thinking Inside The Box♣

"A Dream Is A Prophecy In Miniature."   -The Talmud

SOME CONVERTERS THAT FOLKS LIKE

***The Zinwell ZAT-970A converter box is a good all around box. It comes with the analog pass-through feature. The picture quality of the Zinwell ZAT-970A TV converter box is very good. It provides a picture that is a very significant upgrade over analog television. This TV converter box's picture rates at just below the top TV converter boxes. The sound is very good. It provides viewers with a very crisp sound that rivals top converter boxes.

The ZAT 970A converter box comes with the ability to search for all channels at once or set channels one by one. Finally this box comes with an electronic program guide that is better than that with most other boxes. The Zinwell ZAT-970A TV converter box provides a great set of features, also selling below $55, which makes it an affordable buy.

♣Oh What A Tangled Web We Weave♣

"You Ain't Heard Nothing Yet, Folks."   -Al Jolson

This is probably as good a point as any in our journey through High Definition to take a deep breath and compare theory with reality.

One important reason that digital TV appears to be sharper and clearer than analog is that analog signals can be affected by interference and still viewed on our TVs, although at degraded quality. In contrast, digital transmissions either come in with a strong clear signal or they are unwatchable, with no sound; so if we're able to view a digital channel at all, that usually means that the signal is going to be good. With analog, the picture and sound are independent; with digital, they are integrated. A weak signal not only makes the picture unwatchable, it kills the sound.

An important benefit of the switch to digital-only broadcasting is that it will free up parts of the broadcast spectrum for public safety communications... police departments, fire departments, rescue personnel, etc... a whopping four channels→ 63, 64, 68, and 69. So we have Four New Channels for public safety, and we have Fourteen New Channels auctioned off.

Although the radio voice-communication part of the public safety efforts is proceeding, first responders will not have an emergency broadband data network any time soon. The 698 to 806 MHz freed up by the switch to digital TV has hit some snags. And so there will be no national broadband data network for safety purposes anytime soon, all four channels of it.

Digital is more efficient than analog. Rather than being limited to providing a single analog program, a digital broadcaster is able to offer a sharp "High Definition" (HD) digital program (if the viewer has an expensive sharp HDTV set); or he can provide multiple "Standard Definition" (SD) digital programs through digital "multicasting" simultaneously.

Broadcasters have the option of running a breaking news story on one of their sub-channels. For example, during the Rod Blagojevich impeachment trial, one ABC station showed the impeachment coverage on one of its digital sub-channels, while running soap operas as normal on another sub-channel.

♣Get An HDMI Cable Or Get Component Cables?♣

"With Man, Most Of His Misfortunes Are Occasioned By Man."   -Pliny The Elder

   
DIGITAL-------->SET-TOP BOX--------->======ONE HDMI CABLE=========>HDTV SET (Best)
 CABLE           TO DECODE          |
ENTERS         DIGITAL CABLE        |=====>5 Component Cables=====>HDTV SET (2nd Best)
 HOUSE     (WITH HD SUBSCRIPTION)
   |
   |
   |---------->DIGITAL VIDEO-------->======ONE HDMI CABLE=========>HDTV SET (Best)
               RECORDER (DVR)       |
           (WITH HD SUBSCRIPTION)   |=====>5 Component Cables=====>HDTV SET (2nd Best)



CABLE--------->SATELLITE------------>======ONE HDMI CABLE=========>HDTV SET (Best)
ENTERS         HD RECEIVER          |
HOUSE          SET-TOP BOX          |=====>5 Component Cables=====>HDTV SET (2nd Best)
FROM       (WITH HD SUBSCRIPTION)
SATELLITE
DISH

Component Video Jacks On HDTV An HDMI Jack (Vers. 1.3) On HDTV

THE NITTY GRITTY→ Unless your antenna is receiving an over-the-air broadcast, you will need to connect the digital decoder box that you rent from your cable or FiOS or satellite provider to your HDTV set. You have two choices in your cabling→

1. An HDMI cable. (A "High Definition Multimedia Interface" Cable)

2. 5 Component cables.

THE REAL STORY→ Dog Wolf... Help. We need an explanation of why HDMI is better than just a plain old coaxial cable. Like, the signal arrives through a coaxial cable, so how can a different interface from the set-top box to your HDTV set provide a better (or worse) quality picture and sound?

ANSWER→ Let's assume that you subscribe to digital cable. The set-top box is there to tune and decode "digital cable", NOT analog. And this is the key to why an HDMI interface is far better than standard coax. With new display technologies, such as DLP, LCD, LCoS, and Plasma able to receive digital signals, and HDMI able to transmit uncompressed digital signals, the HDTV set is receiving the most error free signal possible. HDMI supports, on a single cable, any TV video format, including standard definition (SD), enhanced definition (ED), or high definition (HD) video, plus multi-channel Dolby (5.1) sound.

If you were to use the RF output from the set-top box, or even use the S-Video or baseband outputs, the box still must first receive and decode the incoming digital signal, and then do a digital to analog conversion, then re-modulate the signal for use on an analog input on an HDTV set. As you can guess, you LOSE information in doing this, not even to mention that these outputs are not capable of carrying High Definition.

Even though the YPbPr outputs are capable of passing 1080i30, there is a Digital-to-Analog conversion that has inherent signal degradation. With HDMI able to transmit multi-channel digital audio as well as HD video, we surely would use HDMI over coax. In fact, the worst picture we've ever seen on a flat screen HDTV set was on a 52" 1080p60 LCD Samsung... the HDTV set was connected to a DVR box by a length of coaxial cable.

Even a "minimalist" cable box, which takes in its video content in analog form ("Basic" Analog Cable TV), will improve the picture on the screen with an HDMI connection. Here, the analog signal is converted one time, in the set-top box, and then sent to the TV in digital form over the HDMI cable. The analog-to-digital conversion in the set-top box is actually quite good in most cases, even though the incoming analog is limited to Standard Definition (SD), and can never be HD. (Remember we talked about no HD with coax cable? Great.)

Problems With HDMI→ There are three potential problems that will prevent you from using an HDMI cable and force you into using component cables→

1. If your source (like a DVR you're renting from your cable provider) does not have an HDMI output, obviously you're not going to get very far with HDMI. (You may want to try politely screaming at your provider, asking for a DVR or set-top box with HDMI outputs. Also mention the word "FiOS" to your cable company; that may help.)

2. HDMI has a built-in copyright protection system called HDCP; if the microcode in your cable or satellite or FiOS set-top box is not correct, HDMI will block an HD signal from ever reaching your HDTV set. Your cable provider can update your microcode remotely, but the newer microcode has to be available, they have to have the inclination, etc. There is a famous blog about the incompatibility between Comcast's microcode in a Motorola set-top box and a Samsung HDTV.

3. HDMI will NOT carry closed captioning. The specifications for HDMI allow only the part of the frame you actually can see to be carried by the HDMI cable; there are no "invisible" lines transmitted, which is how closed captioning is sent to the TV screen. If this is a problem, then you want to use component cables. Alternatively, the source (DVD, cable box, whatever) can be adjusted to extract the closed captioning text and insert it onto the visible portion of the frame; this then will go across the HDMI cable just fine, and it will display on your HDTV set.

Assuming that you have neither of these three problems, you want to use an HDMI cable. HDMI video quality is better than component cables. The HD picture using component cables is fine; but the picture with an HDMI cable is better.

Now, all HDMI cables are not created equal. We tried one from RadioShack, and the picture was comparable to component cables... still nice HD. But when we tried an HDMI cable that was purchased by Precise Networking from an alternate source, we improved our HD picture; you may want to experiment.

And remember that HDMI is not only a cable that carries high resolution video plus Dolby Surround Sound in digital format, it’s also a standard; and that standard will continue to evolve, continue to incorporate new features to make connecting various components easier, to support future higher resolution devices. The newest version of HDMI, announced 27 May 2009, is HDMI 1.4; it adds many new features that gradually will become more and more important to folks over the next several years.

DEEP COLOR→ HDMI 1.3 supports 30-bit, 36-bit, and 48-bit (RGB or YCbCr) color depths, up from the 24-bit depths in previous versions of the HDMI specification. This allows HDTV sets to go from millions of colors to billions of colors. It eliminates on-screen color banding, for smooth tonal transitions and subtle gradations between colors. (Banding is something bad; if you have an image with a smooth gradation from, say, black to red, and if your HDTV, instead of showing that smooth change, instead shows bands of several intermediate colors, then that is banding.) HDMI 1.3 also enables an increased contrast ratio.

HDMI 1.3 can represent many more shades of gray between black and white. At 30-bit pixel depth, four times more shades of gray would be the minimum, and the typical improvement would be eight times or more. HDMI 1.3 promises finer gradients between colors and a wider gamut (range) of displayable colors. It's a good idea to assure HDMI 1.3 cable compatibility when shopping for a new HDTV set.

The broader color space in version 1.3 of HDMI cables virtually removes all limits on color selection. Next-generation xvYCC color space supports 1.8 times as many colors as existing HDTV signals. HDMI 1.3 allows HDTV sets to display colors more accurately. And it enables displays with more natural and vivid colors.

Since the single HDMI cable carries uncompressed video (using the TMDS format), there IS some encoding/ decoding performed at each end (decompression and recompression); but this is not as severe as component cables doing digital source→ analog component cables→ digital HDTV set conversions.

The all-digital video with an HDMI cable is sharper than component connections; and using a single HDMI cable eliminates the softness and the ghosting occasionally associated with other types of connections. High contrast details bring out this difference the most. Be sure that your HDTV set accepts at least one HDMI cable.

However, component cables easily handle 1080p120, and the picture is never "bad"; it is just not quite as good as HDMI. Component cables, having to convert digital to analog at the set-top box, and analog back to digital at the HDTV set, is slightly "lossy"; detail is lost. Conversely, HDMI remains digital throughout. A six foot HDMI cable purchased online will cost about $20 (2010); don't let the Big Box stores hoodwink you.

(Yes, we know that DVD→HDTV requires reclocking with slight loss in picture quality using an HDMI cable; in our humble opinion, an HDMI cable still looks better than component cables. If you're unsure, as we were, simply try both; but use a quality HDMI cable.)

Separate video, abbreviated S-Video, with brightness and color separated, will not support HD.

♣NOW AT LAST WE'RE READY FOR THE HD STEAKS
AND THE TOP SIX ATSC HD STEAKHOUSES- MMMMM♣

"High Definition Has To Be Given Meaning Because Of The Obvious Fact That It Has No Meaning."   -Henry Miller

In order to understand "HD" (High Definition Television), we first must understand the new ATSC DIGITAL standards... especially the ATSC digital broadcast formats.

The ATSC standards for digital television have "format codes". You can vary the number of rows of pixels, you can vary the number of screens (frames) that "refresh" per second, the number of pixels per line will vary, and so on. In fact. there are 18 digital ATSC TV formats. Here's a table listing all 18 approved ATSC video formats→

ATSC DIGITAL TV FORMAT CHART

* US HD broadcasts use one of two formats, either 1080i30 or 720p60. (Or so the story goes.) But MPEG-2 actually is quite flexible. Some NBC stations (and others) have been broadcasting with 1080p24 and 1080p30 formats. "Metadata" is broadcast along with the video and cues the decoder in your HDTV set to turn these 1080p rates into 60 interlaced 540-line fields per second (1080i30), to remove flicker. (How your HDTV set responds to this metadata is up to your particular set.)

** In addition to the two video formats used for HD broadcasts in the US, 1080p60 is defined by the ITU (International Telecommunication Union) as a standard for HD, though it is not used for over-the-air broadcasts in the US. Why not?→

• First, 1080p60 would take more bandwidth than the 6 Mhz that the FCC allows for each digital channel.

• Second, until late 2006, HDTV sets which could display 1080p60 could not input 1080p60 from DVDs. Today (January 2010), 1080p HDTV sets CAN input 1080p60; and you can get 1080p60 from the new generation of DVD players (Blu-ray Disc).

• DISH Satellite has presented video on demand movies at 1080p60.

♣MORE ON HD♣

"Television Has Raised Writing To A New Low."   -Samuel Goldwyn

Imagine that the 18 ATSC approved digital TV picture formats (the FCC's "Zagat Of Digital") are 18 steakhouses; then HD is the best six of the 18... the Smith & Wollensky, the Peter Luger, etc... the six digital video formats that rank at the very top in video quality, the six video formats serving up only dry-aged, prime digital video steaks... the very best quality pictures that are ATSC approved. (One of DAwn's life passions is a medium-rare prime dry-aged NY strip, served up at some great steakhouse.)

THE REAL STORY→ If you are an engineering-type, you might also like to have an engineering type definition for HD→ If the data rate from studio cameras and other video sources going into the broadcast Video Subsystem, the place where Video Source Coding and Compression take place is 1.0 GBits/sec (One Billion bits per second) or higher (1.485 GBit/sec is a very common number), then we say that the source of the broadcast is High Definition, and that we are almost always transmitting the program in High Definition.

At such a data rate, the video subsystem normally produces video at 720p60 or 1080i30 (different TV networks and TV stations use different video formats; e.g., ABC uses 720p60) almost always with an aspect ratio of 16:9. (Sometimes the original material will have an aspect ratio of 4:3, but it will have a data rate that qualifies for High Definition; the video subsystem usually will stretch out the aspect ratio to 16:9.)

EVEN DEEPER→ There is another subtle difference between HD digital and analog and non-HD digital. If for some reason a black and white source (an old Bogart movie, for example) is broadcast in HD, you may notice that it looks a little greenish; but if it's broadcast in SD (standard definition), it'll look fine.

Television creates black and white for "color sets", including all HDTV sets, by mixing red, green, and blue. HD adds in a little more green than analog did (or SD digital does). HD does this because it allows a slightly wider range of colors to be displayed; and HD wasn't created with black and white in mind.

For the technical-minded, HD black and white brightness is created by mixing colors in these amounts→ (0.21 × Red) + (0.72 × Green) + (0.07 × Blue). This mixture is also known as the "Rec. 709" color space.

SD and analog create a black and white picture by mixing colors in these amounts→ (0.30 × Red) + (0.59 x Green) + (0.11 × Blue), known as the "Rec. 601" color space. (Rec. 601 is a nickname for "ITU-R Recommendation BT.601, the international standard for television studios' non-HDTV digital signals".) These proportions were selected in 1950 or so, when getting black and white as right as possible was paramount, and showing a wide range of colors was a thought for the distant future, for the 21st Century.

IS 480p HD?→ 480p is not HD. 480p (the family that includes 480p60, 480p30, and 480p24) is defined in the US ATSC standards as ED (Enhanced Definition). (Folks who mistakenly have connected a converter-box to HDTV sets... yes, they do... Precise Networking Solutions has had to "de-convert" several HDTVs... can watch all broadcasts as ED (with an aspect ratio of 4:3) .) EDTVs (Enhanced Definition TVs) display 480p fine, but they cannot display HD (720p, 1080i, or 1080p).

♣COLOR ME RED♣

"The Best Way To Predict The Future Is To Invent It. "   -Alan Kay & The FCC

♣Why Plasma And LCD HDTVs Display
All "i" Pictures As "p" Pictures♣

"Very Often The Only Way To Get A Quality In Reality Is To Start Behaving As If You Had It Already."   -C.S.Lewis

First→ Note that with 1080i, all of the odd numbered lines reach your HDTV set before the first even numbered line is broadcast.

If a 1080i30 HD picture were displayed on an HDTV set's screen, alternating rows of pixels ("interlaced" rows) would be energized... first the odd numbered rows, and then the even numbered rows.

So now the $64 question→ Wouldn't having only half of the pixels active at any given time during a 1080i HD broadcast reduce the brightness of the screen by ½ if we had say a 720p60 HDTV set? Our set would display lines 1,3,5,7... and at any instant, half the lines on the screen would be dark. Because a 720p60 HDTV set is not designed to display an interlaced broadcast. The same for a 1080p60 HDTV set.

On the old CRT "picture tubes", we don't have this problem. In the picture tube, an electron beam hits phosphors coating the inside of the front of the tube. WHAM... A point of light forms where the beam hits the phosphors. And that point of light runs across the screen just as we read a page in a book. There is no problem with half the phosphors on the tube not working during each ½ frame.

The Real Story→ All flat panel screens (LCD and Plasma) are designed to be progressive, no matter what format they are receiving.

Every flat panel HDTV set converts every incoming "i" picture to a "p" picture before it displays it. Some do it well, some not so well. Because every LCD and plasma HDTV set is going to display every frame progressively... 1,2,3,4,5... (it doesn't want to know from "interlaced"). So what happens when a 480i30 (a non-HD broadcast) or a 1080i30 (an HD broadcast) video signal comes into our flat panel HDTV?

The odd lines (field number one) first go into "memory" in the plasma or LCD HDTV, where they just sit and wait. Then the even numbered lines for the next field arrive (field number two) 1/60th of a second later, and they're added to the half-picture that's been waiting in memory. And then, and only then, the whole picture is displayed from memory (every LCD and plasma HDTV set have this memory)... lines 1,2,3,4,5... almost as if it were a genuine "p" broadcast.

LCD and plasma flat screen HDTV sets use what engineers call a scaling processor, combined with "frame memory", to convert an incoming "i" (interlaced) picture into a "p" (progressive) display. This process of converting "i" to "p" is called de-interlacing. Here, de-interlacing is the process of converting interlaced video (a sequence of 60 fields per second) to a non-interlaced format (a sequence of 30 progressive frames per second).

The screens on LCD and plasma HDTV sets think they're getting a progressive signal every 1/30th of a second when they're receiving an interlaced broadcast; we don't tell them about 1080i30.

Now, let's quickly scoot back to the TV station. During the creation of the 1080i picture at the TV station, objects in front of the camera (like a football) often move between the camera capturing an image for its odd lines, and capturing an image for its even lines. And therein lies a problem.

In other words, the two frames that are combined in our HDTV set were taken at different times, even though the different fields were created only 1/60th of a second apart. This can create a blur and/or some creative artifacts. The good news is that folks don't concentrate on details when objects are moving rapidly. If a football player is running for a touchdown, we don't care if the team logo on his helmet is a bit blurry.

Technically, there is no reason why plasma or LCD displays could not display the odd numbered lines on their screens first, and then they could display the even numbered lines next, without bothering to de-interlace anything, exactly as CRT "picture tubes" do. And this would eliminate the flicker and the artifact associated with de-interlacing.

But this means that at any given point in time, only ½ of the pixels on the screen would be energized (the other ½ would be dark); and this would halve the brightness of the HDTV set.

For this and other reasons (such as the lower response times of the new flat-panel technologies, LCD and plasma, compared to the older "picture tube" technology), interlaced video just does not work well with the new flat-panel HDTV sets. It works fine with the older CRT technology, so we don't need de-interlacers in our picture tube sets.

De-interlacing is not a perfect process; it's what engineers call a "lossy" process; information is lost. We're combining two pictures taken 1/60th of a second apart, and we're pretending that they can be combined into a single whole screen picture. Will we notice the time distortion in a slow, passionate love scene? No. In an action-packed basketball game? Much more likely.

In fact, when the ATSC digital standards were being drawn up in the 1990's, there were arguments made against the 1080i broadcast video format; but it's here to stay, at least for many years. (Go tell NBC and CBS and HBO, all of whom broadcast their HD programming in 1080i, to swap out all of their expensive 1080i studio equipment, including cameras, and replace it with 720p.)

The Reality Of Interlaced Transmissions→ In one broadcast, any network or local TV station may flip from interlaced to progressive, back and forth. A commercial may be interlaced, the main program may be progressive. One song may be progressive, the next may be interlaced. In fact, one frame (like a congressional discussion on the economy), with a tiny picture "keyed" into it (the LA daily high speed chase), may be both progressive and interlaced.

Good HDTV sets handle this chaos well; sets of lesser quality handle it less well. The next time you have six people on the screen (Fox News) all talking at the same time about why President Obama is to blame for the extinction of the dodo, think to yourself... "Hmmm, I wonder which of the six small pictures are "p" and which are "i", and I hope my HDTV set doesn't catch fire from the strain of de-interlacing the "i"'s."

And of course, there is no reason that a local network affiliate station cannot change the format of the "feed" that it gets from the network.

Thus, if station WQQQ is an ABC affiliate, and ABC feeds "Grey's Anatomy" in HD at 720p to WQQQ, WQQQ would normally broadcast "Grey's Anatomy" to its local area at 720p.

BUT... If for some reason WQQQ switches and becomes an NBC affiliate, do you think it will swap out millions of $$$ of 720p equipment, just so they can match the NBC HD standard and transmit at 1080i? WQQQ will simply take the NBC feed of say "Leno", at 1080i, "down-convert" the NBC feed to 720p, and transmit "Leno" at 720p. Works just fine.

So what's our point here?→

Our point is that sometimes it's hard to know which HD video format (including the resolution) is coming into your HDTV set over-the-air (or via cable... same thing). The only thing you can be sure of when watching an HD broadcast is this→ If your HDTV set is a 720p, then no matter what comes into it, it will display it as 720p60.

And if your HDTV set is a 1080p plasma or LCD, it will display 1080p60 no matter which format ATSC HD (or even analog 480i30) comes into it. (Yes, HDTV sets convert an analog input automatically into digital.) HDTV sets are accomodating and smart.

Returning to the subject of interlacing for just another moment→ Some folks will claim that the old picture tubes (the "CRTs"... Cathode Ray Tubes) retain the two FIELDS, 1/60th of a second apart, on the phosphors that coat the picture tube screen... and that the CRT retains its picture for so long that de-interlacing isn't required... folks call this alleged retentivity effect "afterglow".

DEEP DARK SECRET→ Afterglow's a myth.

The fact is that on the CRT "picture tube", fewer than 100 lines out of 480 visible lines per frame are illuminated at any instant. In other words, pictures on CRT picture tubes fade rapidly; they fade in about 1/200th of a second.

(Ever take a photograph of the screen of a CRT at 1/1000th of a second to see what's really going on? 90% of the CRT screen will be dark at any given instant. But set your shutter speed at 1/30th of a second and your photo of the screen should be just fine.)

Anyway, de-interlacing is not a perfect process. As we said, you are combining two sets of lines that were captured at slightly different times. What if the football moves between the captures of the odd and even lines? When you combine the two sets of lines for simultaneous presentation, you will have artifact (stuff appearing on your HDTV screen that does not exist in reality, like "teeth" on the right side of the football... or even "ghosts"). (Think this may be why ESPN (Entertainment and Sports Programming Network) chose to broadcast in 720p60 instead of 1080i30?)

Fortunately, there are expensive and exotic ways to perform de-interlacing, and these account for a great many patents. Higher-quality HDTV sets usually have higher-quality de-interlacers, employing more complex techniques to reduce artifacts and reduce flicker.

The use of inexpensive de-interlacing hardware is a common difference between lower-priced and higher-priced flat panel displays.

Summary→ When receiving a 1080i30 broadcast, to get all 1080 interlaced lines on a flat panel HDTV screen at the same time (progressive display), the processor in the HDTV set "weaves" together both 540-line half-frames (fields) to form a full 1080p frame. It does this by holding the first half-frame in its memory, receiving the next field, and then "knitting" the two fields together. In this process, many de-interlacers must throw out some lines and/or pixels to reduce flicker and artifact.

Now you can see why 1080i broadcasts are perhaps only 10% sharper on average than 720p, despite having more than twice as many pixels. 1080i has to throw away details during the de-interlacing process, while 720p is not interlaced and doesn't need to be de-interlaced; and thus doesn't need to throw away any details.

The combined fields are displayed at once as a complete 1080p frame (or they're reduced to 720p if the HDTV set is a 720p... "We try to do what we can." -Hoke Colburn (Morgan Freeman), in Driving Miss Daisy, 1989).

Once Again Please→ Ok, let's assume that we have a Panasonic plasma 720p60. This HDTV set can receive the old analog NTSC 480i30 just fine. So that's great, but how do we get from 480i30 to 720p60?

In three steps→

1. First, we do an analog-to-digital conversion... our input was analog, our TV is digital.

2. Second, we perform de-interlacing, which converts 480i30 to 480p30.

3. Third, we perform "interpolation", which uses "spatial filtering" to generate extra lines (that aren't actually broadcast) but reduce the visibility of "pixelation" on any type of display. Interpolation gets us from 480p30 to 720p60. Interpolation is performed by a video scaler (or "video processor")... usually inside your HDTV set, though you can purchase scalers as stand-alone external boxes.

   Video scalers convert video to one of many resolutions (480p30, 720p60, 1080i30, 1080p60). Scalers perform up-conversion, a process by which a video signal is taken from an "inferior format" to a "superior format"... like from component video to DVI/HDMI (remember HDMI?). Quality video scalers can make standard definition (SD) television (480i30) look cleaner and more appealing when viewed on an HDTV set's screen than when viewed on an old CRT.

   Today, the term Video Processor is more common than "Video Scaler". Technically, a Video Processor covers all the possible features, such as de-interlacing, scaling, and conversion to different video formats. The Video Processor is largely responsible for the improvement that non-HD digital and NTSC analog undergo when displayed on an HDTV set.

And this is why even the old NTSC 480i30 broadcasts and other non-HD video signals generally look better on the new HDTV displays. (And the improved video quality of LCD and plasma, and the big 16:9 screens don't hurt either.)

♣Oh, Is HD Ever Compressed♣

"The gods Give To Mortals Not Everything At The Same Time."   -Homer and ATSC

Digital TV transmits a stream of bits. Since a bit stream rather than a continuous waveform are broadcast, digital allows near perfect transmission. But "near perfect" comes at the expense of bandwidth; digital requires much more bandwidth than analog. But the FCC said, "Bandwidth is dollars; use the same 6 MHz channels for digital and analog. So the only solution was to compress the digital signal. This compression was accomplished by using the MPEG-2 technique.

Ever see an HD camera working away in some TV studio? Studio HD TV cameras produce a raw video stream in excess of one billion bits per second. If the picture from an HD camera is ever going to fit in the skinny 6 MHz broadcast channel allowed by the FCC, something dramatic has to be done first.

That "something dramatic" that the ATSC digital standards require is called MPEG-2 Compression, which occurs in the Video Subsystem at the TV station. In a nutshell, MPEG-2 is the process that ATSC digital TV uses to convert the images in the TV camera from a torrential bit stream into neat rows of data packets, achieving massive compression in the process. (We'll dissect MPEG-2 in more detail in a bit.)

Don't Be Confused→ MPEG-2 is the ATSC standard for compressing digital television signals prior to 8-VSB modulation (in the "exciter") and broadcast from the antenna on top of the TV tower.

Cable and satellite providers add compression in addition to studio MPEG-2, compression which is also "lossy". Lossy means that when you decode the additional compression added to over-the-air broadcasts by cable and satellite providers→

1. The picture is degraded. When your set-top box de-compresses the additional compression added by the cable company, some of the bits that were present when the cable and satellite providers performed the additional compression are not recovered, lost forever. Hence the term "lossy".

2. You still have an ATSC digital TV signal that follows over-the-air standards and is MPEG-2 compressed. (It's up to your HDTV receiver to de-compress only the MPEG-2 compression applied at the TV station. The additional compression added by your cable and satellite provider must be removed by their set-top boxes.)

Note→ Many cable providers and most satellite providers are currently using a newer collection of compression methods called MPEG-4. This MPEG-4 compression is stacked on top of the MPEG-2 compression applied at the TV station. MPEG-4 is a ten year old set of standards; it's up to each provider to decide exactly which compression standards he wants to employ.

FiOS→ FiOS adds no additional compression to HD programs.

SO... In sending ATSC digital TV over cable and satellite, a TV signal whose picture is already MPEG-2 compressed (and whose Dolby sound is already AC-3 compressed), is further compressed by non-standard, non-ATSC, cable and satellite techniques. FiOS is not further compressed; and this is why over-the-air and FiOS programs look (and sound) better than cable or satellite.

Using ATSC Digital Standards (A/53), an ordinary 6 MHz TV channel can broadcast over-the-air at up to 19.39 million bits/second of video and sound and data. And before we actually modulate and broadcast over-the-air, we add a lot more bits, mainly for "forward error correction". We thus end up transmitting well over 30 million bits/second; but 19.39 MBit/sec is the bit rate limit that can exit our Service Multiplex, the guy who combines our compressed video and our compressed audio plus a little ancillary and control data.

So some of that 19.39 million bits/second must be subtracted for sound (recall Dolby?) plus other non-video data. This leaves about 18 million bits/second for the HD picture, the video, with a format of either 720p60 or 1080i30.

SO... MPEG-2 compression has to take us from way over a billion bits per second to under 20 million bits per second. That's better than a 50:1 squeeze. So what is this MPEG-2 anyway?

What'S MPEG-2→ ("MPEG" stands for "Motion Pictures Experts Group".) MPEG-2 is the "packaging" format for the streams of bits and bytes in a digital television signal. MPEG-2, in this capacity, is also the standard for video compression that's specified in the ATSC Digital Standards. (MPEG-2 also was a contender for the ATSC digital audio compression technique, but it lost that battle to AC-3 compression; AC-3 is Dolby.)

MPEG-2 typically can squeeze a billion+ bits/second all the way down to 20 million bits/second. MPEG-2 uses several tricks to achieve this feat; one is to make sure that anything in the picture that has not changed since the last frame is not re-broadcast. A cowboy rides through our picture under a blue sky, the sky doesn't change, MPEG-2 says "Do not rebroadcast the sky".

And there are things in the picture we can throw out. We never notice.

Never even notice?

Nope, and that's what makes possible the "thinning out" process that MPEG-2 uses. Every pixel coming from an HD TV camera is associated with three numbers... one for the brightness, two for color. (To be overly technical for a moment, these three numbers are the three axes of a "color space". Given brightness plus two color numbers, we can locate a unique color.) But can we really discern the color of each pixel? No way, José.

So MPEG-2 compression may throw out perhaps half the color values. Make the pixels without color values be the same color as their nearest neighbors. We don't notice this trickery because our eyes are more sensitive to brightness than to color.

In HD, we end up with about 270 rows containing color, out of a total of 720 or 1080 rows. In analog, we have 120 lines with color, out of 480 total lines.

MPEG-2 compression uses the notation 4:2:2 to mean that half of the color values have been deleted. And the notation 4:2:0 says that three-quarters of the pixel color values have been deleted. (If no color information has been deleted, the notation is 4:4:4.) MPEG-2 compression permits all of these options. 720p60 and 1080i30 both use 4:2:0 (meaning three-quarters of the pixel color information is thrown out). ¿Está claro?

How else does MPEG-2 compress, besides not re-transmitting the same stuff and throwing out color values? Here are three additional techniques employed in digital broadcasting's MPEG-2 compression→

1. Static information in an image, like an intricate design, can be transmitted infrequently, relying on information stored in memory.

2. Pictures like a woman dancing rapidly across a ballroom floor must be transmitted more frequently than static, intricate patterns. But they can be transmitted with less detail. The human eye cannot follow patterns that change rapidly.

3. And what about intricate patterns that move slowly (e.g., face shots of two lovers slowly approaching to kiss)? Engineers supplement the usual transmissions with "motion vectors" that provide the receiver with additional data.

But as always, there are no free lunches. MPEG-2 is a "lossy" compression technique, meaning that the more you compress a TV picture before broadcasting it, the poorer the picture's quality when it's received and expanded again. In other words, there's a loss in picture quality between the HD TV camera and what appears on your HDTV set, but nothing too serious. MPEG-2 is relatively "gentle".

Now... why can't we use compression to broadcast 1080p (1080p60)?

Well, HD at 1080p60 generates twice the bit rate of 720p or 1080i; namely, it needs to transmit at 36 million bits/second. And the ATSC Digital Standards for over-the-air broadcasting do not permit such a high bit rate in a 6 MHz TV channel; neither does current engineering technology.

Let's look at cable for possible relief. A 6 MHz digital cable channel can accomodate 38.78 Mbps, using 256-QAM modulation (we'll explain that in a bit). And so a 6 MHz digital cable channel can carry 1080p60 video, well, in theory, at least.

Will cable providers send 1080p60 video to our homes? Cable providers are having enough trouble finding the bandwidth they need to carry 720p60 and 1080i30 HD broadcasts. The new Blu-ray Disc DVDs can inject your 1080p HDTV set with HD at 1080p60, and it looks good.

The Real Story→ For engineering reasons, progressively scanned images ("p") compress better than interlaced images ("i"). At the TV studio, the "raw" data coming from an HD TV camera pours out, usually at 1.485 billion bits per second.

Well, this looks better→ Since "p" compresses better than "i", and 1080p60 is "p", and 1080p60 generates twice the data flow from the TV camera compared to 720p or 1080i, we might expect that 1080p60 would fit nicely in a cable channel that can hold 38.78 million bits per second.

It does fit. But it does not look good on an HDTV screen after its compression. Compression produces artifact, and compressing 1080p60 from about three billion bits per second in the TV studio down to 38 million bits/second simply produces too much artifact, at the limits of current technology. The picture looks worse than Standard Definition (SD). It looks awful.

Technology continues to improve, but it's far more likely that broadcasters will continue compressing 720p60 and 1080i30, so that they can squeeze more stuff into a broadcast (or cable) channel.

♣What Is Analog? What Is Digital?♣

"A Man's Mind Is Known By The Company It Keeps."   -James Lowell

Analog TV Vs. Digital TV→ At a TV station, a program starts by focusing a television camera. The camera changes the brightness and color of the scene into an electrical video signal; that is, the light image in the TV camera is changed into an electrical image on a "target". This electrical image is "read" from the back of the target by an electron beam scanning the image.

With analog TV, as the camera scans the image, the video signal from the camera varies, depending on the brightness and color of each point in the scene. When the electron beam in the camera hits a bright spot in the image, the camera emits a high voltage; hit a dim spot, you get a lower voltage. This video voltage is sent through cables at the TV station, ending up finally at the TV transmitter, which broadcasts the video signal to our TV sets.

At any instant in time, the signal being sent by the analog transmitter represents the brightness and color being scanned inside the camera. After the transmission is received by our analog TV set, one or more electron beams in the set recreate the original image on our set's picture tube. In theory, the video signal broadcast from the TV station can have ANY value between white and black. It has many, many discrete states. And the brightness and color of the beam(s) in our TV set are identical to the brightness and color of the image being scanned in the TV camera at that instant; or in the terminology of TV broadcasting, they are in "sync"; there is no holding buffer. The sound is totally separate; it is sent using a a separate FM transmitter, and there is an FM receiver in every TV set.

With digital TV, the video signal that is produced by scanning an image is converted into binary numbers; most digital signals are generated at about 1.5 BILLION bits per second. This video signal is then compressed. The sound, also a string of bits, is also compressed; the video and sound bit streams (and other stuff) are then mixed together. This combined bit stream eventually reaches the digital transmitter.

The digital TV transmitter sends out one bit stream representing the picture, the sound, and other stuff roughly ten million times each second, each time as a string of 3-bits (eight different values, recall?). Our digital TV receiver samples the signal from the transmitter ten million times per second, and it expects to get one of eight discrete values. After all, it is digital. And moreover, the picture ultimately displayed on our home digital TV screen is NOT in sync with the picture in the camera; there is buffering, and picture and sound lag a little behind reality. (More on this coming up.)

♣Analog And Digital... The Past And The Future Of TV♣

"With Regard To Excellence, It Is Not Enough To Know It, But We Must Try To Have It And Use It."   -Aristotle

Since 1946, millions of us have been receiving analog broadcasts from antennas on thousands of TV towers across the US, some on top of skyscrapers serving cities, some 2,000 feet high self-supporting on the plains serving entire states. And most parts of the TV signal transmitted from analog TV antennas varied continuously, signals taking on a great many values, more values than we could count.

Introduction→ The analog signals radiating from transmission antennas high up on TV towers are different from the newer digital signals. The signal voltages broadcast from an analog antenna is wavelike. And theoretically, ignoring quantum effects and noise, there are an infinite number of values that the analog wave leaving the TV antenna can assume, anywhere from some minimum amplitude to some maximum amplitude.

On the other hand, a digital transmission is based on a digital format, based on strings of ones and zeros (bits). The digital signal cannot have an infinite number of values; in fact, under the ATSC plan, it only can have eight discrete values. Eight values are represented by 3-bits.

NTSC Analog→ Video carriers being VSB modulated (a type of AM modulation), color carriers being quadrature modulated (before they are suppressed), sound carriers being frquency modulated... all analog. And we care about their values continuously; we don't simply sample the signal from the analog TV station 10.7 million times per second; we feed those signals through circuits in our analog TV set's picture tube and speakers continuously.

In an analog TV transmission, very small changes in the picture part of the signal, intentional or unintentional, will change the picture on our analog TV sets. Maybe the grass will turn blue (unintentional); maybe bursts of "snow" will interfere with the game-winning touchdown (unintentional, unless the game is being played in Buffalo), and so on.

And on the other hand, the components of an ATSC digital broadcast are represented by just three bits, representing eight "symbols", 10,760,000 of these "trellis numbers" broadcast each second... three bits that must be sampled at just the right time in digital receivers. Because between the "right times" there is only garbage.

ATSC decided that the TV signal can have precisely eight discrete values at a very precise sampling time. (Between these 11 million sampling times each second, we have garbage. And if we get something other than one of these eight values when we sample, we may guess at what the value should have been, as long as we're not too far off the mark; or perhaps we'll let the digital TV station realize that it made an error, and we'll let it retransmit a little bit of the digital signal.

Analog broadcasts are like the old vinyl records; the single groove on these records varied continuously, they were ANALOG records. In Analog TV Broadcasts, the frequencies received by our analog TV sets have been varying continuously, the amplitudes have been varying continuously, the phases have been varying continuously; virtually anything that could vary has been varying continuously. And any value is legitimate; our analog sets can't tell snow (interference) from snow (during winter weather).

Another way to think of digital is that digital TV is sent out as 0's and 1's (as BITS), not as a continuously varying analog signal. 10,760,000 of these "trellis numbers" per second, each three bits long, each in one of eight possible states, each representing one of eight symbols.

Quick Review Of Digital Bandwidth→ All digital information is a series of 0's and 1's. The more zeros and ones broadcast per second, the higher the data rate; the maximum for a digital broadcast in the US is a data rate of 19.39 million bits per second coming from the multiplexer, going into the exciter; and after forward error correction bits are added, perhaps 32 million bits per second are flying from the transmitting antenna, in their assigned 6 MHz channel.

In digital TV studios (studios that prepare programs that are broadcast digitally), prior to compression by MPEG-2, High Definition data is routed around at 1.485 Billion Bits per Second. See the need for compression? Digital TV takes up more bandwidth than analog, but it compresses better.

(Clarification→ The TV broadcast tower raises the TV broadcast antenna as high as possible (without interfering with other antennas that may be sharing the same tower). The objective is to get that piece of metal, the TV broadcast antenna, up as high as possible. The higher up you get it, the farther it's signal will travel. The tower itself is just there to get that piece of metal, the antenna, as high as possible. The tower doesn't broadcast anything. AM radio towers, on the other hand, broadcast from top to bottom→ the entire commercial AM radio tower is the antenna.)

Stated Another Way→ The conversion to digital-only is changing the language that TV transmitters use to speak to your television. Since about 1946, US TV stations have transmitted using standards laid out by the NTSC in 1941. Big TV towers (get that antenna up there as high as possible) transmitted pictures and sound over-the-air speaking the NTSC language; these transmissions were received by the little antenna on your TV set. And inside your TV, there was of course an NTSC tuner, converting analog broadcasts to analog pictures and analog sound.

The June 2009 conversion to digital-only filled the ether with a new language, new standards designed by the ATSC. All of the US broadcast stations, the ones categorized as "full-power", no longer are speaking NTSC; instead, they're speaking ONLY ATSC. Your old analog TV can't translate the new ATSC into pictures and sound. Almost all TVs sold in the US since March 2007 have ATSC tuners and can understand ATSC digital. (Sets bought before that probably have no ATSC tuner.)

The Solution→ A converter box is not much more than an ATSC tuner sitting between your antenna and your old NTSC TV set. The ATSC digital signals reach your antenna, then are passed to the converter box that translates the ATSC signals into the old NTSC language.

♣Strange  Channels♣

"Plurality Should Not Be Posited Unnecessarily."   -Ockham

The analog NTSC signal is a brilliant example of technical compromise. When the system was being formulated, engineers appreciated that the scarce resource was the limited number of affordable vacuum tubes. They were expensive, large, and they ate a lot of power. And only a handful could fit in a receiver (physically and dollar-wise). So the compromise was to design a signal that required only a few tubes, even if that meant that the analog signal used the spectrum inefficiently. What good would an efficient use of the spectrum be if no one could afford a TV set?

Then along came the transistor and solved a lot of the vacuum tube’s problems. But real progress came when transistors began to be used for digital purposes... like electronic memory.

The #1 thing that analog circuits had problems with was memory. There simply was no way to store a lot of data in analog circuits in a way that allowed rapid access and low cost. Memory makes digital TV possible. Digital TV squeezes 6 (or more) SD signals into the same 6 MHz channel used for one analog signal. Digital pictures are compared to find elements that are the same. Then only the differences from frame to frame are transmitted. But that is possible only if the pictures can be stored in memory and rapidly accessed. The more pictures that can be stored and compared, the more information that can be squeezed into a digital channel.

♣How Many Minors In A Major?♣

"If A Man In A Forest Speaks, And There Is No Woman Around To Hear Him, Is He Still Wrong?"   -Sign On Hot Dog Vendress

In our area, the local ABC affiliate is WMAR. We can enter "38.1" to get the first service on digital channel 38. Alternatively, you can enter "2-1" on your HDTV set.

How? WMAR also transmits some data called a Virtual Channel Table (VCT), which your HDTV decodes. A digital channel follows the ATSC standards, a set of rules or "protocols" for what goes where in the 6 MHz major channel.

In this case, the protocol of interest is called "The Program And System Information Protocol", which carries lots of "metadata" about each minor channel. The VCT is simply one of the tables included in this protocol; it assigns "virtual" numbers to every minor channel that the station transmits. And using virtual channels, you can "remap" channel numbers from their actual radio frequency channel (38 in this example) to ANY other number from 1 to 99. This permits a new digital channel to be associated with its previous analog channel (2 in this example), just one more way that digital TV can be flexible.

♣Bits And TV Channels And HD And Loose Ends♣

"How Come We Didn't Need Converter Boxes And Chaos When We Went To Color TV?"   -Georges Valensi

An entire 6 MHz digital broadcast channel (a major channel) can accomodate 19.39 million bits per second of compressed video, compressed audio, and ancillary data. So what can we actually "buy" with 19.39  Mbps? We could use about 18 Mbps of the maximum 19.39 Mbps to broadcast a program in High Definition using the 1080i30 format, as CBS does.

However, progressive ("p") compresses better than interlaced ("i"). If a station broadcasts its HD programs in 720p60, as ABC does, then an HD program can be compressed all the way down to 8-11 Mbps. Many digital TV stations are presently transmitting High Definition programming using just 8-11 Mbps. And the picture and the sound are good. And this leaves up to 12 Mbps in the digital channel for "other stuff".

Programs where stuff doesn't move much, like the weather forecast, do quite well with only 4 Mbps, or even less. About five compressed minor channels (five programs) was the limit for digital channel capacity back in the 1990's, when the ATSC standards ("A/53") for digital broadcasts were finalized.

BUT... With improved compression technology today (2010), you might squeeze twenty programs into one 6 MHz digital channel. This could come in handy, like when you have forty basketball teams, all of them playing at once... pick the game that you like. And it can come in handy to TV station managers for airing massive numbers of commercials.

In the Real World→ 20 minor channels multicast in a major channel and attempting to show basketball games would look worse than awful; far too much compression.

TV stations are required to provide only one minor channel to the public free; the rest of their many minor channels could be "pay-fers"... like the drug adverts you see in your doctor's office... you know, "Ask your doctor if acetylsalicylic acid could be right for you." Or, broadcasters could provide infomercials continuously 24/7.

And so, the theory goes→ Since by using MPEG-2 compression we can get many programs into one 6 MHz digital channel, then we'll need fewer channels for TV. And so with digital TV, all of our over-the-air programming needs could in theory be accomodated by just channels 2-51.

We'll see if this works in practice, because there have been no channels 52-69 for full-power TV since June 2009. (TV channels 70-83 bit the dust long ago... in 1966, the FCC stopped issuing licenses for stations above channel 69; and in 1970, the FCC took away channels 70-83 (807-890 MHz) and allocated them for the old analog cell phones, BlackBerrys, and such. ($$$) )

But there is another factor, aside from multicasting, that has lead the FCC to believe that all our TV needs can be handled with only channels 2-51.

With analog TV, you (and the FCC) didn't like to have adjacent channels broadcasting in the same area. You wouldn't want to assign both analog channels 11 and 12 to broadcast in Baltimore; they could interfere with one another. (Though with modern analog transmitters and TV sets, they wouldn't have.)

♣Why Convert From Analog Steaks To Digital Steaks In June 2009?♣

"If you only have a hammer, you tend to see every problem as a nail."   -Abraham Maslow

It's now time that we asked the question "Why Now?" Why convert from analog to digital in June 2009?

The "electromagnetic spectrum", the place where everything that is broadcast over-the-air lives, has been divided up by the FCC ever since 1934. Perhaps in 50 or 100 years we won't have need for broadcasting using the electromatic spectrum. But before the FCC came along in 1934, there was the Federal Radio Commission from 1927-1934. (And back between 1922-1927, when commercial AM radio first began spreading wildly to towns across the US, anyone who wanted a station license could just have one.)

But since 1934, the FCC has acted like a traffic cop, directing different kinds of broadcasts into different lanes of the spectrum. So, for example, the spectrum from 88 MHz to 108 MHz was (after a couple of false starts) allocated by the FCC to all the commercial and educational FM radio stations in the US.

And AM radio, after 1922, was given 550 to 1700 KHz. (KHz = thousands of cycles per second, MHz = millions of cycles per second.) If some radio station broadcasts at 1.0 MHz (one million cycles per second), the radio waves coming from its broadcast tower(s) "cycle" from plus to minus and back to plus one million times per second. And 1.0 MHz is then called that station's frequency... how "frequently" the radio waves coming from the station go through a complete plus to minus to plus cycle... a complete sine wave, if you remember your high school trig. And it is because different broadcasts have different frequencies that we can "tune" in just the station we want and ignore all the others.

In this way, one type of broadcast doesn't collide with another (in theory). And one station doesn't collide with another. However, the FCC has to approve who goes where. Like we can't have cell phone towers transmitting on the same part of the spectrum where we have TV channels.

NOW... When the frequencies for commercial TV channels were assigned around 1946, broadcast transmitters and TV receivers were more limited than they are today (think vacuum tubes and discrete components, no chips), and there was no cable (and definitely no satellites). So TV channels 2-6 got 54-88 MHz; and channels 7-13 got 174-216 MHz; and in 1952, channels 14-83 (UHF) got 470-890 MHz.

SO... Because of what was put where in the early days of TV (in 1946 for channels 2-13 and in 1952 for channels 14-83), the analog TV stations used big, valuable chunks of the electromagnetic spectrum for each program that they broadcast. Analog TV channels 2-83 used a lot of spectrum space; because each analog channel needed six million cycles per second (6 MHz) to broadcast just one program.

And the 1946 analog design, even with "compatible" analog color added in 1953 to analog black and white TV, and analog stereo sound squeezed into many 6 MHz channels in 1984, still used 492 MHz of prime real estate for channels 2-83, according to our calculator.

The conversion to 100% full-power digital TV has freed up fourteen TV channels for more cell phones (and BlackBerrys and such); and companies like AT&T/Cingular and Verizon will be able to sell more subscriptions and make more $$$. By eliminating all TV channels above channel 51, there now is 84 MHz of additional broadcast spectrum available for cell phones and similar stuff.

So present TV channels 52-62 and 65-67 were (partially) auctioned to "communications companies" for wonderful "new services." ("New Services" is the FCC's terminology.)

In addition, the conversion to digital TV, by eliminating all TV channels above channel 51, freed up four tv channels (24 MHz) that will go to police, fire, FEMA, and other emergency rescue services. (Old TV channels 63, 64, 68, and 69 have converted from use as TV channels to public safety use = 4 × 6 MHz = 24 MHz. And eventually, probably in several years, public safety will be using these frequencies.)

SO... beginning 24 January 2008, the FCC started auctioning away 14 UHF TV channels for more cell phones (and similar devices), and it will allocate 4 UHF TV channels for public safety broadcasting. ¿Está claro? Four channels for public safety, 14 channels for "commercial purposes".

Just To Summarize Who Got/Gets What→

• 14 channels where TV channels 70-83 were→ For cell phones, many years ago.

• 4 channels located where TV channels 63, 64, 68, 69 presently broadcast→ For public safety communications such as police, fire, and emergency rescue.

• 14 channels where TV channels 52-62 and 65-67 presently broadcast→ Auctioned off to telephone companies for exciting new devices like iTwitter.

So, The Telephone Companies will get the huge prime Peter Luger steaks, dry aged. "Public service" will get just a tiny Applebee's chuck steak. But even so, notice how they always seemed to mention "public service" first, as the reason for going digital-only; "new telephone services" comes second, if it's mentioned at all.

For Example→ “Well, basically, they want to give those frequencies to first responders and other emergency organizations, and also make them available for some new emerging wireless technologies,” states WTAP General Manager Roger Sheppard. (WTAP serves Parkersburg, West Virginia and Marietta, Ohio.)

"But the neat thing about it is, it also allows for television stations to broadcast in high definition or offer more than one program at a time by multicasting." Multicasting... we think of it as giving every TV station multiple free broadcast licenses. And of course, the more programs that a TV station multicasts, the worse each program looks. (See "No Free Lunch" Theorem.)

But why NOW, and why not just the four channels needed for public safety purposes?... We have heard no satisfactory justification. Yes, digital permits HD; yes, digital permits multicasts (which we desperately need, with only hundreds of virtual channels on digital cable); yes, digital may provide new interactive services (or not); yes, digital removes some forms of static (and creates exotic new ones); no, no more broadcast TV in your RV or tour bus when it's moving, with the ATSC digital system that the FCC approved (the converter box won't help); yes, auctioning off the "700 MHz" will raise a few dollars for Congress, in spite of the FCC diluting the value of the auctioned spectrum by adding strange and exotic new rules.

(UPDATE→ The problem with the original ATSC standard is its physical layer could not handle moving receivers. A new compatible standard has been developed by the ATSC (ATSC-M/H = ATSC Mobile/ Hand-Held) and can be deployed now (2010). ATSC-M/H requires new broadcast equipement (a new "Exciter" that is backward-compatible with the existing 8-VSB exciter, about $200-$300,000 per TV transmitter) and new "in-motion" receivers.)

But this entire "conversion" to digital-only seemed to hinge on propaganda that's intent on making everyone believe that the switch from analog to digital television reception is better than the advent of color in 1953. (It's not even vaguely close.) And etc, etc. But cutting some folks off from TV (and some were cut off from TV for a long while after June 2009), sounded like a risky proposition, at a time when emergency communication via TV could well have become critical in a moment's notice.

Some Comments On Digital Reception→

• “I could only pick up about five channels, eight at most. And they were never the same ones. Twist the antenna one way and get ABC and NBC. Turn it another way and get CBS and Fox. I couldn’t get any PBS stations at all, which were the real reason I wanted to get a better signal in the first place.”

• “A digital signal is affected by practically everything – where your TV set is located in your house, the walls in your house, the number of trees in your yard, how close it is to other electronic devices, birds migrating south in the fall. No kidding. A Washington Post story described how a woman who lived on the 20th floor of an apartment building would lose her signal for a few moments every time a plane landed or took off from Reagan National airport.”

• “With my converter box the picture is crystal clear - when I can get it. I tried it for a while and the whole family agreed - snow is better than the picture dropping out completely every few minutes. I tried positioning an antenna all over the outside of my house but with no better luck than the indoor one. We had some success with the front door open and the antenna balanced and pointed in a particular direction but that did not seem like a good long term solution.”

♣Some Digital And HD Secrets♣

"I Am Inclined To Think We Are All Ghosts- Every One Of Us."   -Henrik Ibsen

By using digital TV to broadcast a program in High Definition, TV programs now can display a sharper picture, and more and more accurate colors. But note that digital alone does not make HD. You must be receiving a broadcast in one of the top six steakhouse formats; and in the US, it will be just two out of the top six→ 720p60 or 1080i30; and note that a wide screen (16:9) HDTV set alone does not give you HD. It may simply look better, but it's not a sufficient condition for HD.

You need a digital transmission in order to have HD, but digital alone does not give you HD. just as USDA prime steaks alone don't give you Smith & Wollensky or Peter Luger. So what do you need to watch a program in High Definition?

1. If you are a cable customer, you must subscribe to digital cable, or your provider must be 100% digital.

2. You must be tuned to a program that is being broadcast in HD, whether you are watching the program over-the-air or via cable or via satellite or via FiOS.

3. If you are watching programming via cable or satellite or FiOS, you must subscribe to HD. in which case, you will have a digital decoder box that is programmed by the provider to receive (not block) HD.

AGAIN→ ATSC digital provides for wide screen broadcasts. But these broadcasts may be or may not be High Definition (HD). HD is almost always wide screen (16:9); but wide screen (16:9) may be HD, or it may be SD (Standard Definition, 480i30). ATSC digital can handle HD wide screen and SD wide screen. (Or even non-wide screen (4:3) SD or HD digital.)

For The Purists→ We have said repeatedly that HD is one of the six (out of 18) ATSC video formats with the highest resolutions. And these top six video formats are all defined with a width to height (aspect) ratio of 16:9. And 99% of the time, the aspect ratio is 16:9. But there are some programs that are transmitted in HD (1080i30), and with an HD bit-rate, but with an aspect ratio of 4:3; e.g., "The Wire" on HBO HD, for example. How can this be, dog Wolf?

Ok... 16:9 is the preferred format for HD, but a 4:3 screen is not necessarily standard definition (SD). HD really refers to material that is sent to an HD encoder (in the MPEG-2 phase) before transmission. Several NBC shows are shot in a 4:3 aspect ratio, but then are then scanned in as a high definition source. This provides HD picture quality, and it is considered HD. On the other hand, sending an SD 4:3 signal into an HD encoder and "line doubling" will produce a 16:9 picture with the same bit rate as HD, but this is not considered an HD signal. Whatever.

Now... Why do we need digital for HD? Because digital TV is compressed down to the point where it uses less of the spectrum for each program broadcast, compared to analog TV. HD is broadcast digitally in the US, because digital television requires less bandwidth, if it employs enough video compression. But don't overcook that digital steak. Too much compression will give the picture either...

1. A washed-out appearance, or...

2. Moderate to severe tiny tiles of digital noise (pixelation), or...

3. Non-existense (blank screen).

If HD in the US were broadcast in analog format, as HD has been in Japan for quite some time, (yes, it is possible, works fine, takes 12-20 MHz, comes from satellite... no, HD does not require digital), it would gobble up huge additional chunks of communications spectrum space for every program broadcast... or we'd be tying pairs of 6 MHz TV channels together for a single HD program. And then we'd not have all that bandwidth to auction for billions of dollars where channels 52 to 69 now live.

Of course, it's up to the TV station whether or not they want to broadcast great looking programs, or instead broadcast not so great looking programs (Applebee's steaks instead of Peter Luger's); but then the station can broadcast more than one program per digital channel (multicasting); this may = more $.

Maybe the station manager will use some of her digital bits to advertise drugs in doctors' offices; or maybe she'll have a program that constantly gives you the latest local weather (low bandwidth requirements here). Whatever matches her sense of giving to the community vs the community's sense of giving dollars to her station. (Among broadcasters, multicasting is often referred to as distributing one's "bit budget".)

So it is up to the broadcaster whether she wants to use the whole 19.39 million bits per second for one single program and provide the maximum possible quality picture; or whether she wants to send out one "main" channel plus three "sub-channels" at 5 million bits per second each... less quality but four different programs. (Maybe it changes with the time of day; maybe she saves her best picture and sound for the evening hours.

The Real World→ Digital requires more bandwidth than analog. But digital can be compressed more than analog can; digital can be compressed so much that we can do multicasting, even though it secretly takes up a lot more spectrum than analog (were it not compressed). And with Nyquist filtering technology, we can crowbar a digital channel into 6 MHz, just like analog.

♣STEAKHOUSE  FUBARS  AND  GOVERNMENT  WOOL♣

"50% of what we tell you is wrong; but we don't know which 50% it is."
-Yale School of Medicine To Incoming Students & US Govt On Digital TV

“Those Who Sow The Wind They Shall Reap The Whirlwind” 1 July 2009→ It sounded so very simple. Just buy a subsidized converter box, plug it in, and sit back and enjoy the wonderful world of digital. At least that’s what we were told. But for some New Jersey folks, the 12 June conversion to digital-only has been a huge headache, a frustrating exercise of contradictions, of speculation and fuzzy or non-existent reception. “I have four TV's; at any time, I get reception in one room, but not the others,” says Rosanne Hurley of Paramus, NJ. “We were promised improved service, more stations, and clearer reception,” Elisabeth Salfelder of Fair Lawn, NJ, stated. “Instead, we have fewer channels and are frustrated and disappointed with what is available to us. We should have known better than to believe all the hype. For example, I was watching Channel 5 in the back room and when I went to cook and watch the TV in the kitchen, it said 'no signal'.” For many like Salfelder relying on antennas, the digital revolution has meant glitches leading to TV reception that's fair to non-existent, depending on the channel, the time of day, and the phase of the moon. It seems clear that the government and TV stations, which did a great job preparing the public for the conversion to digital, did a miserable job of telling us what it would take to get reception. Meanwhile, on the left coast, San Francisco Bay Area stations KTVU, KPIX, KBCW and KTFK (Sacramento) have asked the FCC for permission to build new antennas to extend their digital signals. Oakland-based KTVU (channel 2), wants to broadcast on channel 48 from Monument Peak, east of Milpitas, thus reaching the South Bay. KPIX (channel 5) wants to broadcast on channel 42 from Mount Vaca near Napa; and KBCW (channel 44) wants to broadcast from the same location, but on channel 31. KTFK, a Spanish language station in the Central Valley, wants to put an antenna atop Mount Diablo in the East Bay. How weak digital has turned out to be. 2 July→ There was a time when truck driver James Loeffler could rely on the television signals from far-off Denver to get news about Colorado. But the dawn of digital-only television has left Loeffler, a longtime resident of Stratton in Colorado's flat northeast, taking more stock of the doings in bordering Kansas than in his home state. “I don't know what's going on in Colorado anymore, but I do know what's happening in Colby, Kan.,” the 61-year-old quipped. “And they said this was an improvement?” When digital-only broadcasts became the Law on 12 June, over-the-air television feeds from Denver went dark in Stratton, the result of analog translators in Stratton that couldn't transmit the new signals. The only thing residents have now — save for those who subscribe to cable or satellite TV — are the analog broadcasts from Rocky Mountain PBS and from Colby to the east. And PBS will shut off its analog signal 12 July. Nationally, experts believe that as many as two of every five translators went dark at the transition. That's because the aged mechanisms were from the bygone analog era and needed updating, either by adapting them to convert the new digital signals back to analog or revamping them entirely to receive and transmit the digital signal intact. Kit Carson County, where Stratton is located, paid to maintain its translators and refused to update them, citing the expense. Then the man who knew how to fix them died shortly after the digital switchover. “It's pretty unbelievable that this is all we've got now,” Loeffler said. “It really stinks.” 3 July→ The FCC is working with dozens of TV stations that are still difficult to receive by antenna since they switched to new frequencies as part of the digital-only transition, the government said yesterday. Most of these stations, in cities like San Francisco, Philadelphia, New York, Miami, and Dallas, moved their digital broadcasts from UHF to VHF on 12 June. The VHF band (2-13) previously had been used primarily for analog; VHF was largely untried for digital. While some of the UHF band (14-69) often can be received using indoor antennas, VHF works better with the large rooftop units that were ubiquitous in the early 1950's. The problem→ Many antennas that were sold as "digital" were in fact UHF-only. (There are no "digital" antennas per se; any VHF/UHF antenna that worked with analog will work with digital; this is not to say it will work successfully, however.) The FCC has sent engineers to affected cities, and it has granted temporary permission to some stations to increase the strength of their signals as it hunts for a long-term solution, according to the FCC's Robert Ratcliffe. ABC affiliate Channel 6 in Philadelphia lost viewers after the transition, and it has received permission to boost its power output temporarily; and the ABC station in Chicago also has problems. Just after 12 June, about 30% of callers to the FCC's help line complained that they could not receive one or more digital stations. That figure has declined in the most recent week to just above 20%, the FCC said. 5 July→ Port Huron is a city in southeast Michigan. When Thelma DeVoogd of Port Huron turned on her TV back on 12 June, she was excited. The 76-year-old had done more than merely survive the switch from analog to digital TV; she did so with a crystal-clear picture and crisp sound. But a couple of hours later, her TV screen was blank. DeVoogd simply lives too far from most television stations to pick up the new signal, and there's not much she can do about it. “I am frustrated,” she said. “We've had television for over 50 years. It wasn't always great, but there was something I could watch. Now, there's nothing.” DeVoogd is far from alone in her TV troubles. Despite buying a digital converter box and following instructions issued by the FCC, many people who live in rural areas such as "Michigan's Thumb", were left out in the cold when technology "advanced" last month. The problem is simple→ 8-VSB Digital television signals don't carry as far as analog signals, creating dead spots that didn't exist before the switch. But fixing the problem is a bear. Just ask Gerald Weber, 60, of Forestville, Michegan. Weber used to use an antenna to tune in to Bay City and Saginaw stations, which he depended on for local news and weather warnings; but after the switch, he's been unable to get those stations. Weber said he has tried just about everything possible, including writing letters to officials. “Nobody cares. Politicians don't care. And the FCC lies about it,” Weber said. “I don't know where else to go. If there was someplace else I could go, I'd go.” Edie Herman, a spokeswoman for the FCC, acknowledged the problem; but she said there's nothing the agency can do. While officials spent months preaching the message about the importance of converter boxes, the agency didn't publicly announce that some people would be left in dead zones. The FCC did create a Web site where people could check which stations they could receive in digital (with weak 8-VSB modulation). “With analog, you could still get a fuzzy picture, but with digital you hit a cliff; it's really good until it stops,” Herman said. (DAwn had a great-uncle with a pacemaker just like that.) Herman said that people can try using a large outdoor antenna to draw in programming; or they can try moving it around to catch what they can, but there are no guarantees. (If you look at a Constellation Diagram (we will in a bit), you'll see why the 8 possible symbols that come from the TV transmitter 10,670,000 times each second have to be pretty close to 8 vertical lines at sample time for 8-VSB digital modulation to work. It's a fragile modulation system. But then, they knew that.) Matt Tamme, director of engineering at WNEM (channel 5) in Saginaw, said that homes within 40 miles of the station will pick up a digital signal. While that's not a large change from the analog coverage, there are contributing factors that hurt the digital signal, he said. “There are a lot of low areas with trees around that really cause problems,” Tamme said. (Trees? Trees?) While people might have been able to fiddle with a tuner before and catch what was coming through, albeit fuzzy, digital signals don't work that way — it's either there or it's not. (Or you can have pixelation, which changes the picture to tiny colored squares and kills the sound and is probably worse than no picture and no sound.) Tamme said one thing people can try is rescaning their digital converter box (sigh... please) because some stations (almost all stations) changed channels with the transition. Also, he said, splitting a signal between TVs can reduce reception. Something else people should check, he said, is whether their antenna picks up both UHF and VHF signals. David Munsell, senior engineer for WEYI (Ch. 25) and CW 46 (WBSF) out of Saginaw, said the broadcasting area for the station changed just 2% during the transition. Munsell said people having trouble in the Michegan Thumb might be able to pick up CW on 46-1 and NBC on 46-2. But, he added→ “It can be a real challenge for people on fringe areas. At some point, the signal just drops off.” Thelma DeVoogd knows that all too well. She can pick up a signal only from about 6 AM to 9 AM. Subscribing to cable or satellite really isn't worth the money, she said, because she typically watches only “the weather in the morning and nothing else.” Her solution for additional programming→ “I have to watch Canadian television.” Canada won't switch to 8-VSB digital television until 31 August 2011. 6 July→ Weak over-the-air reception wasn't the only problem for 6ABC (WPVI) in Philadelphia. 6ABC, which has the region's top-rated news, also lost its popular radio position at 87.7 FM. “You don't know how much I miss it. Who are they to come in here and take that out of my car?” a distraught David Farina of Ardmore, PA, said. Farina said he drove about 200 miles a day around the Philadelphia area and listened to 87.7 FM... until it went off the air for good on 12 June. “I want it back,” he said. Because 6ABC broadcast on channel 6, folks could tune in to 6ABC's audio at 87.7 FM. This was possible because 6ABC, like any analog channel 6, broadcast on a frequency adjacent to FM radio stations. But when the TV world went digital-only on 12 June, FM radios could no longer receive signals from 6ABC. (Digital TV does not use FM to broadcast its audio.) “We are looking at some options of how people might hear us, but 87.7 is probably gone for good,” said Caroline Welch, a spokeswoman for 6ABC (owned and operated by the Walt Disney Company). She could not estimate how many people listened to 6ABC on their radios, but she noted that “it was a pretty large number.” The Philadelphia region has about three million households with TVs. One problem is that if 6ABC were to simulcast its content on TV and FM radio intentionally, the Philadelphia TV station would have to secure new intellectual-property rights. Under the pre-12 June broadcast situation, people could hear the 6ABC programming "by accident". So 6ABC did not have to acquire intellectual-property rights to be heard on radio. But the station's bigger problem with the digital transition is that many thousands of people lost their TV reception because 8-VSB digital signals behave differently from the old analog signals. (The FCC has granted 6ABC a temporary approval to quadruple its TV signal to improve reception.) 7 July→ Cable, satellite, and Fiber-to-The-Premises (mainly FiOS) will add about 655,000 new subscribers because of the transition to digital-only, according to Wachovia analysts, who had expected the increase to reach about 900,000. 8 July→ Chicago TV stations spent months preparing viewers for the transition to digital-only on 12 June. In spite of that effort, thousands of Chicago area residents have been unable to get signals from CBS-owned WBBM (Channel 2) and ABC-owned WLS (Channel 7) since the transition. “WLS in Chicago is probably one of the three or four stations that is mentioned most in calls about signal problems,” said FCC spokesman Bill Lake. The continuing problems may be linked to the power of WLS' digital signal, which is relatively low. The FCC is letting WLS increase its power (on a trial basis) to see if that helps. WLS General Manager Emily Barr said that her station continues to hear from a few households still having problems because they live in buildings constructed out of materials that may prevent the digital signal from fully reaching their TV set. Calls to the FCC help line now are down to about 10,000 to 12,000 a day, about 2% of them from Chicago, Lake said. But Radio Shack stores report that they continue to get calls from people who say they can't get Channels 2, 5, 7 and 66. Our Opinion→ The best solution would be a rooftop antenna, a signal amplifier, and a quality converter box. This one-time investment would be cheaper than 3 months of cable or satellite service. 9 July→ When analog vanished a month ago for Carolyn Lork and Walter Jagans, so did Cincinnati's TV stations. Lork, who lives 55 miles east of downtown, and Jagans, who lives 40 miles west, each have an outdoor antenna that received analog Channels 5, 9, 12 and 19. These two folks are the most extreme cases of people losing channels since the 12 June switch to digital-only TV. “I’m really disgusted. We did everything they told us to do, and it didn’t work,” says Lork, who has a converter box. For years, Lork relied on Cincinnati TV for news, weather, entertainment, and information about mall sales. Now she only can watch Portsmouth’s PBS station and a CBS affiliate in Charleston, W.Va. “We get the Charleston weather, which doesn’t do us any good, because the storms have already passed,” says Lork, a retired nurse. She’s another victim of digital TV’s “cliff effect.” Unlike analog, with digital there is no “ghosting” or snow. Either viewers get a perfect picture, pixelation without sound, or nothing at all. But not all digital problems are in distant counties. Viewers throughout Greater Cincinnati and Northern Kentucky say that some Cincinnati channels also have fallen off the digital cliff, randomly. Phil Bart no longer gets Channels 5 and 9 with an outdoor antenna aimed at Cincinnati. Dayton TV comes in better, even though Dayton is father away. “We had good reception on analog. Switching to digital should not be this frustrating,” says Bart, an engineer. Judy McCuinn can’t watch Channels 9, 14 and 19 with an amplified indoor antenna. Margaret Hombs has lost Channel 12 on her rooftop antenna. “We’re on high ground. We should have great reception. It doesn’t make any sense,” Hombs says. Sharon Heckort, who has a rooftop antenna, finally ordered DirecTV, after the FCC's call center told her she was too far away for Cincinnati TV reception. A month after analog ended, TV stations in Cincinnati continue to receive complaints from upset viewers, where 14.5% (129,095 homes) rely on antenna reception. (The US average is 10.5 percent.) Channels 9 and 12 have received the most complaints, because they’re VHF. Digital VHF signals (Channels 2-13) do not penetrate buildings, walls, or trees as well as UHF (Channels 14-51), says Kurt Thelen, Channel 12 chief engineer. Unlike analog, digital signals can be blocked by hills, people walking by, or multiple reflections off metal or moisture (everything from cars and aluminum siding to wet leaves). Similar VHF stations nationwide are having the same problems, says Lynn Claudy, National Association of Broadcasters vice-president for science and technology. Channel 9 (WCPO) has heard hundreds of complaints since last year about its weak signal. Bill Fee, general manager, says he hopes most problems will be fixed when Channel 9 finally broadcasts at full power from a new antenna due to be fully installed shortly. Reception complaints are new to Channel 12 (WKRC), which switched its digital from UHF Channel 31 to VHF Channel 12 on 12 June. Thelen, the chief engineer at channel 12, regrets the decision to move digital to channel 12. “I hate to say this is a step backward, but I won’t say it’s a step forward. This is a massive frustration for broadcasters, too,” Thelen says. “People don’t like paying to get ‘free’ TV, but the one-time expense of a rooftop antenna is a pretty good investment,” Lynn Claudy says. FCC spokesman Mike Lake says stations have several options to reach lost viewers→ Increasing power, changing channels, or constructing “repeater” stations to fill gaps. But each option takes time and money. And increasing power to reach outlying counties could overpower TVs close to the tower, Thelen of Channel 12 says. So what should disgruntled viewers do? Broadcasters and the FCC preach patience. “It took us 60+ years to get analog TV right, and we’re just starting with digital. I hope that in the next months and years, there will be technological advances,” the FCC's Lake says. (No, do NOT put tin foil on your rabbit-ears antenna.) Channel 12’s Thelen is “trying to come up with something. I wish it could be fixed overnight, but it won't.” “I see this going on for the rest of the year (2009). We’ve got to solve this problem one viewer at a time,” Channel 9's Fee says. Top Secret→ Meanwhile, according to "George Q", “I still clearly remember when the 8-VSB modulation method was selected as the digital transmission system; and at that time, it was well-known that the digital signal would cover only 80% of the analog 'grade B contour' (the area that gets a decent, viewable signal). Outlying viewers were doomed from the start.” 14 July→ Before 12 June, Chicago area resident Roger Grubb would turn on his HDTV and marvel over the picture quality of his over-the-air channels, especially WLS (Channel 7). “It was spectacular,” said Grubb, whose six-story apartment building has a master antenna on the roof. “Of all the channels I got, it was Channel 7 where you would drool over it. And then, all of a sudden, it was gone.” Since that Friday, when broadcasters across the country switched from analog to digital-only signals, Grubb hasn't been getting WLS. And neither have pockets of other viewers in the Chicago area. The FCC and WLS are still working to resolve the the pockets of nothingness. “This is one of the few stations in the country that we're looking at most closely,” says FCC spokesman Bill Lake. For those viewers who have done everything right to get digital signals, the FCC believes the problems may center on signal power. The FCC has dispatched additional staff to Chicago, as well as Philadelphia and New York, shortly after the digital transition. The agency granted WLS permission to experiment with power levels for two weeks. That testing period ended last week, and now engineers will analyze the results. “The hope, obviously, would be that increased power would relieve the reception problems and wouldn't cause interference” with neighboring stations, said Lake. “But if that's not the case, we and the station need to continue to look for solutions.” Chicago viewers need antennas that can handle both VHF and UHF, one reason some viewers may not be getting all the new stations in VHF digital. In the transition, WLS moved its digital signal from Channel 52 to Channel 7. Emily Barr, president and general manager at WLS, said early testing showed that signals would cover the same geographical area as before. “What we were unable to replicate was what happens if you're in a stone building, a stone house or a high-rise and using an indoor antenna,” Barr said. VHF frequencies (Channels 2-13) can have difficulty getting through windows, as well as brick and stone, when they are digital, modulated by the 8-VBS technique; everyone knew this. That's part of the reason Chicago and other metropolitan areas with tall buildings continue to have lingering problems with the 8-VSB digital signals. And unlike in the analog era, when a weak signal might just result in a snowy picture, a digital picture appears crystal clear or not at all. (Or a garbage pixelated pixture with the sound cut off.) Barr said the trouble spots appear isolated (discontinuous), and that WLS viewership, as indicated by ratings, has stayed steady. (You win some, you lose some.) Marlene Peterson, a self-professed "news junkie," has switched from watching "ABC World News" to "NBC Nightly News" because the WLS signal is inconsistent. “Sometimes it comes in beautifully, and other times it doesn't come in at all,” said Peterson, who has an antenna on the roof of her ranch home. “For example, I just turned it on now, and it's fine. Last night we wanted to watch the news and we didn't get it at all.” The FCC and WLS are continuing to recommend that consumers look into outdoor antennas. For those who aren't able to do so, “we've told them to check back with us because we're interested in resolving this for absolutely everybody,” Barr said. “We know it's not a significant number of people, but, frankly, even one person is significant.” Hint→ If one person was having problems, the FCC would not have sent in an engineering team. An Engineering Analysis→ Impulse noise on Channels 2-13 has been an issue in analog for some time, particularly for channels 2-6. Chanels 2 and 5, for example, are at harmonics of citizen band frequencies. Have someone in your neighborhood using an illegal linear amp or a mistuned (deliberately or otherwise) CB transmitter, and you'd end up with various impairments in audio, video or both. But, people either ignored them, or got accustomed to them. There was much discussion and little debate about a decade ago in RF circles about how lowered power levels would make low band VHF particularly problematic. The PSIP standard permitted broadcasters to keep their existing channel numbering (virtual channels) regardless of what channel they ended up on (real channel). The FCC also made those on VHF subject to interference, Yet, some seemily smart people in dense regions thought they would violate the laws of physics. Not all those remaining on Low band VHF have issues, due to regional development patterns. But, actions have consequences. Unfortunately, Julius Genachowski, like all FCC commissioners to date -- as non-engineers -- don't have a clue as to physical issues. They approach such issues as attorneys, which is to say, badly. If they had the ability to do math, they would have become doctors or engineers, not attorneys. An Engineer Speaks→ I've been working as a TV engineer since HDTV was being talked about as an analog system. The FCC WANTED to make the new digital TV system a UHF-only system, but big groups like MSTV (the Association for Maximum Service Television, Inc.), backed by the big networks, which owned mostly VHF stations, pressured the FCC to allow digital TV on VHF. Now they're paying for it. 8-VSB digital modulation simply doesn't work well on channels 2-13; it doesn't penetrate solid objects as well as many UHF channels. 26 July→ Long-time video engineer Mark de Leon Martinez writes to us→ “For a person that is 'off the cliff' of a digital broadcast, there ARE possible solutions→ If you're using an indoor antenna, put up an external antenna. If you're using an external antenna, invest in a more directional antenna, a higher gain antenna, or an antenna tuned for the specific channel you are receiving (with a tuned combining network where applicable). The possible use of a high quality RF amplifier also may help in some marginal cases, where the tuner really needs only a little more signal.” 29 July→ 98.9% of US households now are able to receive digital TV. Since 12 June, 1.3 million homes went from unprepared to prepared, according to figures from Nielsen released today. Only 1.2 million homes (1.1%) still are unable to receive digital signals. That is down from 2.5% unprepared just a few days before the 12 June transition. Most unready sets are located in spare bedrooms, kitchens, basements, offices, and garages. 54% of unready homes earn less than $25,000 per year, and 29% earn between $25,000 and $50,000. Some unready homes may not be completely without TV, watching low power stations or stations broadcasting from Mexico or Canada, all still analog. 59% of completely unready homes receive at least one low power or foreign station. On average, homes that receive low power stations have 3.3 stations available to them. As of mid-July, 60.7% of completely unready homes have no Internet access. Households headed by older adults are the most prepared, with less than 1% in that age group completely unready. Among homes with the head of household under 35, 2.7% are completely unready. There remain a number of markets that lag behind the general population, such Albuquerque-Santa Fe, Las Vegas, San Antonio, San Diego, and Dallas-Ft. Worth. 30 July→ More than 10,000 WWBT viewers (Richmond, Channel 12) complained of a sudden inability to receive the station’s digital signal, particularly through antennas located indoors, since the switch to digital-only last month. This morning, the FCC authorized an increase to the local NBC affiliate’s digital power from 6,000 to 26,000 watts. Viewers of other stations that switched to channels 7-13 (high VHF) for post-transition operation likewise have reported reception problems. 31 July→ What Exactly Is The Problem With VHF Digital→ Most digital TV stations now are broadcasting in the UHF band (channels 14-51), primarily because of problems with impulse noise on low-VHF (channels 2-6), and to some extent on high-VHF (channels 7-13). While virtual numbering schemes that ATSC digital permits display virtual channel numbers such as 2.1 or 6.2 for the sake of branding, these channels are most often really UHF broadcasts. (For example, in this area "WMAR ABC2" is not really broadcasting on channel 2, but rather on channel 38.) NOW... ATSC digital television uses forward error correction techniques; a channel is assumed to have a random bit error rate, and so additional data bits are transmitted to allow these errors to be corrected at the receiver. While this error correction works quite well for UHF signals, where interference consists primarily of "white noise" (continuous electromagnetic waves of a wide range of frequencies), forward error correction is inadequate on lower VHF channels (2-6), and to some extent on upper VHF channels (7-13), where bursts of impulse noise disrupt the entire digital channel for short lengths of time. A short impulse of noise was just a small annoyance with analog broadcasts; due to the fixed timing and repetitive nature of analog video synchronization, short impulse noise usually was recoverable. But the same interference is severe enough to prevent reliable reception of the more fragile and highly-compressed ATSC digital signal. Power limits also are lower on VHF, especially on low-VHF; but a digital UHF station may be licensed to transmit up to perhaps a million watts of effective radiated power (ERP). Thus very few stations returned to VHF channels 2-6 after the June digital transition. At least three quarters of all full-power digital broadcasts now are using UHF. And in some US markets, such as Syracuse, New York, no digital stations returned to VHF after the June digital transition. Also, UHF penetrates walls and obstacles better than VHF does (which is why the UHF frequencies were selected in the first place, and why channels 52-69 brought (and will bring) so much $$$ from phone companies in the FCC auctions). Analog could deal with less VHF penetration, but digital cannot. And finally, the error-correction technology in the ATSC signal works better at UHF frequencies than in the VHF range. (UHF transmissions also work better in the “concrete canyons” of a major city.) It was strongly suggested that ALL digital TV be broadcast on UHF, but the powers that were said "no, we want to maintain our VHF stations". And they did. A great many video engineers knew that ATSC 8-VSB digital would have serious problems on VHF; the current problems are no great surprise to many. 11 August→ “I agree with [Name] in White Plains ("Digital conversion isn't working for everyone," August 1 letter). Just what is the FCC going to do about this mess it created? The only channel I get is NBC Channel 4. As far as I'm concerned, we better go back to the way we were. Refund us our money; this whole digital idea stinks. Converter boxes do not work, I can't get through on the hotline; no one is there. I can't write what I really feel, you would never print my letter. “If I didn't have my TV in my living room, I wouldn't have a thing to watch. I like to watch the news while having breakfast in my kitchen. But that's gone, just like my $40 for the converter box. Both breakfast and money "gone with the wind." I even spent money on a new HD antenna. All for nothing. What did I get out of this? Nothing but aggravation. They got the money and I got the shaft. “So come on FCC, do the right thing, either fix the situation or let us go back to the way we were.” 19 August→ The Problem→ The problem with digital television, telecommunication engineering experts say, is the technology's "all-or-nothing" shortcomings. Unlike analog signals, which "degrade gracefully," resulting in an imperfect but viewable picture, digital signals cut out totally when not enough data is being transmitted. The result is either a pictureless blank screen that scrolls the words "No signal", or an image that is either frozen or broken up into jigsaw-like fragments (pixelation). As a result, digital signals are much more susceptible to interference than analog. Digital signals can be disrupted by a host of obstacles, including terrain, other buildings, and the amount of dense foliage between a TV station's transmission tower and a home antenna. Bad weather also can break up a signal, and pinpointing specific signal gaps is difficult. Summary→ Digital just doesn't work. 9 September→ We had hoped to get more channels, but it's not just meant to be in this area. We have the Winegard HD 8200U antenna, 40 ft. tower, Channel Master CM 7777 Titan Antenna Preamplifier and Channel Master CM 9521A Antenna Rotator. We used Belden 1829AC Single RG-6 Coax Cable and still can't get anything but KXII (Channel 12, Texoma (Texas and Oklahoma) ) and KTEN (Channel 26, Texoma (Texas And Oklahoma) ). I contacted WFAA in Dallas, and Don Guemmer, who is the Transmitter Supervisor/ Chief Operator responded after entering my address: Not gonna work. You guys have even more dirt between Cedar Hill and your house than the other guy up there. Attached is a plot of the terrain vs distance to your house. That red line is the terrain. Your antenna mast will be nearly 300’ tall to make it work. The Dallas stations towers are 80 miles from us. 10 September→ The latest attempt by NBC affiliate WLBZ, Bangor, Maine, to get its digital TV signal to reach more viewers has met with uncertain results. WLBZ switched its digital signal from UHF Channel 25 to VHF Channel 2, effective 11 September, a move that the Gannett station officials hoped would improve conditions for a number of viewers using antennas. Since the switchover from analog to digital signals on 12 June,“we have received calls from viewers who used to get our analog signal but cannot receive the digital one,” Judy Horan, WLBZ president and general manager, explained in late August when the change was first announced. “By going back to VHF Channel 2, we’re confident that some of those over-the-air viewers will be able to watch us again. We can’t make a promise to everyone who lost us, but it’s an improvement.” The change requires all over-the-air viewers to perform a rescan on their digital television sets or converter boxes in order to receive the new channel. According to the July Nielsen ratings, about 18%, (26,000 households) in eastern and central Maine, receive their local channels over the air. Horan estimated that her station has heard from about 2% of over-the-air viewers. Horan said it was impossible to quantify how many people the latest change has helped. “There’s no way really to gauge it,” she said. “We’re going to end up with a net gain in viewers. But we’re hearing from those with problems now.” Much of the blame goes to the science of broadcast signals and the Federal Communications Commission, which mandated that all TV stations broadcast exclusively digitally starting last 12 June. According to the WLBZ Web site, over-the-air viewers who have been watching WLBZ using indoor antennas, such as “rabbit ears,” could have difficulty tuning in the VHF frequency. The VHF — or very high frequency — signal should improve reception for a number of viewers using outside antennas. VHF is much more forgiving when it comes to terrain because it tends to hug the earth better than UHF, making it possible for viewers who lost reception after the 12 June conversion to tune in the station. The same flexibility can be problematic as the VHF signal enters a building. Moving an indoor antenna toward an exterior window may help, but ultimately the best solution is to use a rooftop antenna. “The FCC has always based signal reception standards on a rooftop or exterior antenna, not indoor models,” Horan said on the Web site. “That’s frustrating to the viewer who didn’t want to invest in a rooftop antenna or is prevented from doing so at a rental property.” Viewers who opt to install a rooftop antenna should make sure that the model they select is capable of receiving “low” VHF signals, which are channels 2 through 6. Conclusion→ Digital wasn't ready for prime. 11 September→ The FCC has approved the move of CBS affiliate station KTVT in Fort Worth, Texas, from channel 11 to channel 19. In an order released Friday (9/11), the FCC said it was in the public interest to allow the move, which also requires KTXA (Independent) there to move from its real digital channel 19 to real channel 29. The FCC has been flexible in letting stations strike deals among themselves, trying to resolve the problems of lost viewership after the 12 June switch to digital-only. The request for the double switch was filed jointly by KTVT and the owners of KTXA. CBS had told the FCC that KTVT had lost 57% of its over-the-air audience after the switch, and that getting viewers to rescan their converter boxes for its new post-transition digital channel had not helped reclaim many of those viewers. The commission said it was in the public interest to allow the switch and to make it effective immediately on publication in the Federal Register. The FCC has had to work with two to three dozen stations that had major issues with signal loss after the transition, granting temporary authority for channel moves and power boosts, as well as some permanent moves like the one approved Friday (9/11). 12 September→ Three months after the 12 June digital transition, more than 99 percent of homes in the United States are able to receive digital TV signals, according to a report released this week by the Nielsen Company. The report found that as of 30 August, only 0.6 percent of homes in the U.S. -- about 710,000 households -- were unable to receive digital signals. When the transition took place in June, that number was at 2.2 percent. 18 September 2009→ Think the digital TV transition is over? Not quite. Many viewers have found that they can't pick up certain stations after the switch, even with the right TVs or converter boxes. The stations are still trying to figure out ways to help them tune in. The main problem is that when the last major stations turned off their analog TV on 12 June to broadcast entirely in digital, some of them moved their digital signals from the UHF frequency band (channels 14 to 69) to VHF (channels 2 to 13). To most viewers, these channels are just different numbers on the remote. But as signals in the airwaves, they have very different characteristics. VHF hadn't been used much for digital signals, and there were indications that there would be problems with the switch, partly because viewers had inadequate indoor antennas. Still, the switch went ahead. Since then, at least 20 VHF stations have asked the Federal Communications Commission to move their digital signals back to UHF, and more would like to do so. However, the government has sold off some of the UHF band to cell phone carriers, leaving less space for TV channels. Another portion is planned to be used for emergency services, which was another reason for the digital TV transition. Philadelphia's ABC affiliate, WPVI, switched its digital signal to channel 6 on June 12, and got thousands of calls per day from viewers who couldn't find the station on their sets any more. Within a week, WPVI got emergency permission from the FCC to quadruple its transmission power. It could do that because the closest station that also uses channel 6, in Binghamton, N.Y., also wanted to increase its power, which meant it wouldn't be overwhelmed by the stronger signal from Philly. But in other cases, increasing power is a complicated proposition involving several stations. More than 50 VHF stations have applied to increase their signal power. The power increase helped WVPI punch through to a lot of viewers, but the station still gets calls every day. Hank Volpe, director of engineering at WVPI, says he understands the loss of the station's UHF slot, “but I would have loved to have a UHF channel to play with.” Mark Colombo, a TV enthusiast and electrical engineering student who maintains an online database of the country's TV stations, said “"everyone who had any sense“ knew that broadcasting digitally on channel 6 or lower would yield terrible reception. Those channels are susceptible to interference from household electronics, spark plugs in passing cars and distant thunderstorms. What was more surprising was that channels 7 to 13 also had problems, though there had been clues it would happen there, too. WVUE in New Orleans, a Fox station, turned off its analog signal last December, before most other stations, and moved its digital signal to channel 8. The reaction was immediate. “We fielded thousands of phone calls,” said Al Domescik, WVUE's director of engineering. “We did everything we could. We talked to people on the phone. We sent technicians out to people's houses. We brought antennas to people's houses. We just kept beating our heads against the wall for months.” In June, the station started simulcasting on UHF, which mollified most viewers. WVUE's experience was repeated more than six months later, when Chicago's ABC station, WLS, tried to move its digital signal to channel 7. It says it got nearly 7,000 calls from viewers about reception problems in the week after the transition. Nearly half of the homes visited by the FCC in WLS's service area in late June had inadequate indoor reception. WLS tried doubling its output power, but it wasn't enough. Now the FCC is letting it move to UHF channel 44. TV consultant Peter Putman said a lot of reception problems for digital VHF channels can be attributed to the fact that VHF antennas need to be large. The long rods on an outdoor antenna are for VHF reception, and it's difficult to make a compact indoor antenna with good VHF performance. TV watchers with indoor antennas had the same problem with VHF stations when they were analog, but often suffered through it. They would get a poor, snowy picture and decent sound, and considered that good enough. But because digital is an "all-or-nothing" technology, the weak signal they get on digital isn't enough to produce a picture at all. Some TV viewers simply have the wrong antennas. For years, "HDTV" antennas were sold that brought in only UHF. Andy Couch, a Web developer in Austin, Texas, installed one in his attic and was happy with it until this summer, when the local Fox station, KTBC, disappeared from his set. It had moved its digital signal from UHF to VHF. “Now I have to get a VHF antenna for just one channel? No thanks,” he said. Another problem is that FM radio stations can interfere with VHF TV channels. Volpe at WPVI in Philadelphia said FM interference is easily dealt with by installing an "FM filter" or "FM trap" on antennas. Analog TV manufacturers incorporated such filters in their sets. However, digital TVs and converter boxes lack these filters, since they do nothing to improve digital reception in UHF, where digital signals mainly had been until this year's transition. FCC spokeswoman Janice Wise noted that relatively few stations out of the more than 1,800 in the country have reported reception problems after the transition, and said the agency is working closely with them to resolve their issues. “People are figuring there's someone out there to blame for this,” Volpe said. “Well, there's nobody to blame.”" But the nature of the DTV transition — with nearly all major-city stations turning off on the same day as mandated by Congress — didn't make it easier to identify and deal with reception issues. Colombo, the TV enthusiast, points out that in Wilmington, N.C., where the FCC encouraged TV stations to shut down last September as a test for the big day, all the digital TV stations used UHF. The area also lacks large hills that can block signals. “It was basically the ideal market, he said. “You could not ask for an easier market to deal with than Wilmington.” 4 January 2010→ “Is digital TV a scam? We get only one station, WPXQ, a local station. We purchased converter boxes and even a booster – nothing happens. We live in Foster, RI, and can’t get satellite because there are too many trees. Cable won’t run in our town, and Verizon stopped installing Fios three miles down our road. We haven’t had television since the digital-only transition last June. Wrote to my local representative, but no luck. Don’t know what to do. The government and local politicians don’t care – they have their TV”. In Our Opinion→ “First, US digital standards are needlessly complex by reason that they had to be translated to English from the original weak Swahili 8-VSB modulation. Second, the Digital TV experiment was not yet ready for prime time. It was in development for over 10 years, and it just does not work in the real world. And third, the carriers and sync pulses that made it so easy for analog receivers to grab onto a weak channel are gone with analog, replaced with only the weak Fisher-Price phony "pilot". [“Ah, that is clearly a metaphysical speculation, and like most metaphysical speculations, has very little relation to the actual facts of real life as we know them.” -Oscar Wilde, The Importance of Being Earnest.] And to be able to auction channels 52-69 (less the four channels to be used for "safety" some day), the FCC had to cut way back on transmitter power to squeeze everyone into channels 2-51 (channel 7 was a special favorite). Why? Because there is very little dead space in a 6 MHz digital channel. Digital uses a lot more space than analog; there is huge compression (MPEG-2) and filtering (Nyquist technique) to get digital into 6 MHz; this is one reason that 8 MHz British digital looks So much better.” But what a boon for US cable and satellite companies, most of whom receive local TV stations over expensive fiber optic cables instead of wobbly over-the-air broadcasts. (Wells Fargo Securities broadcasting and cable analyst Marci Ryvicker estimates that the digital transition added about 655,000 new subscribers to cable, satellite, and FiOS, short of the 900,000 that had been predicted... like, there's a number that's easy to estimate... the reason that folks signed up with cable, etc. (Last great estimate→ Estimated deaths from run-of-the-mill influenza in a run-of-the-mill flu season = 36,000; actual death certificates stating "influenza" as the cause of death in a run-of-the-mill flu season = 839. So much for statistical models.) Our estimate of folks subscribing to cable, satellite, and FiOS out of fear of the digital-only fiasco, corrected by the "influenza" coefficient→ total = 15,265.)

• America's transition to over-the-air digital TV, which has already netted the government $19 billion in its botched first wireless spectrum auction, was doomed from the start, thanks to a flawed voucher program and a time frame that left the country stranded between administrations, with folks climbing their icy roofs in subzero windchills to install new antennas in mid-February.

• In computing the post-transition transmitter power for every full-power station, the FCC apparently assumed that every home in the viewing area had an efficient VHF/UHF rooftop antenna. As a result, many transmitters had insufficient power to reach all of their viewing area after the transition. (But what a boon to cable and satellite.)

• In the midst of the transition, the Obama administration, rightly, was coming in and saying, "Look, we're going to have to deal with this giant transition that we've been handed in the first twenty days of our administration, and we're looking at potentially tens of millions of people losing access to their television and to emergency news and information, and that's not a situation we want to find ourselves in".

• The former FCC never went out in the field to determine that digital actually worked, or that converter boxes actually worked. As a result, to take but one example, the folks in rural New Jersey lost their New York and Philadelphia TV stations in Digital. The former FCC also appeared oblivious to the pixelation problem and cliff problem that are endemic to ATSC digital.

• Worried about offending the TV broadcasters, the FCC never developed any standards for High Definition broadcasting. TV is big business→
  2006→ Revenues From TV Broadcasting + Cable + TV Advertising + TV Set Sales = $182,000,000,000  (Source, Time, June 22, 2009).

• Congress low-balled the number of consumers affected, budgeting only $1.34 billion for the $40 coupons — less than half of what the Consumers Union, the non-profit publisher of Consumer Reports, warned was needed.

Many of those who rely on rabbit ears are elderly, disabled, or low-income. For them, TV isn't a luxury; it's their only connection to the outside world and to emergency information. Congress either could have cut them off on 17 February or delayed the majority of the switch.

Fortunately, it appears that only a handful of public safety programs were ready to jump to their new spectrum right away, so the postponement didn't hold up life-saving communications. Delay turned out to be the best option. Seven million households shouldn't have had to pay for their government's ineptitude.

AND NOW AT LAST
WE FINALLY REMOVE THE BACKS
FROM THE CLOCKS OF TELEVISION

(Was Mankind Really Supposed To Know This Stuff?)

♣NTSC - BLACK AND WHITE AND COLOR♣

"All Things Change, But Nothing Dies."   -Ovid

Why worry about analog when US TV became digital in June 2009? Answer→ Because there are billions and billions of analog TVs connected to billions and billions of converter boxes and cable boxes and FiOS boxes and satellite boxes, receiving Canadian and Mexican broadcasts, receiving billions and billions of low-power analog transmissions from now until the 12th of Never. And besides, ATSC digital makes infinitely more sense when we look at NTSC analog TV first; and so, let's.

Let's start with a look back at just how NTSC color was squeezed into NTSC black and white in 1953; because you see, there never was a "black and white" digital TV... color is an integral, inherent part of digital TV. In analog TV, color crawls in and out of the modulation process; in digital, once we're finished with MPEG-2 in the Video Subsystem, where compression and packetization are performed, we're done worrying about color; we never even give it another thought when we get to phase two of digital, "The Exciter"... where modulation and transmission take place. There is no color carrier, no color modulation, in Digital.

The new ATSC digital standards fix LOTS of problems that have plagued analog TV over its 60 years... including "compatible" color broadcasts added to black and white "with a crowbar" in December 1953→

The colors like "yellow" and "red" and "green" would shift when the ANALOG NTSC TV signal was broadcast over-the-air in the early days. Why? There were changes in the chrominance, carried in the analog TV broadcast signal... some say the colors in early analog TV changed with the phase of the moon. But there is no color carrier in ATSC digital TV, no separate color information, nothing depending on lunar phase.

ATSC digital color is very different from analog NTSC color. Every pixel in High Definition TV is attached to three 8-bit numbers (not signals, numbers), one for brightness, and two for color. Yes, some of the color numbers are thrown away during MPEG-2 compression in the Video Subsystem (a process called "sub-sampling"); but those numbers that remain are virtually unchangable... carved in stone... as long as our digital receivers sample the digital broadcast at exactly the proper instant.

But even ATSC digital relies upon the French engineer Georges Valensi's 1938 patent for transmitting separate luminance and chrominance. And thus, to understand ATSC digital (and why the color works so well), it's helpful first to understand NTSC analog color (and why it never has worked all that well).

And so we're now going to chat a bit about just how analog NTSC color works, how it has worked for the last 55 years; and then we're going to contrast it with the simplicity and elegance of the new ATSC digital... digital that's replete with Trellis Coders and Solomon-Reed Encoders and Nyquist filters. (Hey, it's 2010.)

First let's state for the record that analog NTSC color TV is actually "Very Simple Stuff" (VSS); yeah, TV gurus like to tweak our minds with analog color just a little... make it seem complexicated. And some TV engineers will swear on their "flyback transformers" that NTSC stands for "Never Twice the Same Color".

But as we know, NTSC is the set of standards for ANALOG TV. It was developed for black and white (B&W) TV, carved in stone (approved by the FCC) in May 1941; and then about twelve years later, the NTSC standards were modified to include ANALOG COLOR TV and re-carved in stone. But analog color TV had to be simple stuff, in order to fit in with the B&W analog TV that was already starting to run wild by 1953... and vacuum tubes were a limiting resource... like, it was October 1953 when Arthur Godfrey (who?) fired Julius La Rosa on the air... live... but sadly, only in black and white.

And then 17 December 1953 came. The US FCC reversed its 1951 decision (approving the CBS non-compatible color system) and approved the RCA/NTSC color system. NBC broadcast the NBC chimes image at 5:31:17 p.m. using NTSC color standards. But CBS broadcast the first live color program at 6:15 p.m.; NBC followed with a live color program at 6:30 p.m. And on 1 January 1954, NBC broadcast the Rose Parade in color on 21 stations. These were surely exciting times in which to have lived. (Remember this stuff when next you play "Trivial Pursuit".)

The tricky part of coming up with NTSC analog color TV was figuring out a system that would be compatible with existing black and white (B&W) TV sets... Compatible? Yeah... B&W TVs had to display their normal monochrome pictures, even when they were tuned to broadcasts in color; and color TVs had to display a monochrome picture during B&W broadcasts. But once the puzzle pieces were jammed together back in late 1953... well, it's now a piece of cake to understand what the folks did... fitting analog color into analog B&W in a compatible fashion was in fact quite brilliant. (Except for changing the frame rate unnecessarily.)

DEEP DARK SECRET→ The compatible reception of black and white NTSC now has been long gone and irrelevant for many, many years... color NTSC is used even to broadcast occasional monochrome material today, like some great B&W movie from back in the early forties.

(NOW... If only some genius had devised a compatible system of digital TV that could be received on analog sets.)

And hey, by April 1956, WNBQ in Chicago had replaced all of its black-and-white equipment with color equipment, becoming the first TV station to broadcast all of its local programming in color. (Perhaps the US Congress should have simply decreed that after April 1956, there would be no more black and white broadcasts; and the FCC could have given every station TWO channels, one for B&W and one for color; and the government could subsidize color converter boxes; and... no, that would be ridiculous.)

♣THE TRUTH OF TV TRANSMISSION♣

"You Are, When All Is Done... Just What You Are."   -Goethe

AIN'T NO BLATs→ Before we dig further into NTSC analog TV, we need to clarify a point that few folks understand (perhaps 0.1%). A very common misconception is that, in analog TV, the entire picture (field) plus the sound gradually fill up the 6 MHz channel, like water filling a bathtub.

Then, when the TV camera at the TV station has scanned a complete image, and the 6 MHz channel "somewhere out there" has filled up, then BLAT... the TV transmitter wakes up from its snooze and broadcasts the whole thing to our TV sets, at the speed of light. And then the 6 MHz channel slowly starts to refill with the next picture, until there is another BLAT. And we have 60 (or maybe 30) BLATs per second in NTSC analog television, and the TV transmitter simply sleeps most of the time between the BLATs, and... and... NO.

NO. NO. NO.

NO. That's not what happens... except for the transmission traveling from the TV antenna atop the tower to our TV sets at nearly the speed of light... that fragment was accurate. But the rest? The rest was pure 100% Flubbitz-Doodle.

OK... Here is what really happens→ (We'll skip the sound; analog TV sound is easily understood, and it's irrelevant to the issue.) At any instant in time, the antenna on the TV tower is busily transmitting, very much like an AM radio transmitter does. But instead of an announcer shouting an annoying commercial, the TV transmitter is transmitting the brightness and color of the point that is being scanned in the TV camera at that instant.

When the design for black and white analog TV was created in the first half of the last century, the technology to store and later display a complete frame simply did not exist. Thus, images in NTSC analog TV are transmitted using "line scanning" techniques. As an image is being scanned by the TV camera, it is encoded, transmitted, received, decoded, and scanned onto the picture tube in the TV set.

In analog TV, the horizontal lines are not broken up in the studio into dots or into "pixels". As the light-sensitive image is scanned line by line by an electron beam behind the image, a signal is sent out from the station's transmitter as a continuous, varying current. The scan lines in the camera are recreated, line for line, on home TV sets tuned to this station. Eventually, in 1/60th of a second, the TV camera completes its scan of odd (or even) lines behind its video image; and the home TV receiver has created a near duplicate of the image in the TV camera.

At any instant, the brightness of the point that the TV camera is scanning is reflected by the strength of the signal coming off of the station's transmitting antenna; and a similar brightness is created on screens of home receivers by a rapidly moving electron beam that creates a point of light where it touches the TV receiver's screen. The screen is covered with phosphors; when the electron beam hits a point on the phosphors, that phosphor glows for an instant, often brightly.

When the TV camera scans a bright spot on its image, the signal to the transmitter drops (yeah, it works upside down; but there's a good reason why a 12% signal strength is white and a 75% is black), the signal to the antenna drops, the signal to your set drops. But the set is smart, however; and it knows that a low voltage means "Paint wherever we may be... Brightly."

When the TV camera scans a darker spot on its image, the TV antenna transmits a higher signal. This way, the beam in your TV set painting on the picture tube's phosphors is in "sync" with what is being scanned inside the station's TV camera. How do we keep the TV camera and the TV set in sync? We literally transmit the same "sync pulses" to both. (We're assuming that you have an analog TV set with a picture tube because most TVs today still use picture tubes (also called "CRTs" (Cathode Ray Tubes) ) ).

The TV station also transmits other amazing things to your TV receiver... things that say "Shut up", and "Move down to the next line"; or "Shut up and move back to the upper left" (because a brand new field is about to begin). TV engineers call the "Shut Up" commands blanking pulses; "blanking pulses" turn off your set's electron beam(s), so that you don't see lines on your screen when the electron beam is repositioning itself for drawing the next line or the next field. Blanking pulses simply turn off the set's electron beam(s) for an instant.

SO... At any given instant, the TV transmitter might be sending out brightness and color information about the point in the image that the TV camera is currently scanning; or it might be sending out a "blanking pulse" to briefly shut off our TV set's electron beam(s). Or it might be sending a "sync pulse" to keep the beam(s) in our TV receiver in sync with the beam(s) in the station's camera; but it's definitely sending out something. TV transmitters are almost always busy sending out something. (There are absolutely no BLATs.)

So tell us, dog Wolf, why then do we need that big 6 MHz channel? Simple... because the TV camera rapidly changes what it's sending out. It has to get out all of the picture information for 525 lines to your analog TV screen, plus all the sync and blanking pulses, in 1/30^th of a second. Remember, analog NTSC TV runs only at 30 complete frames per second. (And yeah, you need some room for the audio (sound) carrier and its "sidebands".)

525 lines x 30/sec = 15,750 lines/sec. And when you use AM modulation (ok, technically "VSB" modulation, a type of AM) to "modulate" the TV station's video (picture) "carrier" at that speed, information flows out from the video carrier wave to the left and right on the frequency spectrum, instantly filling about 4 MHz of the 6 MHz that's available. The stuff that instantly ends up right and left from the video carrier is called the "sidebands"; all modulation of electromagnetic waves, any kind of modulation, produces sidebands... just not always the same kind of sidebands... but sidebands.

So, the energy that was in the video carrier flows out to the left and the right of the carrier; the carrier gets smaller (has less energy); and the energy that the carrier loses flows out to create the four MHz or so of sidebands... because the actual video information lives in the sidebands; the carrier is really there as an energy resevoir, ready to flow left and right and to create the sidebands.

Note that NTSC analog TV in the US transmits only at 480i30. Only that one format is ever transmitted in NTSC analog; and that's the only video format that the analog NTSC TV set ever expects to receive. This makes analog TV sets simpler (and cheaper).

SO... Bear in mind that the antenna on the TV tower is transmitting to homes (and sometimes to cable and satellite providers, though a fiber-optic link is more common today between the TV studio and cable/satellite providers) instantaneously whatever point the TV camera is scanning (or whatever the videotape is playing, or whatever the film is showing, etc). Most of the time, that 6 MHz, and the TV transmitter, and the TV tower's antenna are busy transmitting what is happening at the station (or at the network studio) at that instant. Analog NTSC TV stations do not send one frame (or one field) at a time. They send one point at at a time.

Note carefully→ ATSC Digital is very different. What is being broadcast at any instant in ATSC digital is not what is being scanned in the camera at that instant. At any instant, ATSC digital is sending tiny, tiny pieces from across a few milli-seconds of camera scanning; we'll explain all the how's and why's of digital transmission shortly... but like, it's the 21^st century; all things, good and bad, must change.

And now... you're among the 0.1% of "TV gurus" who understand the actual workings of NTSC analog.

How DO They Do The Sound?→ Alright, I hear curious minds out there asking, "Dog Wolf, how DO they do the sound?" Simple→ At an NTSC analog TV station, the sound goes into a perfectly normal FM transmitter, except that the output of the transmitter uses only 1/3 of the space on the electromagnetic spectrum (50 KHz) as does an FM radio transmitter (150 KHz); this reduces the "dynamic range" of analog TV's sound (the difference between the softest and loudest sounds). But in ATSC digital, Dolby gives us the true dynamic range (and then some). You can adjust the dynamic range on most digital TVs to be more or less dramatic.

The FM sound (FM carrier and FM sidebands and all) is inserted into the upper part of the 6 MHz channel, above the video carrier and video sidebands.

In NTSC analog, the same antenna at the top of the same tower transmits both the video and audio. A gismo called a diplexer provides the necessary isolation between the picture and sound transmitters. The sound and the picture signals are "combined" in the diplexer, and they go from the diplexer directly to the shared antenna. The FM transmitter often has to reduce its power, however, so that the FM sound covers the same geographic area as the picture. In any case, the FM sound transmitter cannot exceed 22% of the peak radiated power of the video carrier... "cannot" because the FCC says so.

Tidbit→ If you tune your FM radio receiver to 87.75 (just below 88.0), you may hear the sound from channel 6, if channel 6 is active in your area; channel 6 is allocated 82 MHz to 88 MHz. Listening to sound from channel 6 without watching the picture can be theraputic. But that therapy ended on transition day; once we went to digital-only, there no longer was any FM sound on channel 6 (or any other channel, for that matter). Digital TV does not use FM (nor a separate transmitter) for its Dolby sound.

---

♣THE SIGNALS OF NTSC TV♣

"Never Does Nature Say One Thing And Wisdom Another."   -Juvenal

Slightly More Detail→ The camera "tube" contains a photosensitive area, or "target", composed of hundreds of thousands of light-sensitive isolated elements, upon which an optical lens system focuses the scene to be broadcast into an image. An electron scanning beam moves over this surface (from the front or from the back). This electron beam releases an electrical video signal with a voltage depending on how bright each point is on the camera tube.

In NTSC analog, the light image in the studio TV camera is scanned by electron beams moving horizontally from left to right across the camera's tube; and the same image is re-created in the home TV sets by electron beams moving horizontally from left to right.

Simultaneously with these horizontal scans of the electron beams, the beams are also slowly working their way down to the bottoms of the camera tube and to the bottom of the TV set's screen; and upon reaching the bottom, the beams are then sent flying back to the top again. First the odd lines are scanned (and painted in the receiver) horizontally, then the even lines are painted on the TV set's screen horizontally; it's a process called "interlacing", developed to reduce flicker in our visual systems. (Flicker is not a technology or engineering thing, it's a biologocal thing. We use tricks like interlacing to reduce flicker in our visual systems.)

Note that the motion of the beams, both in the camera and the TV set, is similar to the way we read a page of text; left to right, left to right, slowly working our way down the page.

Each scan of the image, from top to bottom, produces a "field" and involves half of the total horizontal lines (262½). Two complete scans are required to accumulate the 525 lines for a complete picture, called a "frame." 60 fields are scanned each second (30 frames per second); and 60 fields are "painted" on the picture tube in the receiver each second.

Continuing on our journey through the in's and out's of NTSC analog TV, we recall what we said above→ Under the NTSC standards, 30 complete pictures (frames) are transmitted to viewers each second. But they are not transmitted as complete frames (or even complete fields). Instead, each image is, in effect, broken down into several hundred thousand points on a light sensitive target in an NTSC TV camera, and these elements then are transmitted in sequential order, one point at a time, as the TV camera methodically scans its image, methodically scans each point.

The TV camera in the studio focuses on an image, and the image is focused on three light-sensitive targets in the color TV camera. Each target is made up of hundreds of thousands of light sensitive points. A separate electron beam scans each of the three light sensitive targets (red, green, and blue), and an electrical signal proportional to the brightness of each point being scanned is produced, as the electron scanning beam touches each light sensitive point. A decrease in brightness (luminance) at any point causes an increase in radiated power from the broadcast antenna... this is termed negative transmission... the brighter the point, the lower the transmitted signal current.

Thus 12% of the maximum transmitted signal strength means we are scanning the brightest white, and that point ultimately becomes a bright white (for an instant) on our TV set's picture tube. Conversely, 75% of the maximum transmitted signal strength translates into a black (for an instant) on our TV set's picture tube.

There also is another (sneaky) way that we can produce black on our picture tube, besides scanning a black point on an image. We can briefly shut off the electron beams in the picture tube in our TV set. If we transmit a blanking pulse which is always transmitted at 75% of maximum signal strength; or if we go for broke and transmit a sync pulse (synchronization pulse) which is always transmitted at 100% of maximum signal strength, we will produce black. Anytime that we transmit at between 75% to 100% of maximum, we shut down the electron beam(s) in our receiver for as long as we transmit in this 75-100% range. This can be useful if things are happening that we don't want displayed on the TV set... like the beams flying from lower right to upper left, to begin a new field.

So you can see now why we use negative transmission→ Because the various sync pulses all are transmitted at maximum power (100%="super-black"), and we don't want to see them; and by making both black and blanking pulses equal to (75%="black"), we don't see them either.

There also is a stability advantage in making the sync pulses 100%; they are easier for the TV receiver to find, to locate, to grab onto, even if the actual video signal from the TV station is noisy and snowy and weak. So most TV systems (France being a notable exception) have sync pulses that are as high as possible, usually 100%.

So NTSC analog TV uses 12.5% of the maximum peak video carrier to represent white. And it uses 75% and up to represent black (or to cut off the electron beams, same thing).

Forward To Digital For A Moment→ ATSC digital has no sync pulses sticking up at 100%, nothing there for the digital TV set to "grab onto". And so before transmitting the digital signal from the broadcast antenna, we insert a "fake" carrier called the pilot at 310 KHz above the start of the 6 MHz channel... just something for the digital TV receiver to "grab onto" when things get rough... like when we change the channel. (A diagram of the digital 6 MHz with the pliot is coming up. Stay tuned.)

Back To NTSC Analog TV→ The three signals, derived from the brightness of the three light-sensitive targets in the TV camera, are used to control the three electron beams in our TV receiver; the brighter a point is on a target, the more intense that the control grids in our receiver's picture tube will make the electron beams hitting the red, green, and blue phosphors at that one point, and the brighter that point will be painted on the TV screen (for an instant).

Stated another way, in a color TV camera, there are three targets, one sensitive to red, one to green, and one to blue. As these three targets are scanned by three electron beams, three electron beams in the viewer's home TV are also scanning, drawing the same image, point by point, that the scanning beams find in the color TV camera. The electron beams in both the TV camera and in our home TV sets move methodically from upper left to lower right, just the way we read a page in a book. The original brightness and color of every point in the images in the TV camera are recovered in the color TV receiver and painted on its phosphor screen, point by point.

---

NOW→ For this to work, the beams in the color NTSC TV set must be synchronized exactly with the beams scanning along in the NTSC color camera. At the TV station, a piece of hardware called the sync generator creates the sync pulses that→

1. Drive the three electron beams in the color TV camera.

2. Are transmitted to our homes to drive the beams in all TV sets that are tuned to this station.

The sync pulses are sent over a cable from the sync generator at the TV station directly to the camera's circuits that control its scanning beam(s); and identical sync pulses are inserted into the TV signal broadcast from the station's transmitter. This was very clever stuff in 1941. Here is how we scan and blank out the retraces→

Simplified TV Scanning

---

Ok, let's take a slightly more detailed peek at just what we have been sending out over-the-air from NTSC analog TV stations since 1953 (the beginning of color TV)→

1. A video signal, based on the brightness and color of each point in the image in the TV camera, at any given instant. And recall, the brighter the point, the lower the signal we transmit; 12% modulation is usually maximum brightness; and 75% modulation (and above) is black.

2. "Rectangular" sync pulses. Rectangular pulses means that if we graph the strength of the pulse against time, the pulse looks like a rectangle; it rises very rapidly from 75% to 100%, then stays at a constant voltage at 100% for a few microseconds, and then it moves back to 75% very rapidly. (A picture showing sync pulses follows shortly.)

Why does analog NTSC use "sync" pulses? Because when analog NTSC television was developed, no technology existed for storing the video signals; and so, the signal had to be generated and transmitted at exactly the point in time at which it was displayed on the TV set's picture tube. And so, it was essential to keep the scanning in the TV camera in exact synchronization ("sync") with the scanning in the television set.

OK→ so just what are some of these NTSC analog "sync pulses" and "blanking pulses" that the TV station transmits to keep everything in "sync" and to "blank" stuff we don't want to see?→

• The Horizontal Sync Pulse (also sometimes called the horizontal drive pulse) synchronizes the horizontal motion (left to right) of the beams in the TV camera and the beams in the receiving TV set. This is analagous to reading across a line of text in a book.

• The Horizontal Blanking Pulse shuts off the beam during the "horizontal retrace interval" (when the beam is flying from the right end of the line that we just finished leftward, to the beginning of next line). This is analagous to moving our eyes to the next line on a page of text that we're reading. It prevents extraneous "retrace lines" from appearing on the TV screen. After all, we don't read anything when going from the end of one line to the start of the next line; similarly, we don't want to see the electron beam in our sets doing this, so we "blank" the beam, we shut it off for about 10 microseconds until it's ready to start "drawing" the next line.

  NOTE→ This "horizontal sync pulse" is transmitted at 100%. And it sits in the middle of the "horizontal blanking pulse, which is sent at 75%.

---

THE  HORIZONTAL SYNC PULSE  AND THE HORIZONTAL BLANKING PULSE

-------100%------ "Back Porch" | | | |Horizontal Sync| ↓ 75%________| Pulse |___________75% / \ / \ / <-------Horizontal Blanking Pulse--------> \ / 10µs \ ~~~~~End Of Line........ .........Start of Next Line~~~~~

---

• The Vertical Synchronizing Pulse moves the scanning beams in the camera, and the picture painting beams in the TV receiver, from the top of the image to the bottom of the image... just as when we read lines of text on a page, we slowly work our way down the page from the top line to the bottom line.

• The Vertical Blanking Pulse shuts off the beam in the camera and in the TV set when we have finished the bottom line of the image (finished with this field), and we're moving from the right end of the bottom line up to the beginning of the top line of the next field... just as when we're reading, and we move from the lower right of a page to the upper left of the next page... we don't want to read anything as our eyes make this trip; and we don't want to see anything on our TV sets when we're going from lower right back to the upper left for the next field. No spurious lines. So we blank the electron beam during this time.

• The Equalizing Pulse maintains a "rigid interlace", so that lines on the picture tube screen are equally spaced; this assures that the even lines are "drawn" precisely midway between the odd lines, from the top to the bottom of the screen. Same for the TV camera. It is this pulse that makes interlacing possible. (There is no analogy here to reading lines of text, since we don't read the odd lines first and then the even lines next; at least, most of us don't... with DAwn, anything is possible.)

• The Color Sync Burst most often called The Colorburst sits on the "back porch" of the horizontal blanking pulse. At the end of each line of video on the screen, the screen goes dark (the electron beam shuts off) for about 10 µs; after we're into this blanking pulse for about 1.65 µs, a 4.65 µs horizontal sync pulse comes along, points the electron beam at the beginning of the next line, and starts it moving left to right again.

  But before the ten µs of the "horizontal blanking pulse" expires, after the "horizontal sync pulse" has ended, at about 7 µs into the blanking pulse, eight or more cycles of the Color Sync Burst are inserted by the TV station at the beginning of the new line to "sync up" the color oscillator in the TV set. The TV station sends this colorburst to keep any frequency error in the set's oscillator to a minimum, in case the oscillator is out of tolerance. The color burst lasts about 2.5 µs, and then we reach the end of the "horizontal blanking pulse"; ah, we see video again. (See diagram below.)

  The "color burst" allows us to re-create the color carrier in our TV receiver, so that we can extract the color information that was broadcast without a carrier. (A diagram to show just where the colorburst lives at the beginning of each line of video is below.) The color sync burst is a "fleeting" reference signal that pops up at the start of each line of video. This color burst is an unmodulated sample of the 3.58 MHz color carrier that was deleted in the TV transmitter before it ever could be broadcast.

NOTE→ In the early days of color TV, no colorburst was sent from the TV station when the program was in B&W. This feature was used to "kill the color" on color sets; when there was no colorburst, the color amplifiers shut down. This was a cool thing back then; it prevented any colored snow from appearing during black and white broadcasts.

While it takes the electron beam 63.5 µs (a minor eternity today) to go from left to right across the picture tube, including the 10 microsecond blanking pulse that darkens the screen, and the sync pulse sitting on top of it that moves the beam to the start of the next line, the colorburst chews up just 2.5 µs at the beginning of each line, out of 63.5 µs for the whole line... very cool back in 1953.

Note→ Since ATSC digital has no color carrier and never did, it needs no "colorburst" reference signal to recreate the color carrier in the digital TV set. In fact, ATSC digital doesn't need any of the sync or blanking pulses that seem to be running wild in the Old Wild West of NTSC analog color TV.

Colorburst On "Back Porch" Of Blanking Pulse (75%) After Horizontal Sync Pulse (100% Maximum Signal)

The diagram above also shows three values, H, L, and S; ignore them. We are primarily interested in showing you where the colorburst lives, at the start of every line of video... we come to the end of a line on the TV screen, a "line blanking pulse" is transmitted, which turns off the electron beams in the TV set; and then the "horizontal sync pulse" repositions the turned off beam back to the left side of the tube, and the beam begins moving left to right; but first, WHAM... there is the colorburst, which the receiver uses to reconstruct the 3.58 MHz color carrier that the TV station deleted on the way to the video transmitter, in order to prevent interference between too many carriers... video, color, and sound.

So just what is happening in the diagram again? Once more, a little slower, please, dog Wolf→

1. OK... follow along in the diagram above... We come to end of a line in the camera, and to the end of a line in our analog TV screen.

2. A 10 microsecond "Horizontal Blanking Pulse" is received, turns off the elctron beam in our sets, and prevents us from seeing anything on the screen for the next 10 µs. During this 10 µs "Horizontal Blanking Interval" (occasionally, called the HBI), incoming data is not displayed on the TV screen... the electron beam has been blanked to avoid displaying the "horizontal retrace line"... in other words, after displaying one line of video, the scanning beam in the picture tube (the CRT) must go dark while it travels back to the left side of the picture tube. The Horizontal Blanking Pulse itself is transmitted as black (75%).

3. Now it gets even better. After 1.65 microseconds have elapsed into the Horizontal Blanking Pulse, the 4.65 microsecond "Horizontal Sync Pulse" is received. (So why did we need that 1.65 µs of nothing (blanking) before the Horizontal Sync Pulse came? (This 1.65 µs, by the way, is called the "Front Porch".)

   Well... This 1.65 microsecond interval was inserted (in 1941) between the end of each transmitted line's actual video and the beginning edge of Horizontal Sync Pulse (the guy who takes us to the next line) to allow voltage levels to stabilize in the TV sets of the early 1940's... to prevent interference between the picture lines. Today, well... it's just there. The Horizontal Sync Pulse is transmitted at 100%, which as we know is "blacker than black", and it's a nice thing for the TV set to grab onto.

4. Now we're at 6.15 microseconds after the start of the the Horizontal Blanking Pulse, with the electron beam in the TV set now aimed at the next line; now a 3.58 MHz sine wave Color Burst Signal is added to the start of the new line, on the "Back Porch". (See it in the diagram?) The back porch is that part of the 10 µs Horizontal Blanking Signal that still remains after the Horizontal Sync Pulse has ended.

5. The "Back Porch" lasts 4.79 microseconds. When a color signal is being broadcast, the Color Burst is inserted into the "Back Porch". When a B&W signal is transmitted, there is no colorburst on the back porch. And at the end of the horizontal blanking pulse, now we're ready to start transmitting the actual video for the next line. Note that the "back porch" was just "there" when the original NTSC analog TV was approved in 1941; when color was added to TV, the "back porch" must have seemed like a cool place to put a "color burst".

SO... How many times does the TV camera scan across each sensor (red sensor, blue sensor, green sensor)? 262½ lines every 1/60^th of a second? Correct. Then the scanning returns quickly to the top and scans the "in-between" lines. (Recall, odd lines/ even lines "interlacing" to reduce flicker?) The screens on TV receivers are made to go black for all "retracing"; this "made to go black" is called "blanking". There is no apparent flicker because of the interlacing of the lines and because of the two "vertical-retrace blanking intervals" for each complete picture (frame). (Yes, the analog TV set's screen goes black 60 times per second, twice per complete picture... and we never even notice.)

NOW... In the color TV camera, a horizontal-deflection waveform causes the scanning beam to travel from left to right across a scanning line, one beam for a sensor for each of the primary colors (red, green, and blue). And as you read this line, your built-in horizontal-deflection waveform causes your eyes to travel across the curves and serifs of our characters (sounds very sexy) from left to right. At the end of each scan line in the TV camera, the "horizontal-deflection waveform" whips the electron scanning beam back to the left side of the light-sensitive surface.

And of course, there is also a vertical-deflection waveform. This waveform gradually pulls the scanning beam down the surface, top to bottom. So, when the scanning beam is turned on again, after being blanked at the end of each line during retrace, it is just a tad lower than it was when it scanned the last line. Thus a series of lines are traced rapidly from left to right.

And when the beam hits the lower right, WHAM... the beam is turned off, and the "vertical-deflection" field in the TV camera (AKA "the vertical retrace") whizzes the beam back to the upper left; 1/60th of a second has now passed, and it's time to begin scanning the 262½ lines of the next image that falls on the light sensitive surfaces... it's time to create the next field.

SUMMARY→ To reproduce the image being scanned by the camera, both the camera and TV set must be scanning the same part of the image (picture) at the same time. This synchronization applies both to the horizontal and vertical motion of the beams. At the end of each line, the beam returns to the left side of the screen, called horizontal retrace; it is controlled by the horizontal sync pulse. At the end of one field (262½ lines), the beams must be sent to the top of the screen. This is called vertical retrace and is controlled by the vertical sync pulse. When the beams are undergoing horizontal or vertical retrace, they must be turned off in the camera and receiver. This is called blanking and is controlled by sending a blanking pulse, a signal level at or above black (75%) to turn off the beams. The turned off beams move to their new positions before being signaled to turn back on. An equalizing pulse also is inserted during vertical blanking, which causes the fields to begin at the proper points to achieve interlacing. (We don't want odd lines and even lines overlapping.) And to allow extracting the color information, the station sends a colorburst (a few cycles of the color carrier) at the start of each line.

---

(Without delving into genlock (generator lock), all the cameras and video tapes and whatever at a TV station are usually synchronized from a common source; before this was done, the picture on home TVs would roll a frame whenever the source was switched between cameras.)

Starting to see a pattern? The picture painted on the phosphors of your TV screen is done much like reading a page. And usually, there is a "blanking pulse" to turn off the beam (so you don't see the retraces), and then the sync pulse comes that actually repositions the beam, before the beam is turned on again.

Good News→ Does ATSC digital TV need these blanking and synchronizing and whatever signals? Nope. (Saved the best for last. It does use→ A Pilot and two sync pulses, but for very, very different reasons, as we shall see.)

♣THE VALENSI SOLUTION♣

"Everything Looks Worse In Black And White."   -Paul Simon

NOW... Just how does NTSC color TV work? Well, it works rather well, once you get into the late 1960's and the early 1970's; it took about 15 years after NTSC compatible color was introduced in December 1953 for it to get its act completely together... for TV sets to add enough advanced circuitry to get the color accurate and stabilized.

We look at our 1992 20" Sharp analog color set, we look at the movies and live programs in color, everything today is transmitted in color, and we marvel that this NTSC analog color set actually works as well as it does. And B&W also looks fine, even though the set has never been repaired or aligned (of course, alignment is automatic).

Georges Valensi's Solution→ To add color to 1941 NTSC B&W (Black and White), engineers, one in particular, figured out a clever way to send, in effect, two additonal signals containing all the color information you need for a complete color picture, along with the sharp B&W picture. This "key" to getting a working color television system in the same 6 MHz bandwidth as B&W was patented in 1938. (Georges Valensi's solution also was the key to compatible color television.)

Instead of trying to transmit red, green, and blue (the raw signals that come out of a color TV camera)... which had always lead to techniques that were incompatible with B&W broadcasts and/or took a huge amount of bandwidth, far more than the 6 MHz that each TV channel in the US was allocated, French engineer Georges Valensi came up with the idea of transmitting a color signal as luminance (brightness or "Y") and chrominance (color information or "C")... one sharp B&W picture plus two "fuzzy" color signals, in effect.

Then, these three signals could be "massaged" in the color TV receiver to produce red, green, and blue, which would drive the three electron "guns" in a color picture tube. Mixtures of red, green and blue light on the color picture tube's screen trick the human eye into seeing most visible colors, even though only three narrow bands of color can actually be generated on color picture tubes. The trick works because the eye can only detect colors centered around yellow, green, and violet anyway... although our eyes can see more colors than an NTSC color picture tube can produce. (We can see perhaps 10 million distinct colors; think the 29 billion or so colors that a plasma HDTV set can produce is another reason that digital on a "flat screen" looks so good?)

NOTE→ Chrominance has two parts→ Hue and Saturation. These are not broadcast directly; instead two "vectors" (two AM signals) are broadcast, called I and Q. The I vector displays the colors between orange and cyan, and the Q vector displays the colors between purple and greenish-yellow. The phase of I and Q determines the hue (which color it is, like red, yellow, etc); and the amplitude of I and Q determines the saturation. (Saturation is how much white is mixed in; the less white mixed in, the greater the saturation; deep cherry is more saturated than light pink.)

Anyway, Voilà, Valensi's idea worked... full color pictures, more vivid even than color slides (eventually... the theory was sound, the practice needed a bit of tweaking).

(So remarkable was Mr. Valensi's idea (and so many years had passed between his patent and the advent of compatible color TV) that his patent received an unprecedented extension to 1971. All current color television standards... NTSC, SÉCAM, PAL and, yes, today's ATSC digital... all implement Mr. Valensi's idea of transmitting a video signal composed of separate luminance (brightness or "Y") and chrominance (color information or "C"). )

And as we said... it only took until the early 1970's for analog NTSC color TV to produce fairly accurate colors that were reasonably stable. (Although... Dr. Steve recalls that his mother won a color TV in early 1967, and that the colors were quite vivid and fairly stable.) Meanwhile, from 1953 until today (2010), black and white sets have kept their sharp B&W pictures; they simply throw out the color information, with its I and Q signals; in other words, black and white TV sets just throw away all the color information that they receive (and that is useless to them). They even receive the colorburst on the back porch and just ignore it.

Engineers call the sharp B&W picture "Y". (Y stands for luminance, how bright or dark the picture is, the amplitude of the video signal.) And by using Y and I and Q, you can derive all the color data that you need... red, green, and blue. We'll explain it all shortly.

SO REMEMBER→ Who was the father of just about all compatible color television, both analog and digital? Right, Georges Valensi.

♣WHY THE "Y"♣

"He Who Neglects To Drink Of The Spring Of Experience
Is Likely To Die Of Thirst In The Desert Of Ignorance."   -Lin Po

[Many folks are curious about why the letter "Y" is used to represent brightness or luminosity. (Many of these folks later regret having wondered about "Y".) But for the truly inquisitive, or for those of you who want to impress a color TV engineer at your next dinner party, here's the origin of the "Y"→

In 1931, the International Commission on Illumination (Commission internationale de l'éclairage, or CIE) developed a Red-Green-Blue (RGB) model of human color vision. Their experimental results were initially combined into the specifications of the "CIE RGB color space".

(What's a color space? Answer→ The human eye has receptors (called cones) for short, medium, and long wavelengths of light. Our nervous system obtains color information by comparing the responses of our three types of cones to light. Thus, three values describe a color sensation. And any method for associating these three numbers (called "tristimulus values") with a given color is called "a color space". (Dr. Steve says he knew that his semester of undergrad optics would come in handy some day; he just didn't expect that it would take forty years.) )

Now the members of this special Commission wanted to develop a mathematically defined color space, one which also related to the first color space (called "RGB"). Was this possible? Well, since the Commission correctly assumed that Grassmann's Law held, the second color space (called "XYZ") could be related to the CIE RGB Space by a linear transformation. (What is Grassmann's Law? Answer→ If you have two beams of light, and an observer sees that the first beam is made up of R1 + G1 + B1, and she sees that the second beam is made up of R2 + G2 + B2, then if these two beams are combined, the RGB values of the combined beam will be R=R1+R2, G=G1+G2, and B=B1+B2. (See? Linear.) )

The desired XYZ color space was created and named→ "The CIE XYZ Color Space for visual three-color stimulus (tristimulus)." (WHEW.) And yes, this XYZ color space was one of the first mathematically defined color spaces in the study of the human perception of color.

AND... This new XYZ color space was designed to have the following property (among several others)→

"The Ÿ(λ) color matching function ("Y-bar of Lambda") would be exactly equal to the photopic luminous efficiency function V(λ) for the 'CIE standard photopic observer' (CIE 1926). The luminance function describes the variation of perceived brightness with wavelength..."

Hence, Ÿ(λ), or simply "Y", came to represent the luminance function, or the luminance. Q.E.D. (Quite easily done.) ]

♣The 6 MHZ Swiss Cheese♣

"Beauty Is A Form Of Genius - Is Higher, Indeed, Than Genius, As It Needs No Explanation."   -Oscar Wilde

6 MHz Analog NTSC Color TV Signal

♣THE FINISH ON COLOR TV→ IQ♣

"There Is No Worse Lie Than A Truth Misunderstood By Those Who Hear It."   -William James

I and Q→ When many folks decide that they want to understand NTSC color television, nothing on Earth seems to make them as crazy as I and Q. Like, there are some really bright people who can solve Einstein's General Relativity equations in five dimensions who have given up on I and Q. But this will not do for us... because I and Q are the key to understanding analog (and, yes, digital) color TV. And as we've told you before, the more complex that they try to make things seem, the more simple they really are... so right here and now, your dog Wolf is going to demystify I and Q. Believe us→ It ain't no big thing.

First, What's I & Q?→ I and Q are the color (also called "chrominance" or "chroma") information that TV stations transmit to our TV receivers. Everything as far as which colors the cameras in the TV studio are "seeing" at the moment, or the colors from the videotape machines or film projectors... all of that information is transmitted in I and Q.

Without I and Q, you have black and white TV. Sure, color information from a TV camera begins as red and green and blue signals leaving the camera; but by the time it reaches the station's TV transmitter, the red and green and blue have become I and Q (and ok, Y... brightness). Start with Red and Green and Blue, end up with I and Q and Y.

There is no color information that is transmitted (unless you count the colorburst... and you shouldn't, because that is just a reference signal that contains no color information about the picture in the TV camera at the moment) that is not in I and Q. I and Q are everything that you need to add to a black and white NTSC 6 MHz signal to make it a color NTSC 6 MHz signal.

Now→ Remember when we took a moment to talk about why "Y" was called "Y". Remember that we talked briefly about color spaces? Great. Let's chat a little more about color spaces; because color spaces are the key to understanding both analog and digital color TV. (Whole lotta keys around here.)

Any method that lets us discriminate (almost) every color that we can see with just three numbers is called a color space. A red value + a green value + a blue value define one possible color space (RGB). So we could transmit from the TV antenna the red, the green, and the blue (RGB) voltages that the TV camera is producing at that instant. We can add these three values to create almost any color; we'd be transmitting the RGB color space, and that would work just fine.

Except For Just Two Problems→

1. The RGB color space won't fit in only 6 MHz; it's too big.

2. Black and white TVs would not get a proper signal for B&W using the RGB color space; we'd lose compatibility.

We need a different color space than RGB; one that will fit in the 6 MHz we have for all our TV signals; and one that a B&W TV can somehow believe is a B&W signal (we need compatibility).

OK... Get Ready... Here It Comes→ Analog NTSC color TV transmits a color space called YIQ, where "Y" is the brightness, and I and Q are the color information that is broadcast. At any given brightness value of Y, I and Q are like our longitude and latitude, our x and y co-ordinates.

Let's take an example and see how this works... we'll make Y=0.5 (half of the maximum of 1.0, half of the maximum brightness) and we'll also scale I and Q (our x-axis and our y-axis) to 1.0 maximum; let's see what we get→

The YIQ color space at Y=0.5 in IQ steps of 0.25.

Hey, looks good. We can use Y and I and Q to locate any color in the YIQ color space. And Hey→ Aren't we transmitting Y (brightness) and I and Q (chrominance)? We might be able to use this for a color TV system, maybe the analog NTSC color TV system, you think? Three numbers can locate almost any color we can see; YIQ is thus a color space.

So, if our friendly regional analog NTSC color TV station transmits Y= 0.5 and I= -1.0 and Q= +1.0 , we'll get a nice deep blue color... check it out on our YIQ color space; this is truly simple stuff. And holy baloney (Aussie phrase)... the YIQ color space fits into 6 MHz.

The only question that comes to mind is this→ If a color camera sees a deep blue (or any other color), how does it know what I and Q should be on the YIQ color space? In fact, while we're at it, how do we know what Y should be? Like, red, green, and blue signals come out of the color cameras; how do we convert from these signals to the Y, I, and Q that we broadcast?

Enter The Matrix→ Let's revisit our TV station, the color studio with the huge RCA color cameras weighing hundreds of lbs, with cables so thick coming from them that a technician had to hold them up so that the cameraman could physically move the camera. (Like 1954 cameras.) The color TV camera has three sensors→ Red, Green, and Blue.

300 Pound RCA Color Camera With Milton Berle Circa 1954.

And so, no matter what color the analog color TV camera was aimed at, it produced some combination of red, green, and blue signals from its three color sensors. Do we transmit these signals? Nah, it would take too much bandwidth, it would not be compatible with B&W TV sets, remember? And that 1938 Valensi patent would have been for naught.

Instead, we feed the R, G, and B signals to a piece of hardware called "The Matrix" (yes, we loved part I)... MORPHEUS→ "It's that feeling you have had all your life. That feeling that something was wrong with the world. You don't know what it is but it's there, like a splinter in your mind, driving you mad, driving you to me. But what is it?" Ok, enough guys... this matrix is a piece of hardware; the RGB signals from the cameras enter the matrix, and out comes a "luminance" signal→ Y, plus a color signal→ C (with two components, I and Q), whose bandwidth is intentionally reduced. (We don't notice, because we see black and white shapes much more sharply than small areas of adjacent colors.)

Are I and Q equally important? Nah, no way. Take a glance at our YIQ color space with Y=0.5. Our eyes are more sensitive to changes along the I-axis (changes from orange to cyan) than they are to changes along the Q-axis (changes from purple to greenish-yellow). Just the way the YIQ color space is set up. Therefore, more bandwidth is required for I (1.3 MHz) than for Q (0.4 MHz). NTSC Color TV takes advantage of all the quirks in human color vision... as well it should.

Now, how's it do that, the matrix, we mean; how does it convert? Very simply... it uses some arithmetic formulas; and all it really cares about is having Y look as much as possible like the picture in black and white transmissions and having I and Q conform to the YIQ color space. OK... Here are the three secret formulas that the matrix uses to create Y and I and Q at the NTSC color TV station→

Y Luminance = (Red × 0.30) + (Green × 0.59) + (Blue × 0.11)

Q Signal = (Red × 0.21) - (Green × 0.52) + (Blue × 0.31)

I Signal = (Red × 0.60) - (Green × 0.28) - (Blue × 0.32)

Summary So Far→ We created a new color space that will fit in 6 MHz and is compatible with B&W analog TVs. The z-axis is "Y"... which is as close as we can get to the brightness signal that you'd see on a B&W TV. The x-axis we now call I and the y-axis we now call Q. We arrange the colors in our three-dimensional color space so that any unique Y and I and Q values give us a unique color; and so we can display just about any color that we can see using three numbers... Y and I and Q, meaning that we have a color space (since any three values give a unique color, and almost every color that we can see can be reached through some unique combination of Y and I and Q). Go Valensi, go.

And just as long as you have a sharp B&W picture with crisp details, folks are unable to detect that colors are not sharply defined, that the colors are lower in detail. You see, as we said, we tend to see colors spread across broad areas; we don't see colors with sharply defined boundaries, either in the real world or on color TV. (Like, did you notice that the second letter in this paragraph is black and not blue, blue like the rest?)

It's the lack of detail (the greatly reduced detail) in the TV color information that has allowed color TV to fit in the same 6 MHz channels that B&W TV was designed for. Stated another way, because our eyes are less sensitive to color than they are to brightness, bandwidth can be saved by collecting more luminance (Y) detail than chrominance (C) detail. In the analog NTSC color system, we have about 525 lines of resolution, 480 of which are displayed; but only 40 of these lines represent "color resolution".

Transition To Digital For Just A Moment→ NOW... deviating for a moment to ATSC Digital, TV studio cameras and other TV studio equipment usually operate "near the limit" of real-time; in other words, if there were much more data coming from a camera (or a video recorder, whatever), then you'd have to start storing the data in buffers; meaning, no "really real-time" TV. To solve this potential "flubbitz", we use the "gibberish" solution Y'C_BC_R 4:2:2, which in English means that we sample the two color components at half the sample rate of the luminance, so color resolution is halved. (The other ½ is thrown out.)

This reduces the bandwidth of the video signals flowing around the TV station by one-third; and so we now can watch Matt Lauer on the NBC Today Show in real time, with little to no detectable visual degradation caused by throwing away ½ of the color data (and with no buffering required).

(We're just thinking that if we didn't have real-time TV, this could really put a crimp in sporting events and their associated wagering.) We call that 4:2:2 thing sub-sampling, and digital TV stations have been sub-sampling for about ten years now, give or take.

Did we say 6 MHz? Dr. Steve was once an engineer at a color TV station, so long ago (the late 1960's) that the frequency of the TV transmitter would drift up and down (but very rarely, he insists). And who can recall when there was a knob inside the big klunky channel selector on their home TV set? The TV receiver would also drift, and folks had to manually adjust the frequency that was being received... adjust it for the best picture. And so really only 4 MHz of the 6 MHz was used to transmit stuff; the other 2 MHz was used as a buffer to prevent interference; nothing was really quite in it. And in the 4 MHz that had stuff, about 88% was luminance stuff; the other 10-12% was for color and sound stuff.

AM Quadrature Suppressed Carrier→ I and Q are both created by modulating a single color carrier; how do we keep them from interfering? Answer→ Using this special type of modulation, we get an I and Q that are separated by 90º... the TV station delays Q by 90º. We do this so that neither I nor Q has any effect on the other. Very clever, "quadrature transmission" allows I and Q signals to be interwoven; in fact, the "Q" actually stands for "quadrature" (meaning offset by 90º), and the "I" stands for "in phase" (it's not delayed).

The NTSC color modulation technique is actually called→ AM, quadrature, suppressed carrier (QAM) because, well, that's exactly what it is. AM means that the amplitude of the I and Q sine waves varies. Quadrature means that the I and Q sine waves are 90º out of phase with one another. And suppressed carrier means that the color carrier is removed before transmission, and that only the information resulting from modulation... the information in the sidebands... the I and the Q... are transmitted.

(We really don't need to transmit a carrier in AM; all of the information of value is in the "sidebands"... in fact, the sideband above the carrier, and the sideband below the carrier, are originally mirror images; we really only need to transmit one of the sidebands with no carrier→ this is called single sideband modulation.)

We may think of it as if we have squeezed the two color signals, I and Q, into the space of one color signal by using QAM modulation. QAM can carry more information in a given bandwidth than ordinary AM (like AM radio).

BUT... Three carriers, three is a crowd. To prevent interference between the color carrier and the video carrier and the sound carrier, one of the three has to go; as Jim Reeves phrased it in 1960 in the Joe Allison composition→ "He'll Have to Go".

And so the broadcast equipment at the TV station suppresses the color carrier, after I and Q are created. In the world of modulation (mixing information with carriers), once you have used a carrier for modulation, then you really don't need the carrier anymore; so in our 6 MHz diagram that we presented above, you see just a tiny line where the "Color Subcarrier" once lived. And now the video, color, and audio carriers no longer can interfere; because there no longer is a color carrier.

Question→ Folks always wonder→ When is the color carrier there, when is it deleted, does it blink back and forth, on and off? Answer→ In the color modulator where I and Q are produced, the color carrier is there (obviously). BUT... On the way to the video transmitter, the color carrier is removed. In other words, the color carrier is there when we need it for modulation to produce I and Q; but we don't let the color carrier reach the transmitter; it's filtered out before it can be transmitted.

HUH? No color carrier? How can the color TV receivers in our homes extract (de-modulate) the chrominance (C) signal to get I and Q? Good question, because without the color carrier, there is no way that your TV can accurately recover the I and Q vectors. Yeah, as we said, once you modulate (mix the information with the carrier), you don't really need the carrier any more. But when it comes time to extract that information, you must have the carrier back again. (What to do, what to do?)

Solution→ Your analog color TV receiver has a device called an oscillator, it oscillates, it recreates the missing color carrier; and that's why we have the colorburst... tada... it locks the oscillator frequency in the TV set with the color carrier frequency at the TV station that was suppressed. The receiver adds the color carrier back into the chrominance (C) signal, so that I and Q may be retrieved.

In other words... what if, while the electron beam in the picture tube in our TV set is starting a brand new line, getting all ready to to start displaying the next lower line (odd or even), during a time when the beam is "blanked"... (Remember→ "blanking pulses"? We don't want to see those "retrace" lines anyway.)... What if during the 10 microseconds when the electron beam has raced from the right edge of the screen back to the left edge and is just starting to paint a new line, the TV station were kind enough to broadcast just 2.5 microseconds of the original color carrier?

No modulation would be required, just the bare color carrier, like maybe at least eight cycles of it? That could work... and we'd probably call these 8+ cycles the colorburst. (There is a picture of the "colorburst" sitting on the "back porch" a little bit back, when we were looking at pulses sent as a part of NTSC.)

This is really just an extension, for color, of all the "sync pulses" that we chatted about just a little while ago. But now, instead of keeping the beam in our TV set in "sync" with the beam in the TV camera at the TV station, we're keeping the "resurrected" color carrier in our color TV sets in phase with the color signal at the TV station, that color carrier who was suppressed before he could ever be transmitted.

Hey... then your TV set could use the 8+ cycles of the colorburst as a reference frequency to keep your TV set's "oscillator" (the guy who's going to restore the color carrier) running at the same phase and at the same frequency as the oscillator in the color modulator at the TV station. The TV station could do it at the beginning of each line to keep any frequency error of the set's oscillator to a minimum, in case the oscillator is "out of tolerance".

Then your set could recreate the exact color carrier that was suppressed before transmission, suppressed to prevent interference between three carriers. (Recall, "Three's A Crowd"?)

And then you would be able to de-modulate the color carrier in your set and get the I and Q color information. So with every horizontal scan line sent to your color set from the TV station, let's add at least eight cycles worth of colorburst at the beginning of the line.

Uh oh... that colorburst somehow changed its phase during its over-the-air transmission. You don't like the phase of the colorburst that you received? (Now the grass is pink.) Problem With NTSC→ The phase of the colorburst can change during over-the-air transmission. (Solution With ATSC→ ATSC digital has no colorburst, never even had a color carrier that needs replacing... so there's no colorburst to change phase during tansmission.) )

That's why color TVs designed to receive broadcasts in the US have a "tint" control; the tint control changes the phase of the colorburst until the colors on your screen suit you (and reality). At this point, we can see that crow-barring color into B&W TV compatibly was only slightly easier than going to the moon. But it works. By synchronizing an oscillator in your TV by using the colorburst at the start of each scan line, a color television receiver is able to restore the suppressed carrier of the chrominance signal, and in turn, decode the I and Q color information.

And of course, black and white TV sets just ignore the colorburst stuck on the "back porch" of the blanking pulse; B&W TVs think, "What's that? Ah, who knows, who cares; just ignore it."

The Real Story→ Y and I and Q are each different linear combinations of red, green, and blue (recall, the secret formulas in the matrix?). The coefficients for Y were selected so that Y comes out looking as close as possible to the brightness in a black and white picture; then a B&W TV set can throw away I and Q, and use Y, exactly as it used to use brightness.

And because Y and I and Q are all different linear combinations of red, green, and blue, we can "solve" them to get the relative intensities of red, green, and blue at a point in the camera's image. There is a component in the receiver called the receiver matrix. Y, I, and Q go into it, and red, green, and blue come out; it "solves" our equations. (Another way to look at it→ The receiver's "reverse" matrix converts colors from the YIQ color space back to the RGB color space.)

But the real reason that I and Q exist is because they provide a way to broadcast three pieces of information (how much Red, Green, and Blue) with three numbers (Y, I, and Q), and they do it in a way that's compatible with black and white TVs. Thus with Y and I and Q (call it the Valensi system), existing black and white television receivers continue to be usable and color programs displayed on them look reasonable.

In analog NTSC receivers, a lot of the complexity of the receiver is there to receive the I and Q signals. Once I and Q are de-modulated, they and Y can recreate the original Red, Green, and Blue signals to display a color image on the picture tube.

In the NTSC analog picture tube, (usually) three electron beams skim across the screen, painting each line. There is one beam for the red phosphors on the screen, one beam for green phosphors, and one beam for blue phosphors. In the beginning, all color TV picture tubes were round. They all used three electron guns with different "lenses" to focus the electron beams, which then passed through a metal sheet with holes, called→ The Shadow Mask. Even today, all color TV picture tubes, aside from Sony, still use the shadow mask, based on 1950's patents.

(In 1968, Sony introduced the Trinitron picture tube. This new tube used only one electron gun with one large "lens" to more accurately control the electron beam. Instead of The Shadow Mask, the Trinitron uses vertical metal slits to align the electron beam with the colors on the screen. In this way, more electrons get through, less heat is produced, and alignment is only critical in one direction. The tube's face is vertically flat. The Trinitron was such an improvement in TV technology that Sony won an Emmy for it.)

The higher the red or green or blue signals, which our analog TV set has re-constructed from the red, green, and blue in the TV camera at the station, the more each red, green, or blue phosphor on the screen is energized by stronger electron beams. The more electrons coming from any of the picture tube's three guns and hitting a given color phosphor, the brighter that color point briefly becomes. Some phosphor is part of the sky, most of the electrons energizing it come from the blue gun and hit the blue part of the phosphor. That phosphor is part of the green forest, most of the electrons hitting the green part of the tri-color phosphor come from the green gun. A work of art... of mind and of engineering... that whole NTSC analog color TV has been poetry for the last 55 years.

And so the NTSC analog color TV picture is painted on the TV screen one "pixel" at a time. Just like black and white TV, the electron beams start in the upper left and work down the screen just the way we read... left to right, until we reach the lower right end of the screen. And each tri-colored phosphor that the three guns are shooting electrons at glows in proportion to the decoded R, G, and B signals.

♣NTSC Color Odds And Ends♣

"Luck Never Made A Man Wise."   -Seneca

Summary I→ The TV station has a matrix that converts the RGB color space to YIQ space. The station suppresses the color carrier to prevent the chance of interference between three carriers (video, color, and sound). The TV transmitter now broadcasts the I and Q signals, produced by modulating the color carrier, to your NTSC analog color set. The station also is gracious enough to send you a colorburst signal, to allow your TV set to replace the color carrier that the station suppressed.

Now your TV set and the color camera in the studio are in sync, and they are in phase. The inverse matrix in your color TV set decodes the Y and I and Q parts of the transmission back to their original Red and Green and Blue. Then your set's RGB signals are sent to the color TV picture tube; and by mixing these three colors, the color picture tube creates millions of different colors in our visual systems. And Voilà... your TV screen appears to be filled with living NTSC analog color.

(It's almost a shame to throw away analog NTSC color for ATSC digital, now that NTSC is working so nicely. Of course, the variety and richness of ATSC digital color exceeds the capabilities of NTSC color. Still it's sad. For its day in 1953, NTSC analog color was an absolute brilliant feat of engineering. And when you get a converter box, you're converting ATSC digital to NTSC analog. And so NTSC color still will be around for many years. But it also amazes us that a lot of this same NTSC analog color stuff takes place in the small NTIA converter box that receives digital signals and creates the same signals (more or less) that are coming from an NTSC analog color TV station. Mirable dictu.)

VIR→ Today, we look at analog TV, and the color is 2nd nature; we expect it, and we don't expect to adjust color controls (and we really never should have to). We don't recall the last time we adjusted anything on our 20" Sharp TV; just ON, OFF. One reason that analog color on analog TV has become so stable is VIR (Vertical Interval Reference) technology.

You see, the standard analog NTSC video image contains some lines (like, for example, lines number 10–21 of the 525 lines transmitted in each frame) which are not visible. As we discussed, the TV station sends out a vertical blanking pulse, which turns off the electron beams in our TV set when the beams are at the bottom of the picture tube. No beams, nothing on the screen.

This is done so that you will not see any of the pulses that control the picture (and some stuff that is used for testing). And neither are lines 10-21 used for sync or equalizing pulses. Lines 10-21 were deliberately left blank (that is, they were blanked during the vertical blanking interval) in the original 1941 NTSC specifications, mainly to provide time for the creaky 1941 electron beam in the old 1941 B&W picture tube to return to the top of the display.

But by the 1970's, that wait surely no longer was required in a set full of fast computer chips with but one vacuum tube, the picture tube; but the unseen lines were (and are) still there. What might we do with these lines? Maybe closed captioning? Good. What else?

Well, we could add something called VIR. VIR attempts to correct some of the problems with color that are inherent in the analog NTSC video system. VIR adds reference data, inserted at the TV station, to the broadcast. VIR adds data about the luminance (Y) and chrominance (C = I and Q) levels; and it adds all this cool stuff on invisible line 19.

Then VIR-equipped television sets employ this reference data to adjust their pictures to be a closer match to the studio image. The VIR signal contains three sections→ The first is 70 percent of luminance ("Y") and the same chrominance as the colorburst signal; and the other two sections are 50 percent and 7.5 percent of the luminance ("Y") respectively.

Another improvement to color, along with VIR, was that many NTSC analog TV sets began changing anything close to a flesh tone to a factory pre-set flesh tone. It looks much better than greenish faces, even if it is analagous to slapping on "make-up".

Comb Filters→ And then TVs began using comb filters. In the NTSC receiver, we have to separate Y from C (C being I and Q). In the early years of color TV, there was no cost-effective way to perform the Y and C separation other than simple notch/bandpass filters; the cost was low, and it was easily implemented. Unfortunately, color information could leak into Y; and brightness information could leak into C; both of these leaks produced artifacts. And the notch/bandpass filter system removed some of the sharpness in Y, and it removed some of the colors in C.

In the late 1970's, more satisfactory methods of separating Y and C using comb filters began to appear in TV sets at a reasonable price. With decent comb filters, very little luminance or chrominance ended up in the wrong processing channel, and less of either signal was lost. By the 1990's, 3D (Motion Adaptive) comb filters were cheap enough to become standard in quality NTSC analog sets, further improving color and reducing artifacts.

Summary II→ Under the old analog NTSC color system in use since 1953, the color camera scans its image and breaks it down into red, green, and blue. The TV station then converts these red, green, and blue signals into luminance (brightness, or "Y"), and two chrominance (color) signals, I and Q.

The analog NTSC color TV station then broadcasts these three signals over-the-air to all the TV sets that are tuned to this station's over-the-air broadcast. Black and white TVs use only the luminance signal to create a black and white picture. They discard the chrominance signals, I and Q. (Yeah, slight waste of energy.)

But color TVs use all three signals (Y, I, and Q) to re-create the original red, green, and blue signals from the TV camera, which become the color picture on their screens. Analog NTSC color television starts with a black & white signal (luminance); then, basically, color information (chrominance) is overlaid on top. Voilà, compatible analog color.

And if we like, we can let Dr. Steve summarize NTSC color television succinctly and elegantly→ Each video source at the studio produces an RGB color space. The engineers switch among the various video cameras and tape machines and such, each producing an RGB color space. But the RGB spaces cannot be broadcast; they would require more than 6 MHz, and there would be no compatibility with black and white TV sets. And so a matrix at the station converts the various RGB color spaces fed to it into a YIQ color space; YIQ will fit in 6 MHz, and it is compatible with black and white receivers. The station now transmits the YIQ color space from its matrix to all receivers tuned to it. Black and white receivers extract only the Y from the YIQ space, and they produce a monochrome picture. Color receivers use an inverse matrix to convert the YIQ color space back into an RGB space, which produces a color picture on the color tube. (Yes, there is a bit of the poet in most engineers.) And that's analog NTSC color TV.♣

♣ATSC Digital Overview♣

"Do Not Believe Hastily."   -Ovid

OVERVIEW OF ATSC DIGITAL TELEVISION

Now let's move from NTSC analog to ATSC digital TV. We'll begin our adventures in ATSC Digital Wonderland by starting with a fly over, looking down at its components from 50,000 ft.

1. The video signal from digital TV cameras and from other video sources in the TV studio enter the Video Subsystem. Here, video source coding and compression take place. Similar compression of the audio occurs in the Audio Subsystem. The purpose of the video/audio source coding is to minimize the number of bits needed to represent the video and audio. The ATSC coding system includes the MPEG-2 Video Stream Syntax (video compression), and the Digital Audio Compression Standard (AC-3 = Dolby = audio compression).

2. From the Video And Audio Subsystems, the compressed video and audio bit streams, along with ancillary and control data, enter the Service Multiplex and Transport layer. Service Multiplex and Transport refers to the means of multiplexing (combining) the video, audio, and ancillary data streams into a single data stream... and then segmenting that data stream into packets. ATSC employs the MPEG-2 Transport Stream syntax for the packetization of video, audio, and data.

3. Finally, from the Service Multiplex and Transport subsystem, packets of video, audio, and data enter the RF/Transmission System, where Channel Coding and Modulation take place. The Channel Coder scans the bit stream and adds additional information that can be used by digital receivers to reconstruct data that contains errors. The physical layer (modulation) uses the digital data stream information to create the signal to be transmitted, using 8-VSB modulation (a form of AM modulation). (In just a bit, we'll examine the steps in RF/Transmission in detail; this is exciting and important stuff. Stay tuned.)

(Note→ We have briefly described the A/53 ATSC Digital Television Standard, Parts 1-6, 2007, as adopted by the Federal Communications Commission. This version adopted by the FCC is not necessarily the current version of the ATSC Television Standard. (There also exists an (April) 2009 ATSC standard called ATSC-M/H (mobile/handheld) which we will not delve into.)

♣ATSC Digital Color♣

"The Mind Is A Strange Machine Which Can Combine
The Materials Offered It In The Most Astonishing Ways."   -Bertrand Russell

OK... Time to travel to ATSC digital in all its glorious colors. We'll encounter some intriguing changes as we travel the road from NTSC analog color to ATSC digital color. ATSC color is not all that different from NTSC color... well, yeah, it really is.

ATSC Digital is more like→ If TV engineers could redesign color television from scratch, and you had year 2000 technology available, and you knew what you knew about problems with NTSC analog color... how would you guys redo it? Oh, and you're still stuck with those 6 MHz channels.

So how did the engineers handle color in ATSC digital? They removed color from the broadcast phase of ATSC. There is no color carrier. There is no colorburst. There are no I and Q vectors that are de-modulated in the digital receiver. But yet, somehow, The Valensi Solution still shines through. When MPEG-2 is compressing the output from a digital camera, there are lots of Y's... and not so many Reds and Blues. And when our broadcast arrives at a black and white digital receiver (yes, they exist), the Y's become the black and white, and the Reds and Blues are discarded, Now let's look at ATSC color in more detail→

First→ In ATSC digital, there is no requirement for color compatibility. All ATSC digital broadcasts are in color, even if the broadcast material is black and white, like one of the great old Bogart movies. (For some reason, networks often transmit old black and white movies (e.g., "It's a Wonderful Life", NBC) in High Definition, which sometimes gives the black and white picture a very slight greenish cast, since the HD color space is "Rec. 709", which has a higher green weighting than does the NTSC "Rec. 601"; the simulcast of "It's a Wonderful Life" in NTSC analog TV had no greenish cast. As we said, the greenish cast is very slight; if it's not, there are problems with your HDTV set green compensation circuitry... adjust your tint. (Surprise, HDTV sets have tint adjustments.) )

(Rec. 709, is a nickname for "ITU-R Recommendation BT.709, the international standard for television studios' HDTV digital signals".)

However, all ATSC digital television receivers are not necessarily color sets. If you have a set-top converter box connected to a black and white analog TV, you obviously will see the broadcast in black and white . Connect the set-top box to a color analog set, and you will see the broadcast in color. So we still need the Valensi Solution. However, in digital television, we use Y, Y minus Red (Pr), and Y minus Blue (Pb)... and these are the three axes of our digital color space.

It should be noted here that comb filtering, used to separate the luminance from the chrominance in NTSC analog receivers, is not necessary for ATSC digital TV, since the luminance and the color information are already separated.

Second→ Color was not an after thought in ATSC digital TV, as it was in NTSC analog TV. It is an integral part of the original ATSC specifications.

Digital broadcasts, from the TV camera to the TV broadcast antenna, can be divided into two parts→

1. MPEG-2, where compression is performed (including throwing away some color data... called sub-sampling), and compressed sound and miscellaneous data are combined (multiplexed) with the video stream; this combined bit stream (video, audio, data) is then tied up into neat packets.

2. The packets then move along the exciter's "conveyor belt", where error correction and 8-VSB modulation take place. (Along with other stuff; we'll look at the "conveyor belt" shortly.)

Third→ Recall what we said just a moment ago→ In NTSC analog TV, the black/white signal (Y) is made by mixing the three colors from the TV camera according to the recipe→ Y = (0.30 × Red) + (0.59 × Green) + (0.11 x Blue)... which was correct... you can't make this stuff up. And the same coefficients are valid for SD (standard definition, 480i30) digital color TV. (Recall... we do this stuff because our eyes are more sensitive to certain color bands; like, see how much more green we need (0.59) than blue (0.11)? )

However, Luminance for HD (high definition, 1080i30 or 720p60) is calculated slightly differently→ Y = (0.21 × Red) + (0.72 × Green) + (0.07 × Blue)... which gives black and white video sources (e.g., an old movie) a slight greenish cast when the network decides to broadcast the old B&W movie in High Definition... meaning that they send the movie to an HD Encoder instead of to an SD Encoder... just a slight bit more green than NTSC analog, where the YIQ color space was designed so that Y was as close to black and white as possible. (HD was not intended to be 100% perfect for rare black and white programs; the idea for HD was to be able to display as many colors as possible, as accurately as possible.)

Hint→ Believe it or not, as we said, HDTV sets have a tint control; they also have a color level control. If the slight green tint on the B&W Roy Orbison documentary bothers you, turn the color level down until the green goes away; be sure to note where the color level was originally.

Sloppiness→ This is just more sloppiness in the ATSC digital design. A bit set here, a bit set there, and your HDTV set could recognize B&W programs broadcast in HD. And then WHAM... the set flips its coefficients, and the greenish cast is gone. This would cost about $0.01 in additonal ciruitry in the HDTV set.

In Other Words→ For folks curious about why the coefficients of Red, Green, and Blue used for computing luminance are modified slightly with High Definition video, we'll begin by reminding you that NTSC analog color and ATSC digital Standard Definition use the YIQ color space, also known as Rec. 601.

High Definition coefficients, on the other hand, are referred to as color space Rec. 709. HD uses a slightly different color space than NTSC analog TV and digital SD. So now the question becomes→ Why does HD use a new color space?

Answer→ HD uses the Rec. 709 color space because 709 permits a broader color gamut (the range of available color shades) than does Rec. 601. For example, the NTSC color space (Rec. 601) cannot display full, deep, rich blacks; the Rec. 709 color space can... especially on a plasma flat-panel screen. The Rec. 601 coefficients were chosen for maximum black and white compatibility, since B&W is created strictly from Y (luminance), with I and Q discarded. The HD coefficients were chosen for maximum color (saturation and hue) possibilities.

Back in 1953, deep, rich blacks were not a concern. There were no plasma sets displaying High Definition back then. In the NTSC analog system, the object was for Y to produce the best black and white picture possible... no slight tints of any color... since black and white transmissions had to be viewed on color sets in... well, in black and white.

SO... Black and white broadcasts in ATSC digital High Definition may have a faint greenish tint which requires a slight adjustment to your HDTV set; and the same black and white program simulcast in NTSC analog will produce a pure black and white picture when viewed on an NTSC analog color receiver. But for color broadcasts... ATSC digital High Definition will display a color space with a wider color range (a wider gamut) than NTSC analog color.

And thus the cat is now out of the bag→ Analog NTSC color and SD can display most of the colors that we can see... but not all (and not as many colors as High Definition).

Fourth → Sound; while sound is not directly related to color, since we sneaked it into our chat on NTSC analog color, we should probably sneak it in here. ATSC digital does not use a separate FM transmitter for sound, as analog NTSC does; in fact, ATSC digital doesn't use any kind of separate transmitter for sound. And unlike NTSC, ATSC does not use a diplexer to combine the video and the sound carriers (radio frequency waves). ATSC does not send the combined video carrier and the sound carrier to the TV transmitting antenna. In the ATSC world, there is but one carrier.

Instead, once the ATSC digital video is compressed, and once the digital Dolby sound is compressed, these two compressed digital signals are combined in a multiplexer (more on multiplexers coming up); and then the output of the multiplexer, the combination of all of the various bit streams, eventually ends up in one single transmitter. (See the ATSC block diagram, a few minutes back.)

Fifth→ In NTSC color TV, we transmit Y, I, and Q signals. In ATSC, we convert Y, Y-R (Pr), and Y-B (Pb) into numbers (we digitize them), just as DAwn's scale converts her weight to three integers.

Sixth→ Recall how, under the NTSC color system, only 10-12% of the truly available 4 MHz was used for color and sound? Some things never change. In the ATSC digital compression system (MPEG-2), a technique called sub-sampled color is used, and most of the digital bits are used for conveying brightness information. Reducing the color detail is done early on in MPEG-2 video compression; we throw out all but the gross color details, to shrink the data that is passed along to represent the image in the TV camera. Yes, we went from analog to digital, but the human eye has not changed; we still throw out the color detail that we cannot see. But now, it's some of the color bits that we trash.

Seventh→ It's 2010. TV transmitter frequencies don't drift (unless the transmitter is in motion). TV receivers long ago abandoned their fine tuning knobs. That 2 MHz of each 6 MHz that was empty in 1946 to prevent channels from interfering with one another no longer has been necessary for a long time. And so, what was a major problem for TV in 1946 is now ancient history.

With ATSC digital TV, almost all of the 6 MHz for every channel has stuff stuffed into it. In digital TV broadcasts, about 5.38 MHz out of 6 MHz is now stuffed with stuff to be transmitted. The electronics of 2010, both for the TV transmitter and for the TV receiver, has long ago stopped being subject to frequency drift and other accuracy fubars.

With these points in mind, we're now prepared to dig slightly deeper into fertile soil of ATSC digital, and to examine just how digital TV handles color.

♣Deeper And Deeper We Dig Into The Fertile Soil Of ATSC Digital♣

"What Is Truth?"   -Pontius Pilate

Quick Tiny Review→ In an NTSC analog color broadcast (we keep repeating "NTSC" because there are other analog color systems used in other countries that are not NTSC, like "PAL", popular in Europe.), a signal is transmitted representing the brightness and the color of a point in the image the TV camera has focused on. And recall, the broadcast process works in reverse; the brighter the point in the camera's image, the lower the percentage of maximum signal... so 12% of maximum is bright white and 75% (and above) is black. This is analog.

In digital transmissions, numbers proportional to the intensity of electrons that we want fired at a pixel, or the voltage that we want applied to a plasma "bead", whatever, are broadcast to our digital TV sets.

HUH?

Ok, let's review, before we now are fully immersed in the cool, cool ocean on this hot June day. (Yes, we are tapping on our laptop this warm June day, with our feet in the cool ocean at Ocean City, MD.) In NTSC analog color, the image in the color TV camera is scanned, creating a signal that's proportional to the brightness of the image at any point. And since it's analog color, any signal strength between 12% and 100% may be transmitted, just as long as it's between 12% and 100% of the maximum... exactly like DAwn's reading lamp with the dimmer... anything percentage-wise goes.

Actual picture information doesn't exceed 75% of the maximum permitted (black). 100% is reserved for "sync pulses", and a few other special things. It's cool to have sync pulses higher than any possible picture data; it gives the receiver something to "grab onto" when sync pulses are competing against static in the picture; the sync pulses are always higher. (We sometimes wish our minds ran at 75% maximum, and had sync pulses at 100% to assure sanity; but pecking away here at the edge of the ocean... we'd like to have something at 100% for sanity to grab onto.)

But as the analog TV camera scans an image, the transmitter gets that image signal anywhere between 12% and 75% of its maximum. In the analog world, anything goes. (Like the book by Wodehouse and Bolton upon which the Cole Porter lyrics and music were based.) Dr. Steve, you ever see "Anything Goes" on Broadway, like in 1934? Ten years before you were born? Ok, we won't involve DAwn in getting into negative ages.

12.2%, 23.87%, 74.983%, you name it. You concoct any number between 12% and 75%, it's Kosher. If your TV set receives some interference at the 15% video signal level... your TV receiver says, "Cool, I'll display that, and brightly too; looks just like white snow." Our analog TV receivers cannot tell interference, perhaps caused by the plywood drying machine three miles away, from the legitimate stuff that the TV station's video transmitter broadcasts; and so, our analog TV set displays it all.

But ATSC digital is different. Remember, the Giant SuperMart is open or closed; it's digital, never 15% open. In ATSC digital, only certain signal strengths from the TV transmitter are legal (at the instant when the digital TV receiver samples the transmission). 8-VSB means that there are only 8 valid modulation levels at the time the signal is sampled. And we can represent those 8 levels with three bits... which is why 8-VSB modulation allows eight level symbols... and these eight level symbols are the three bits of the transition words coming out of the trellis coder. (More on this coming up.)

Maybe the numbers 10%, 20%, 30%, 40%, 50%, 60%, 70% and 80% are legal. When our digital receiver sees the number 35.7% coming in from its antenna (during a sampling instant), our ATSC digital set knows that something is wrong. But it's smart; it can be programmed to throw away a bad number. Maybe the set will wait for the transmitter to rebroadcast the flubbitzed number (if the problem came from the TV station), and there will be a brief "judder" in the program (even plastic surgery leaves faint scars). Or maybe the TV will just ignore the 35.7%, if it seems that the number somehow came from the plywood drying plant; or maybe the digital receiver will use some error correction technique that was inserted at the "exciter" to fix the problem.

The set may be smart enough to figure out that the 35.7% number is really 30%, and it displays it as if it had been received as 30%. (It looks at the past history of numbers received and "learns".) But no matter, the digital TV is NOT going to accept 35.7% for a pixel... unless we decide on Day One that 35.7% is a Kosher value.

There are more ATSC digital differences... MPEG-2 compression is a way of life for the video. If a man on a horse rides through a scene where half of the picture is blue sky, MPEG-2 can tell the digital TV receiver, "A man rode through the scene"; but you don't have to keep transmitting the sky again and again;we get the picture. In NTSC analog, you would just keep transmitting the sky... over and over and over again.) Voilà... a lot of information doesn't need to be re-transmitted in ATSC digital color TV. You can tell the digital TV set, “This bunch of brown horse fur moved in this way". ¿Está claro?

NO?

OK... A digital transmission begins with the digital TV camera scanning light-sensitive targets, just like an analog TV transmission. There are three targets in the TV camera, one for red, one for green, and one for blue. The red target gets the red light, the green target... and so on. The signals from each of the three targets go to the Matrix (recall?). Here we create the luminance, the black and white "Y" signal, from the Red, the Green, and the Blue.

The Big Difference→ Now, here's where things change. Three signals... Y, Pr (Y minus Red), and Pb (Y minus blue)... are now measured at the TV station early on, and they're converted from voltages to numbers... converted to digital... just as when DAwn stands on the digital scale in her bedroom, her analog weight is measured and converted into a number. (DAwn says with pride that her weight this morning, converted to digital, was 107 lbs.)

Every pixel that's scanned in the TV camera gets a Y number; every alternate pixel gets a red and blue number. Digital television requires that raw signals be converted to numbers, so that they can be processed by computer hardware. There are are no analog I and Q signals in digital television... Pr and Pb are close to replacing I and Q, however.

Now, in the world of analog TV, a single line is sent from the TV transmitter to our TV sets as a continuously varying signal... "continuously varying"... the definition of analog.

BUT... in the wonderful world of digital TV, each line is sampled as a string of pixels. On a Standard Definition broadcast (480i30), with a 4:3 aspect ratio and square pixels, if we have 480 lines of picture, we have to have 640 pixels per line (because the pixels on a digital TV set are square, and (4/3) × 480 = 640). And at the TV station, digital TV assigns a number to the brightness ("Y") at each of the 640 pixels on a line... and the digital TV camera also provides numbers indicating the intensity of the red and blue signals at alternate pixels.

Now, ATSC digital arranges Y, Pr, and Pb numbers eight to a row, eight rows to a Block→

Again, each 8x8 "Block" contains 64 luminance numbers, or 64 numbers for (Y - red) pixels, or 64 numbers for (Y - blue) pixels. BUT... Only one type of pixel number (Y, or Pr, or Pb) is allowed in one 8x8 Block. And every number in any type of block is encoded as 8-bits.

Now let's combine four 8x8 luminance blocks with one Pr (Y - Red) 8x8 block and with one Pb (Y - blue) 8x8 block, to create what is called a macro-block. Here is a diagram of many macro-blocks→

MANY MACRO-BLOCKS

For ATSC digital compression, macro-blocks are very cool. Macro-blocks reduce the digits that must be transmitted to create a picture. With each frame, only the macro-blocks that have changed are transmitted.

Each pixel in a block is subtracted from the same pixel in the previous frame. The objective is to transmit as many zero-valued pixels as possible.

Note that in each macro-block, we have thrown away ¾ of the color information for the pixels, which we abbreviate with the code→ (4:2:0). We can get away with this just as we got away with sharply reducing the bandwidth of the chrominance back in NTSC color. We detect color with the three types of cones in our retinas, but cones cannot detect sharp color boundaries.

Each measurement, when we convert analog signal strength to digital numbers, would require 8-bits if we did not "thin" the color blocks. If we don't apply compression to the color, an 8x8 color block would use 512-bits (8 × 8 = 64 elements, × 8-bits = 512-bits). But with (4:2:0) sub-sampling, we can get an entire block (8 × 8 = 64) of color elements all the way down to about 100 bits.

Great... So what does this (4:2:0) have to do with filtering out some of the digital color information?

The 2 represents the horizontal "sub-sampling" of Pr and Pb. Recall that we said that we sample the brightness ("Y") of every pixel, but only the red and blue components of alternate pixels? That's what the 2 means. Color detail is reduced by this sub-sampling of Pr and Pb. But provided that we don't do anything to degrade the luminance, and provided that we keep its full detail, we won't notice the color bit-reduction.

And finally 0. Zero is a code that says we're doing vertical subsampling at a 2:1 ratio. In other words, along a row of pixels, we throw away the color information from every other pixel. And in addition, we throw out the color information in every other row of pixels. And so ¾ of the color numbers are deleted. Throwing away these numbers allows us to get an HD picture that fits into a 6 MHz slot... and (4:2:0) allows us to broadcast more than one program (multicast) in a digital "major channel".

And each block (8 × 8 = 64 8-bit numbers) is further compressed in MPEG-2 by three processes→

1. A Discrete Cosine Transform (DCT) is performed on the block (shrinks the data if the block is not truly random).

2. A Quantization Process is performed (throws away low-order bits of the transform coefficients).

3. Variable Length Encoding (assigns very short codes to common values, longer codes to uncommon values).

Now we're ready to explore modulation. We've used the word a lot in the previous sections (along with 8-VSB), but we've never quite explained it. So now, your dog Wolf will explain it; we mean, after all, this is important stuff. And good stuff.

♣Just What's Modulation And 8-VSB All About?♣

"Nothing Is So Firmly Believed As That Which We Least Know."   -Montaigne

When High Definition television is broadcast from the digital TV transmitter connected to the TV transmitting antenna high up on the TV tower, over-the-air through the ether, to "rabbit-ears" or a roof-top antenna connected to our HDTV set, the TV station relies on a modulation technique called 8-VSB... which stands for 8-level vestigial sideband modulation. (Nah, just a few big words; this is very simple stuff. Modulation is even legal in Maryland.)

Why? Why dog Wolf? Because the ATSC standards (remember them, also known as A/53?) and the FCC who adopted these standards say that when broadcasting digital TV over-the-air in the United States, we have to use the 8-VSB modulation method (well, yeah, ok... and the newer standards approved in 2004 also say we can use the E8-VSB modulation method... where E stands for "enhanced"... but it's pretty much the same modulation method... just more punch, less finicky, more manly and robust, but with a lower data rate). NOW... Hold these thoughts→ Modulation... 8-VSB... Sideband.

(Yes, there is a new set of standards called ATSC-M/H (or A/153) that permits digital TV in the US to be received by mobile receivers and handheld devices; right now, ATSC-M/H seems to be limited to buses riding about Raleigh, NC; it's approval is recent (2009); why was it not a part of the original standards?... ask the ATSC. Someday soon, we may even add a section to this opus on ATSC-M/H; but no, not just yet.)

8-VSB Modulation? HUH? Just What's Modulation?

In order for an electromagnetic wave (also called a "carrier wave" or a "radio wave") generated in a transmitter and radiating from a TV antenna and traveling near the speed of light to carry pictures, sound, and data, the "carrier wave" must be varied in some way. AND... That variation in the "carrier" must have some relationship to the information that we want carried. This is the whole secret of using a radio wave to transmit (carry) information... variation of a carrier wave, to the beat of the drum of some information.

The information that we want to transmit by a radio wave "carrier" is usually low in frequency. Music, for example, is between 20 and 20,000 cycles (ups and downs) per second (now called Hertz... Hz) for humans. But so what? Why don't we just convert the music to a varying electric wave (exactly the way a static telephone does), and then perhaps run the varying electric wave through a sound amplifier, and then just connect the output of the sound amplifier directly to an antenna? Why even bother with "carrier waves" and "modulation"?

ANSWER→ Yes, we could, and it would work. But "radio waves" have different properties at different frequencies... at 1,000,000 Hz (1 MHz), a 50,000 watt transmitter can just about transmit information to most of the 48 states (and a lot of of southern Canada to boot). But with a 1,000 watt sound amplifier, we might be able to transmit radio waves at 20-20,000 Hz (sound waves)... our music... a couple of miles... not very far. And even worse, the interference from electrical equipment in the 20-20,000 Hz range would be devastating. And third... how exactly do we separate our musical broadcast from the station down the road, also broadcasting from a sound amplifier? It doesn't work, we can't separate the stations.

SOLUTION→ What if, somehow, we could "mix" a much higher frequency radio wave with our low frequency sound waves, like mix 1,000,000 Hz with our musical waves? We could call the much higher radio frequency wave a carrier wave. And we could call the process of mixing the low frequency sound waves with the high frequency radio wave→ modulation.

Modulation is mixing lower frequency information that we want to transmit with a higher frequency radio "carrier" wave, a radio wave that will "carry" our information a considerable distance.

And because of the way that radio waves at different frequencies "propagate", we now could cover an entire town or state or country with our carrier wave. Plus, there would be much less interference from electrical devices. Plus, if we used 1.0 MHz as a carrier wave, and the guy down the road used 2.0 MHz as a carrier wave, we could separate the two in our receiver... we could "tune" to only the station that we wanted to receive.

Sounds much better this way. Sounded much better in 1922 also, when commercial AM radio broadcasts were born. The clever trick of "mixing" lower frequency sound waves with higher frequency radio waves in an electronic gismo called a modulator has been named... yes, modulation.

The whole key to modulation is to mix lower frequency information (perhaps audio frequencies, perhaps the 6 MHz of TV picture and sound) together with a higher frequency carrier wave (like 83.25 MHz... TV channel 6). The information that we mix with the carrier is often called by engineers the baseband signal. A baseband signal (or "lowpass signal") is simply a low frequency signal that can include frequencies that are equal to (or very near) zero; for example, a sound wave can be considered a "baseband" signal.

NOW... In some way, the higher frequency carrier wave has to vary in step with the low frequency information that we are mixing with it. This is the key to radio, television, satellite communication, etc. All in all, this idea of modulating a radio carrier wave, an idea that came to fruition in the late 1800's, has been exceedingly cool; like, it has allowed us to broadcast color TV, from men in vehicles on the moon, back to Earth.

By modulating a radio carrier wave with "information", the information can travel much, much farther than if there were no "carrier" wave. (Can you now see why it's called a "carrier wave"?)

So we have a partnership... the radio wave (or "electromagnetic wave" or "carrier wave") goes FAR; and the information mixed into the carrier wave by modulation allows the radio wave to carry the mixed-in information FAR along with it. In the illustration below, a carrier wave is first being "amplitude" modulated→ we vary the amplitude of the carrier wave according to the lower frequency information that we want to transmit.

Second, we have an example of "frequency" modulation→ We vary the frequency of the carrier wave (a little bit) according to the lower frequency information that we want to transmit. (AM and FM both have advantages and disadvantages... see "No Free Lunch" theorem.) But for this modulation idea to work, we have to vary something about the carrier wave.

AND... When the modulated carrier wave reaches our receiver, whether it is on the next block or across the ocean, we remove the carrier wave; this leaves just the low frequency information. If we were transmitting music, after we separate the music from the carrier, we might pass the music on to a small amplifier connected to a speaker. If we're receiving FM (frequency modulation), and we tune to 88.1 MHz... Voilà. "Los Campesinos! - You! Me! Dancing!" Tune our receiver now to 88.3 MHz, receive a different modulated carrier, extract its low frequency information, and it's Dvorak's New World Symphony (No. 9) in E Minor. See, modulation is cool.

A Carrier Wave After Being AM and FM Modulated By A Sound Wave.

AM and FM radio broadcasts are examples of ANALOG modulation. Why? In analog modulation, the low frequency analog modulating information (the "baseband signal"; e.g., music) is continuously mixed with the carrier. Now we want to look at AM; it's very cool... and it's really how analog TV (VSB modulation) and digital TV (8-VSB modulation) do their video thing.

♣Now Let's Look At Amplitude Modulation (AM) In More Detail♣

"Human Beings, Vegetables, Or Cosmic Dust, We All Dance To A Mysterious Tune,
Intoned In The Distance By An Invisible Player."   -Albert Einstein

A LITTLE HISTORY→ Modulation is the KEY to broadcasting anything to anywhere. The concept of modulation was conceived almost 100 years ago, and it was the "golden key" that made radio broadcasting, and later television broadcasting possible. AM radio began with experimental broadcasts in 1906 (by Reginald Fessenden, a name you don't hear every day), and it was used for small-scale voice and music broadcasts.

By 1915, things were chugging along nicely, and voice could be transmitted by AM radio from New York City to San Francisco, thanks to AM modulation of radio waves. And in 1922, commercial AM radio stations began springing up in towns all over the US. And then, on 28 August 1922, it came to pass... The first radio commercial was broadcast by WEAF, New York, for the Queensboro Real Estate Corporation. (The ten-minute live commercial was spoken by H.M. Blackwell, a representative from Queensboro... remember this for Trivial Pursuit.)

OK... AM... we start with an unchanging radio carrier wave; perhaps its frequency is 600 KHz (600,000 complete cycles per second), just like WCAO-AM in the metro Baltimore area. And perhaps the amplitude of the radio wave is ±50 volts... whatever.

Then we begin changing that amplitude in some way... varying the amplitude... modulating our 600 KHz radio carrier wave... modulation... this is the key to all broadcasting, to all cable-casting, to all satellite-casting; the changes to our originally unchanging electromagnetic carrier wave contain the INFORMATION that we want to transmit. (Why do we keep saying that? Why did Johnny Cash keep singing "Walk The Line"? Who knows of such things?)

When the varying "modulated" wave reaches your receiver (AM, FM, TV, cell phone, whatever it may be), the receiver de-modulates the modulated carrier; it separates out the information part that we mixed with the "carrier wave", throws away the carrier wave that allowed it to reach us, leaving only the information part of the broadcast... be it data, HD television, analog TV, FM music, whatever. Just like being a passanger on a train; when we reach our destination, we get off the train; we don't need the train any more, throw it out. Trains can travel very far, but they have to carry something... freight or people or mail (Do trains still carry mail?)... to make them useful... just like radio carrier waves.

(Whew... give broadcast engineers something like modulation to play around with, and they will play with it until it becomes very... well, very cool. (Some read "very cool" as "complex".) For example, engineers have added "subcarriers" to the broadcasts from FM radio stations, permitting FM stereo and FM multicasting and HD Radio^©. But they surely mean well (We love you, Dr. Steve), and with added complexity often comes added function. ATSC digital TV has to combine a bit stream of improved video and a bit stream of Dolby Sound and a bit stream of data. ATSC digital TV transmission starts with various bit streams... yes, there are many steps in mighty 8-VSB modulation, to be sure.

(Actually, engineers tweak and create new modulation methods partly because of something called the Shannon-Hartley Law. (It's a descriptive law, not a prescriptive law; you can't get fined for violating it, and it was never upheld by the Supreme Court.) The S-H law was proved in 1948, quite a while after Mankind started "modulating". But it says that if you tell me the transmission power you're using, and the bandwidth of the channel (e.g., 6 MHz), and the level of noise, then I can tell you the MAXIMUM number of bits per second that can possibly be transmitted over your channel by any modulation method that will ever be conceived. (Kind of like the speed of light... you can't accelerate to the speed of light and exceed that speed, and you can't send more data than the Shannon-Hartley data rate.) And so engineers tweak and create their modulation methods, always trying to get closer and closer to the maximum data rate that the S-H Law predicts is the maximum possible.)

We've talked about AMPLITUDE MODULATION (AM) because 8-VSB uses AM here and there, among other things. And 8-VSB is the way that we get digital (and thus High Definition) onto a TV carrier wave; 8-VSB is our modulation method for ATSC digital television, broadcast over-the-air. We'll look at 8-VSB in some detail when we dissect "The Exciter".

♣Digital TV MPEG-2♣

"There Is Nothing More Difficult To Take In Hand, More Perilous To Conduct, Or More Uncertain In Its Success,
Than To Take The Lead In The Introduction Of A New Order Of Things."   -Machiavelli (The Prince, 1513)

REVIEW→ To convert High Definition video in the TV studio, from streams of billions of bits per second, racing out of digital cameras and digital tape machines, into something way down, at least in the millions of bits per second, into something that we can actually broadcast over-the-air in digital TV format, two BIG multi-part operations are employed→

1. First, The MPEG-2 Encoder.

2. And second, The Exciter.

And as we might expect, two big, expensive pieces of equipment are required to execute these operations... the MPEG-2 encoder and the 8-VSB exciter. And so, let's begin first with a peek at what MPEG-2 can contribute to the wonderful evolutionary leap from NTSC Analog to ATSC Digital TV.

(And then, after we expound a tad on MPEG-2, we'll look at the formidable, but nonetheless exciting, "Exciter".)

OK... Are you familiar with how the Internet works? (You are, if you've read my Dog Wolf article on the subject, THE NET.) Well, moving stuff around the Net is done by a thing (a bunch of sexy protocols) called IP; and what IP is to the Internet, MPEG-2 is to Digital TV.

Yes, MPEG-2 is THAT important.)

So just what is this piece of the Digital TV puzzle anyway, this piece that's called MPEG-2?

OK... Let's imagine that we have a wonderful, vast Interstate "Bitway"; no, not an analog highway like I-95; this one is digital, and we'll call it I-Transport-Stream... it's not an interstate that handles cars and trucks and busses... I-Transport-Stream is an interstate that carries bits... 0's and 1's.

(Makes some sense... after all, it's a Digital interstate "bitway".)

Oh, and one other thing... Our digital bitway is special; it's a one-way highway. And it has very few exits, maybe just one... an exit to Exciter-City. (Guess that was two things.)

The design goal of the MPEG-2 I-Transport-Stream is to allow the "multiplexing" (i.e., combining) of digital video bits and digital audio bits, and to synchronize them. (Yes, it's always nice when the picture and the sound are in sync.)

So... I-Transport-Stream has millions and millions of bits zipping past any point every second... and it can even accomodate up to 38.78 million bits per second for those bits heading to Satellite-Ville and Cable-Town.

After we pass these two exits, our interstate "bitway" narrows... and so for bits going to Exciter-City (bits that we would like to be transmitted over-the air), our interstate "bitway" has a capacity of 19.39 million bits per second.

NOW... Like any good and sensible Interstate Highway, all along I-Transport-Stream are... you guessed it... entrance ramps (Makes sense, yes? We may not have many off-ramps, but we have a slew of on-ramps). And after we travel along our "bitway" a bit, we soon come to realize that the road signs, instead of saying "Entrance Ramp Ahead, Please Merge", all seem to say instead "Elementary Stream Ahead, Please Merge"; and fresh, new bits whiz onto I-Transport-Stream from each these entrance ramps, ramps who are signed with Elementary Stream.

  
>>==^=====================I-TRANSPORT-STREAM===========>>> PID  515---> {TO
>>==|====^================I-TRANSPORT-STREAM===========>>> PID 1023---> THE
>>==|====|=====^==========I-TRANSPORT-STREAM===========>>> PID    2---> MULTIPLEXER
>>==|====|=====|======^===I-TRANSPORT-STREAM===========>>> PID   77---> }
    |    |     |      |
    |    |     |      |  
    |    |     |      |<---(Encoding)--"PID=77"---Elementary Stream (Video)
    |    |     |     
    |    |     |
    |    |     |<-------------"PID=2"-------------Elementary Stream (Data) 
    |    |
    |    |
    |    |<---(Encoding)------"PID=1023"----------Elementary Stream (Video) 
    |
    |
    |<---(Encoding)-----------"PID=515"-----------Elementary Stream (Audio)

Also, watching the bit traffic entering our I-TRANSPORT-STREAM "Bitway", we conclude that there are three different types of entrance ramps→

• Encoded Video entrance ramps.

• Encoded Audio entrance ramps.

• Data entrance ramps.

Now recall our road signs saying "Elementary Stream ahead, please merge". Well, each overhead sign announcing an entrance ramp also has a unique identifier... like most interstates... and the entrance ramps along our "Bitway" each have a unique number called a PID. After a little thought, we decide that PID must stand for Packet identifier.

NOW... The bits from each entrance ramp, the bits comprising each Elementary Stream, all stay in a unique lane in the wide Transport Stream (at least for a bit). The Elementary Streams are like Olympic runners, each in their own lane. We can think of each lane on our interstate "Bitway" as a Sub-Stream of the "Bitway". And all of our bits are well-behaved; like, our Elementary Streams never move into someone else's lane on our "Bitway".

We also notice that our bits seem to enjoy socializing in groups of eight, and so we'll call these cliques of eight bits a byte.

Each Elementary Stream moving along our Transport Stream Bitway is identified by a PID, an ID number from 0 to 8191. And so our Transport Stream (our "Interstate Bitway") is populated by up to 8,192 Elementary Streams; and each of these Elementary Streams has a unique PID.

(After the end of our Bitway (e.g., our HDTV receiver), a De-Multiplexer extracts each Elementary Stream from the total Transport Stream... in part, by searching for packets identified by the same PID number.)

Cool, but that's not all. We also notice that bits entering our bitway from Encoded Video entrance ramps and from Encoded Audio entrance ramps must pay a small "toll" before they are permitted to enter our Digital Interstate; and when they pay their toll, they are then MPEG-2 compressed (encoded). (Technically speaking, encoded audio bits are compressed by Dolby (AC-3) rather than by MPEG-2 compression; Dolby won that War of The Roses.)

In other words, only after an Elementary Stream that is an Audio Stream or a Video Stream pays its toll and is MPEG-2 compressed can it become an Elementary Stream that enters our Bitway.

(Recall our previous chat on MPEG-2 compression? It included techniques like Discrete Cosine Transforms and a Quantization Process and Variable Length Encoding? We're just revisiting it here. Whatever we do for compression has to be done in reverse sequence in our digital TV sets... "decompression ". Makes the old analog TVs seem like Fisher-Price toys by comparison, ey?)

BUT... Why do we say that the bits pay a small "toll"? Because the processes of compression are lossy. When our bits are eventually de-compressed, we find that we have lost some of them; not enough to worry about... but some bits will never re-appear.

Because MPEG-2 compression is lossy, the more you compress a signal, the lower the resulting quality will be, until the receiver literally can no longer display anything but tiled garbage ("pixelation").

Another thing that we notice about our Elementary Stream entrance ramps is the friendly traffic cops controlling the bits on each ramp. The cops let a certain number of bits go through onto the Transport Stream, and then they stop the bit-flow for just an instant. Then they let another Packet of bits through. In effect, our friendly traffic cops on the entrance ramps break each Elementary Stream of bits into Variable Length Packets.

And after our traffic cops divide Elementary Streams into these relatively large packets, we say that we have a Packetized Elementary Stream (PES). The Elementary Stream is packetized by Encapsulating sequential data bytes in the Elementary Stream, and then slapping on front some PES Packet Header bytes.

One reason that our friendly traffic cops perform PES-ing is to make sure that there is tight synchronization between video and audio and "private data". Each variable length packet in an Elementary Stream contains a header and a Payload; and the Payload is a single Frame of video (or of audio)... in other words, it's highly useful data.

But why "PES-ing"? Our friendly cops do this "PES-ing" because of something still up ahead on our Interstate Bitway, something after the last entrance ramp. Just as Interstate Highways often require all trucks to pull off into weighing stations, our Transport Stream requires all Encoded Video packets and all Encoded Audio packets to pull off of our Interstate Bitway for a moment... At The Multiplexer. (Ta da.)

Many of the packets on each lane of our Elementary Streams run into the Multiplexer. These packets are relatively LARGE, designed to make the work of the multiplexer quick and efficient.

And as we examine the multiplexer closely, we see that only one very wide lane exits from it. And our big packets that entered the multiplexer, each with one audio frame or one video frame, are now much smaller upon exiting, just 188 Bytes each... and obviously no longer variable length.

Yes... the multiplexer has broken our "PES's" into tiny packets. And it looks like there are about ten times as many Compressed Video Packets as there are Compressed Audio Packets (which makes sense). Each packet coming out of the multiplexer still is either a 100% video packet or a 100% audio packet... packets never mix, they always have bytes of the same type.

And so our multiplexer has combined lots of lanes of Elementary Streams on our Digital Interstate into one single, very wide lane, with just one Transport Stream, now carrying 188 byte Encoded Audio packets and 188 byte Encoded Video packets, each packet now just a fragment of a complete frame. Tell us why, dog Wolf?

Why the new tiny 188 byte packets? Because that size is ideal for the error correction that soon will take place for these packets on their way through Exciter-City... ideal for the error detection and correction techniques that will be applied to the packets by the Exciter.

We call these 188 byte packets Transport Packets. And every now and then, we notice a 188 byte Data Packet popping out of the multiplexer from, well, nowhere... mixing in to the other packets on our wide Transport Stream. Hmmm... new "secret" elementary streams. Maybe these new packets help to keep track of things? Hold that thought for just a moment.

Why 188 bytes? Really want to know this stuff? MPEG-2 originally expected its packets to be carried over ATM (Asynchronous Transfer Mode) networks... networks using the ATM protocol for sending data around the Internet and elsewhere. At that time (the mid-1980's), ATM cells were supposed to have a payload (data portion) of 47 bytes. And 47 × 4 = 188.

ATM, however, settled on 48 bytes of data and a 5 byte header for each of its fixed length packets. But anyway, it was considered a cool idea to make the MPEG-2 Transport Stream packets compatible with ATM. (Today, ATM is used outside of Local Area Networks (LANs), but it never gained widespread use; and its complexity prevented ATM from becoming "the one integrating network technology", as its designers had intended. Color it obsolete.)

Now... the packets from nowhere... the multiplexer also has inserted its own Elementary Streams... streams that it has created, streams that it has inserted into the Encoded Video packets and the Encoded Audio packets that entered the Multiplexer and which it is combining.

This new information from nowhere, called Service Information, describes the Transport Stream as it leaves the multiplexer. Simple stuff here. MPEG-2 still keeps track of related video, audio, and data packets (maybe it's all some TV program), as if these related packets were an Elementary Stream, even though everything is racing along on the one lane Transport Stream after leaving the multiplexer.

Now... we've made it sound like random Elementary Streams just enter the main Transport Stream. That is not exactly true.

In MPEG-2, we also have the concept of a Program. Certain PIDs (remember PID numbers?) are related and are part of the same Program; MPEG-2 keeps track of PIDs related to a Program. For example, a Transport Stream might contain three Programs; and each program might be a complete television channel.

A receiver wanting to tune in a particular channel just has to decode the Payloads of each PID associated with the desired Program. It can discard the payloads of all the other PIDs.

SO... How does the receiver know which PIDs it needs? Simple... MPEG-2 inserts a table into the Transport Stream, showing all the PIDs for every Program ; the digital TV receiver just reads that table, called a Program Map Table (PMT). Yes, an HDTV set is smart... part display screen, but also part computer (and whatever other parts it needs).

A Program is composed of a series of Events. Every Event is a discrete television program. Yes, major confusion. Let's put things in order.

A Transport Stream contains one or more Programs; each Program might be a unique television channel. So our Transport Stream is composed of Programs. Programs are composed of Events. And each Event is a single television program, like "Monday Night Football".

Small Summary→ The TRANSPORT STREAM enters our HDTV set. The TRANSPORT STREAM is divided into PROGRAMS. These are not TV programs, like "The Mentalist"; often, these programs are an entire channel. Now previously, MPEG-2, at the TV station, inserted a small table into the TRANSPORT STREAM called the PMT (Program Map Table). The PMT lists every PID number for a PROGRAM, and this allows our sets to reconstruct a channel. MPEG-2 PROGRAMS are divided into EVENTS. An EVENT is what we think of as a single TV program, like "CSI". (Yes, it takes a little getting used to; there are holes in the walls at many TV stations where broadcast engineers banged their heads when first they had to deal with MPEG-2.)

   
TRANSPORT STREAM--->|PID 123 (video)|---> Program (channel) 15000---> |Event CSI  |
         |          |PID 170 (audio)|                ...              |Event NEWS |
         |          |PID 780 (data) |                ...              |Event LENO |
         |                                                            |Event CONAN|
         |------------------------------> Program (channel) 15001     
         |                                           ...
         |                                           ...
         |
         |------------------------------> Program (channel) 15002

♣What's This "PSIP" Thing I've Been Hearing About?♣

"No Man Can Tell What The Future May Bring Forth."   -Demosthenes

OK... So we have a Transport Stream that contains Programs (channels), and each Program contains multiple Elementary Streams. Neat. So how do we know which channels are being broadcast? And how do we know which Elementary Streams belong to which channel? How do we even know which TYPE of Elementary Stream we're broadcasting?

ANSWER→ Ever look into the guts of a computer's "operating system"? Like Windows, or like IBM's MVS? Tables... everywhere you look, there are tables of control blocks. Remember our new, secret Elementary Streams that popped into the Transport Stream kind of out of no where, around the time the old Elementary Streams were going into the Multiplexer? Surprise, they're tables.

This Service Information contains the tables that relate various things in the Transport Stream; they are, in effect, the arrows and links in our diagram just above. And your digital TV set has to be able to read these tables to unravel the Transport Stream.

MPEG-2 is simply chock full of tables, which we mercifully won't examine in great detail... because now, our packets flowing along the Transport Stream... consisting of all Elementary Streams that were combined at the Multiplexer... are about to be dumped onto the "conveyor belt" that will carry them through each exciting step in Exciter-City... The Exciter... the final trip to digital modulation Heaven, and ultimately, up the transmission tower to the digital broadcast antenna, through the ether, and into our digital TV receivers.

So, we have all these tables being repeated on the Transport Stream about once each second, decoded and interpreted and acted upon by our smart digital TV sets. It just so happens that "in a previous lifetime" (so young was he), our very own Dr. Steve worked for a company that made mainframe computers and their operating systems; and Dr. Steve wrote healthy chunks of the code (instructions) that made these operating systems work.

The operating systems were filled with tables; that way, Aardvark Amalgamated and Gerbil General, two different companies, could both use the same operating system to control their mainframe computers (the ones that used to fill a large room), just by filling out the operating system's tables in a way that was customized to the requirements of each. And in those days, much of the stuff that went into the tables lived in a "library" (a type of computer file), and the library often was called Sys1.Parmlib.

Interestingly, the idea of filling a table with data from a library was so appealing that, thirty years later, the designers of digital TV used the same concept to fill the tables that are broadcast to our HDTV sets. And so, in the land of digital TV, we have a thing called the PSIP (The Program And System Information Protocol).

Some say that PSIP is the "bolts that hold together digital television". We'd not go THAT far; but we'd surely say "it's the oil that keeps the logical bearings of digital TV lubricated". PSIP is several small tables; the PSIP data is entered at the digital TV station, and it's decoded in all the digital TV receivers tuned to that station.

PSIP data is transmitted as part of a TV station's digital signal; and it contains important messages for every digital TV receiver tuned to that station... stuff about the station and what's being broadcast. One of the more important reasons for keeping PSIP around is to provide a way that digital receivers can identify a particular digital station and to determine exactly how to tune it in. PSIP identifies both the digital channel and the associated NTSC (analog) channel (at least until June 2009).

Now, recall that we told you that you could receive the weather 24/7 in this area by tuning to 11-2? Recall also that 11 was the channel number for analog WBAL-TV? And recall that we're really on RF channel 59 when we tune to 11-2? Thank PSIP for this convenience... it helps maintain the current channel's "branding"; digital receivers will logically associate channels 59 and 11, making it easier for viewers to tune to WBAL's digital station... even if they don't know the digital RF channel number is 59.

In addition to simplifying channel numbers, PSIP tells the digital receiver if multiple program channels are being broadcast (multicasts) and, if so, how to find them. You fill in the PSIP information at the TV station, it puts entries in a table, the table is sent along with the rest of the Transport Stream, it's modulated and broadcast, and it's decoded by the digital receiver.

PSIP tells the receiver if a program is closed captioned, it conveys V-chip information, it provides any data associated with the program, and much more. It gives the name of each Event (TV program). Of course, if broadcasters don't transmit the proper PSIP information, home receivers well could go bonkers and tune to the wrong station and tell you that your kids are watching "Noggin" when they're really tuned to "Caged Fighting"... and so broadcasters need to be careful in filling out their PSIP tables.

♣Digital TV Exciter♣

"Please Don't Scream."   -Dawn McGatney

REVIEW→ To convert High Definition video from the bit streams coming from digital cameras in the TV studio into something that we can actually broadcast in digital TV format over-the-air, two big steps are employed...

1. MPEG-2.

2. The "EXCITER".

And as we would expect, two big, expensive pieces of equipment are required... the MPEG-2 Encoder and the 8-VSB Exciter.

Interestingly, the exciter increases our bit rate (because of the need for error correction codes and other stuff) to over 30 million bits/ second, before actual modulation and transmission take place. (Sometimes less can be more.)

   
MPEG-II---->FRAME------->DATA----->REED------>DATA----->TRELLIS---->SYNC
 DATA   SYNCHRONIZER  RANDOMIZER  SOLOMON  INTERLEAVER  ENCODER  INSERTIONS
                                  ENCODER                            |
                                                                     |     
                                                                     |
                                                                     |     
     TRANSMISSION<----8-VSB POWER<------ANALOG<-------8-VSB<-------PILOT
        ANTENNA        AMPLIFIERS      UPVERSION     MODULATOR   INSERTION

Ok, here we go, our ride is about to begin→ In ATSC digital TV broadcasting, the MPEG-2 Encoder first takes the video bit stream from the cameras and tape machines and whatever and compresses it and breaks it down into data packets. The MPEG-2 Encoder, using a Multiplexer, combines the compressed video bit stream with the compressed digital Dolby (AC-3) bit stream. Some miscellaneous digital data then is added to the output from the Multiplexer. MPEG-2 then breaks up the combined bit streams into 188 Byte Packets and drops them onto the 8-VSB Exciter's imaginary virtual "conveyor belt".

Note→ Video, audio, and data are "tied up" in separate packets; packets are pure, never mixed. Even the sound for each of the 1-6 Dolby channels is placed in separate packets by the MPEG-2 Encoder/Multiplexer. (We wouldn't want our left-front speaker's sound getting mixed up with our low frequency woofer's sound ("the lease breaker"), would we?)

THE FRAME SYNCHRONIZER→ So how does the 8-VSB Exciter know that a new packet has just plopped onto its assembly line? Like, the 8-VSB Exciter has to locate the start and the end of each data packet. This is done by using the MPEG-2 Sync Byte, the very first byte (a byte is 8 bits) in each packet, as the packets enter The Exciter. This byte is read and thrown out (and so we're now down to 187 bytes per packet); the rejected byte eventually will be replaced by a brand new sync byte, created by The Exciter, during a later stage of processing.

THE DATA RANDOMIZER→ Anyway, and this is so very cool, the video data packets flowing through The Exciter must be random, so that every iota of the 6 MHz bandwidth is used equally. This process is later reversed in the HDTV receiver, in order to recover the proper data values. Is this cool or what? The packets on their way to the modulator are changed to be essentially random (engineers call it "pseudo-random"); and then, in the HDTV set, the numbers are restored to their original values... numbers for luminance, numbers for chrominance... engineers live and breathe this stuff.

REED-SOLOMON ENCODING→ As our packet of data bytes continues to move through The Exciter, it encounter Reed-Solomon Encoding. Reed-Solomon examines all 187 bytes in the packet, and it tacks on another 20 bytes... the famous "Reed-Solomon parity bytes". As the digital TV signal is transmitted, starting at the transmitter, errors in bits can (and do) occur; if the errors are not too severe, the Reed-Solomon parity bytes often can be used to correct them in the digital TV receiver. Note that after Reed-Solomon encoding, the data packet now has grown by 20 bytes to 207 bytes. Reed-Solomon encoding is a method of what engineering-types called Forward Error Correction.

(There is today a vast, elegant body of mathematics for detecting and correcting errors... and ATSC digital TV modulation drinks deeply from that knowledge and comes out with a large handful of error correction... because if you can detect an error and correct that error, then there never was an error... and so it almost seems as if 8-VSB doesn't try as hard to prevent errors as it does to detect and correct them. But that would be silly.)

THE DATA INTERLEAVER→ This is similar to the data randomizer that we thought was so cool. Let's examine the watermelon algorithm→ Is it better if we had placed all of our watermelons on a single train, or if we had decided to carry them in 100 tractor-trailers, should an accident occur? Same algorithm→ The Data Interleaver takes tiny fragments from many different 207 byte packets, spread out across time, and it constructs new 207 bytes packets out of "mix and match". Note that unlike NTSC, the transmission from this point on is NOT in sync with what is happening in the TV studio.

Should there be a burst of noise as a packet is being transmitted from the TV transmitter to our TV receiver, we'll lose several different "tractor-trailers" (we'll lose several pieces of packets), instead of losing the entire "train" (losing a complete packet). The interleaving follows a pattern, and so the packets can be re-assembled in the TV receiver. The data interleaver is a form of what engineers call Time Diversity; fragments are taken from packets across about five milliseconds (.005 seconds), using storage buffers which save the most recent five milliseconds of packets coming along.

Time Diversity is used in digital TV to combat error bursts due to time-varying channel conditions. These error bursts may be caused by fading, in combination with a moving receiver, transmitter or obstacle, or by intermittent electromagnetic interference; e.g., from crosstalk in a cable. Time diversity implies that the same data is transmitted multiple times, or that a redundant error code is added.

THE TRELLIS ENCODER→ As our now flubbitzed data packet continues on its journey through The Exciter, it next encounters the Trellis Encoder... similar to the Reed-Solomon encoder and yet... well, different. (Ok, both are forms of Forward Error Correction.) Along comes our seemingly endless stream of Bits through The Exciter.

The Trellis Encoder breaks every byte (each with eight bits) into four two-bit "words". The encoder examines the "transitions" between two-bit words, and it creates a three-bit code describing the transition. It then throws away the two-bit word... throws away the data... two bits in, three bits out. The original two-bit word is later reconstructed in the TV receiver from the 3-bit "transition code" from the Trellis Encoder. It is these three-bit codes that are actually transmitted over-the-air... the transitions, and not the actual data. Three bits can represent eight values. The 8-VSB modulation allows broadcasting eight "level symbols". The eight level symbols are the three bits of the "transition words". Bingo.

ADDING SYNC SIGNALS→ The next "stop" along The Exciter's "conveyor belt" adds three neat things that will (eventually) help the digital receivers in our homes→

1. The Pilot.

2. The Segment Sync.

3. The Field Sync.

Sounds like Greek, doesn't it? (Only the kitchen sync seems to be missing... sorry.) Nah, simple stuff. Would the ATSC have come up with anything complicated? Never. Besides, it's the 21^st century; we can't have something as simple as the old analog NTSC TV for ATSC digital TV, can we now?

Ok, first, The ATSC Pilot→ Here it's simplest to show in a picture the effect of inserting the pilot. The ATSC pilot is a small "residual carrier" that appears near the zero frequency point of the resulting 6 MHz waveform. (It's actually just 0.31 MHz above the bottom of the channel... much lower in the 6 MHz than the real video carrier in NTSC analog.)

This Pilot gives the circuits in digital receivers something to lock onto... something that is independent of whatever "data" is being transmitted. (Note that the ATSC pilot is also much smaller than the video carrier in NTSC analog broadcasts; the pilot consumes only about 7% of the total transmitted power, but it gives your TV receiver something to "grab onto".)

Inserting the ATSC pilot is like inserting a phony carrier near the left end of the waveform. (Notice also that there is NO color carrier and no sound carrier; this now is 8-VSB digital stuff. So in effect, we're inserting a fake carrier here, so that TV receivers have some idea of what is where, no matter what they are receiving as data or as noise.)

In the old analog TV, we had sync signals rising out of the blackness (75% level), to 100%; if the actual picture were weak and snowy and such, the old analog TV always had the sync signals rising another 25% above the highest possible video signal; the ATSC digital pilot is an attempt to give your digital TV receiver a similar type of "something" above the fray to grab onto, if things get rocky... like when we change the channel.

An Example Of A Typical Digital TV Signal Showing The ATSC Pilot.

Next, second, The Segment Sync→ Remember that first byte of the data packet, coming in from the MPEG-2 Encoder, that we threw away (the "packet sync byte")? We're now going to add back that byte that we threw away. This segment sync that we're adding is a positive-negative-positive pulse swinging between the +5 and -5 signal levels. Circuits in the digital receiver "home in" on the repetitive nature of this segment sync byte (it keeps appearing rapidly), in contrast to the background of "random" data... recall the Data Randomizer step? These segment sync pulses are easy for the receiver to spot, at noise levels where data recovery is impossible. This sync pulse, together with the ATSC pilot, allows the receiver to recover quickly from "lock-up" during channel changes and other "spikes".

Finally, third, The Field Sync→ Ok... a data segment is 208 bytes→ the 188 bytes in an MPEG-2 packet + the 20 bytes from Reed-Solomon. These 208 byte data segments each contain 832 "symbols" (2-bit symbols, 4 symbols per byte).

313 consecutive data segments = One Data Field. After a complete Data Field made up of 313 consecutive data segments, there is a Field Sync Segment... a complete 208 byte segment (or packet). The Field Sync Segment has a unique pattern of positive and negative pulses, and the digital TV receiver can recognize it. The received field sync segment is compared with the Field Sync Segment that we know The Exciter sends; differences are errors. These differences are used to eliminate "ghosts", which were a major problem in the early days of ATSC digital TV. Like Segment Sync Pulses, Field Sync Segments are easy for the receiver to spot, at noise levels where data recovery is impossible. Note that unlike the analog NTSC system, the two ATSC digital sync pulses have nothing to do with the image on the screen.

WRAP IT UP→ Now all that we have to do is to amplitude modulate (AM) our 8-level digital signal onto some intermediate carrier frequency, Nyquist VSB filter the intermediate frequency carrier so that it will fit within 6 MHz (the lower sideband is much more severely filtered here than it is in NTSC analog transmissions), and "up-convert" the intermediate carrier to our final transmitter frequency, amplify it, and send it off to the transmitting antenna.

Make it so.

VERY IMPORTANT STUFF→ Almost everything in a digital receiver is there to reverse and decode stuff that previously happened at the digital TV station... in the exciter and, earlier, during the MPEG-2 compression process(es). We want to emphasize that timing in the receiver is VERY critical, so much so that the receiver cannot even be in motion. (The Doppler effect would change the timing of the incoming signal.)

The receiver has its own clock, a clock that's kept synchronized by the Segment Syncs. This clock tells the receiver EXACTLY when to expect the next Trellis Number. At that exact time, the amplitude of the transmitted signal is sampled and compared against eight possible values (which is what the "8" in 8-VSB means). The closest number is passed on to the trellis decoder, which, using the three received bits, recreates the original two bits of actual data. But timing is everything... because 10,760,000 Trellis Numbers are transmitted each second. And between the proper sampling instants, there is absolute and complete chaos... or put more simply, garbage.

(No big deal... when humans bend over on one foot to pick up a ball, our brains have to solve about one million differential equations per second just to get our muscle co-ordination right. The ATSC digital TV receiver is only slightly more complex than the human brain. In our opinion, ATSC digital lacks the elegance of NTSC analog; complexity is not elegance. We see ATSC digital as a series of problems with a series of patches attempting to correct them. Just our opinion... and the opinion of those who saw the need in 2004 to add E8-VSB as an alternate and more robust modulation method.)

Returning to ATSC digital... AC-3 audio packets go to a decoder in the receiver. Here the packets are divided among the (up to) six audio channels. The bit stream for each channel is converted back to analog, amplified, and sent to the speakers, either within the HDTV set or external.

♣Enhanced 8-VSB (E8-VSB)♣

"Those Who Dance Are Thought Mad By Those Who Do Not Hear The Music."   -Dawn McGatney

Ok, so what happened in 2004? One of the milestones was→ The ATSC standards for digital TV were amended to allow a new, optional mode of over-the-air transmission. This new mode was called Enhanced VSB (E-VSB or E8-VSB). Digital TV transmitted in the E8-VSB mode can use additional forward error correction that allows reception of ATSC digital TV broadcasts under weaker signal conditions. E8-VSB is intended to broaden the application of 8-VSB. The E8-VSB transmission system allows better reception of signals.

E8-VSB was developed in response to the desire of broadcasters to add more flexibility to the ATSC digital TV broadcast standard. This "enhanced" alternative to 8-VSB provides optional modes of operation that allow broadcasters to trade-off data rate for a lower noise threshold of transmissions.

Enhanced 8-VSB allows broadcasters to include a second, lower bit-rate program that is more "robust" than the HD program that is having trouble getting through in low signal conditions. If an HDTV set is having trouble receiving an HD program (e.g., say interference is degrading the HD program), the receiver, if equipped, can switch to a lower resolution SD version of the same program, one that is being broadcast along with the HD version... in other words, the same program in HD and SD are "multiplexed" together in the E8-VSB transmission.

OR... Suppose a receiving location has a permanent problem receiving HD broadcasts from a given station; it can switch to the SD version of the station's broadcast... perhaps the receiver only has an indoor antenna and really requires a roof-top antenna. By switching to the SD program, a pair of "rabbit-ears" might work; the SD broadcast has more potent error correction, although a lower bit-rate per second. (See "No Free Lunch" theorem.)

Other examples of potential applications for E8-VSB include delivery of “fall back” audio, programming services targeted at small digital TV receivers with indoor antennas, non-real time transmissions of file-based information to handheld and pedestrian receivers, and robust data broadcasting to devices such as desktop and laptop computers. Note that the choice is up to the TV station when (and if) it wants to use E8-VSB instead of vanilla 8-VSB.

(Fallback Audio mode allows the sound to continue as normal, even when the received signal isn't good enough for the less robust 8-VSB used with the video stream; and remember, with digital TV, when the picture starts to go, the sound also goes, which can make you want to throw your expensive HDTV set through your expensive picture window.)

For Example→ The broadcaster has 19.39 million bits/ second to play with. Let's say that she allocates 15 million of these bits to an HD program and 4 million to the same program in SD. In other words, a scalable fraction of the main service can be used to transmit more robust, more manly encoded data. To use less bandwidth for this more robust, manly SD transmission, the video and sound are more "tightly" compressed.

The E8-VSB exciter (Remember the exciter? Well, this is the new, improved version.) now has three inputs... normal bit-rate, ½ bit-rate, and ¼ bit-rate. Each packet coming to the enhanced exciter has an ID (a PID value) that directs it to the correct input. Of course, the HDTV receiver must have additional circuitry to handle enhanced VSB.

The cool thing is that the data allocated to the lower bit-rate robust stream can also be used at the receiver to improve the reception of the normal stream. How, dog Wolf? Well, simple. The robust data is used by the normal stream of bits as randomly distributed "training symbols", which can be used in the equalizer to improve the receiver performance where dynamic multipath channels (AKA, "ghosts") are a problem.

The ATSC standards have been updated to include a method to signal the presence of one or more enhancements to the digital signal. In the process of developing the enhanced 8-VSB (E8-VSB) transmission mode, it was recognized that other enhancements and extensions to 8-VSB might be adopted in the future. Since HDTV receivers may not be able to determine which enhancements or extensions are being transmitted at a particular time just through interpretation of the received signal, it is important to provide a mechanism for signaling the presence of enhancements... like 3-d TV perhaps.

It also was important that such a notification mechanism be in place before any receivers were manufactured that included enhancements or extensions, so that the earliest of these receivers could recognize (and perhaps ignore) extensions and enhancements for which they were not equipped.

Thus a VSB Extension is a method of enhancing the functionalities of the ATSC 8-VSB modulation system. The E8-VSB transmission mode is an example of such an extension. The "VSB extension signaling mechanism" facilitates E8-VSB as well as other future 8-VSB enhancements, without creating backward compatibility issues.

(UPDATE→ The problem with the original ATSC standard (A/152) is that its "physical layer" cannot handle moving receivers. A new compatible standard has been developed by the ATSC (ATSC-M/H = ATSC Mobile/ Hand-Held) and now (2010) can be deployed. Of course, ATSC-M/H requires new broadcast equipement (a new "Exciter" that is backward-compatible with existing 8-VSB exciters; think $200,000-$300,000) and brand new "in-motion" receivers. Stay tuned.)

♣Squeezing HD Onto Cable- An Introduction To Digital QAM♣

"Genius Is Another Word For Magic, And The Whole Point Of Magic Is That It Is Inexplicable."   -Margo Fonteyn

Let's begin by reviewing and expanding a little bit on Digital Cable→

Cable can carry HD only in Cable Digital format (also called "QAM"); normally, HD requires some additional compression by the cable provider.

Now, with digital cable, each physical 6 MHz "slot" can carry 7 to 12 SD (480i30) virtual cable channels (programs), if we use 256-QAM modulation and MPEG-2 Compression. Each program is thus transmitted from the head-end to the customer using 3–5 Mbits/sec. (In the US, 256-QAM or 64-QAM are the mandated modulation techniques for digital cable.)

This sounds like we actually might be able to get two HD programs to fit into one physical RF (radio frequency) cable channel. Why does your dog Wolf say this? HD requires six times as many bits/sec as does SD, so we'd expect HD to take up about six times the bandwidth of SD. Thus, if we can accomodate 12 SD programs in a 6 MHz digital cable channel, then 12 programs ÷ 6 times ≈ 2 HD programs.

Now... HD is transmitted using either 64-QAM and 256-QAM modulation. 256-QAM is a bit "tighter" than 64-QAM, but 256-QAM is more prone to bit errors. 256-QAM can accomodate 38.4 Mbit/sec on a 6 MHz cable channel. This is about double the data rate that can be broadcast over-the-air→ 19.39 Mbit/sec, recall?

Thus in one physical 6 MHz channel, cable can carry nearly two full digital 19.39 Mbit/sec HD signals. BUT... The compression used by digital cable degrades the quality of the picture; it is "lossy". Over-the-air broadcasts look better. (See again "Free Lunch Theorem".) Note→ FiOS applies no compression to HD programs.

Cable companies are free to provide non-broadcast programming (Showtime HD, ESPN HD, CNN HD, whatever) in HD or not, as they may decide. They are restricted only by the capacity of their physical cable networks, their infrastructure. A typical cable network contains up to 125 six MHz RF slots.

MPEG-4- The Future?→ US cable providers are preparing new set-top boxes for customers, boxes designed to support new HD channels using MPEG-4 compression. For example, HBO uses MPEG-2 compression to transmit HBO HD and Cinemax HD; all other programming is provided to satellite and cable operators in MPEG-4 format; MPEG-4 is simply a more efficient, more cost-effective system of compression. MPEG-2, however, looks better than MPEG-4; MPEG-4 is "lossier"; when you de-compress MPEG-4, you get back fewer of the bits that you started with, compared to MPEG-2.

Important→ Digital Cable also allows for the broadcast of ED (Enhanced Definition, 480p30), as well as HD (720p60, 1080i30) and perhaps someday 1080p60 ("full HD"). But analog cable can transmit programs in only 480i30 format (SD), the lowest definition video that's broadcast over-the-air in digital.

This is worth repeating→ If you have digital cable (if you have Video On Demand), if you pay extra for a digital set-top box (or a DVR recorder), you can receive HD from your cable provider; but if you have the older analog cable, then it's no HD for you.

Let's assume that you don't rent a digital decoder box, and that you don't have a slot for a CableCARD on your new HDTV set. Let's further assume that your new HDTV set also has a QAM tuner (it's "cable ready" for digital cable) in addition to its ATSC digital tuner; you now can scan through the actual (not the "virtual") cable channels, the physical RF cable channels.

Now let's assume that you select channel 11 on your HDTV set; Comcast, in this area, has put "TV Guide" on channel 11. You look at TV Guide for a moment, and then you press the "Go Up One" button on your remote. Do you go to channel 12?

Not in this case. Your new HDTV set is smart. It has detected a second "service" on physical RF cable channel 11. It names this 2nd service "11-1" and tunes it in. Lo and behold, it's WBAL-DT, broadcasting Leno in HD; and we're using no digital decoder box. You press "Go Up One" yet again. What is this? Is it 24/7 InstaWeather? Yep, your HDTV set has detected still a third "service" on cable channel 11; the TV set names this one "11-2". Finally, we press "Go Up One" one more time, and we go to physical cable channel 12, RF channel 12.

Do we pay our cable provider for "11-1", sometimes an HD cable channel? Nope. Anything that's digital and is not encoded is yours to enjoy at no additional charge. But most non-broadcast programming is encoded. If you want a cable-only channel like "Discovery HD", you'll need a set-top digital decoder box (or a DVR), and you'll need to pay (unless you get lucky, and you find that "Discovery HD" is shoved in as the 3rd "service" on, say, channel 59; and it isn't encoded).

Your HDTV set doesn't know that most folks in our area get Discovery HD on virtual (make-believe) cable channel 230. Your TV simply knows that it found it on physical RF channel 59. The numbers that you see in the TV Guide are all virtual channels.

Now, in the old days, the 2-digit channel that you watched was the actual RF channel. But that went out the window with digital cable; and today, your cable provider may stuff up to a dozen "services" into one RF cable channel, each with a 3-digit virtual channel number that appears on your digital decoder box. But both your set-top box and your new digital TV have the smarts to pull apart all of the services in one RF cable channel, including HD services, and display them for you one at a time.

SO... Some HDTVs can receive both 8-VSB (over-the-air digital broadcasts) and QAM (digital cable); but many cannot. The set-top boxes that you rent from your digital cable provider contain one or two QAM receivers (tuners).

And cable companies usually encode most of their HD content; so you pay a little extra every month for that "HD enabled box" to decode (unblock) the HD content. And regardless of whether your HDTV set can receive QAM or not, the cable companies want a little extra $$ from you before they'll let you watch their encoded HD content. That's why even TVs with QAM receivers need a decoder box (or a CableCARD).

Note→ When cable companies run low on digital decoder boxes, they may give you instead a DVR (Digital Video Recorder) box; but it'll be programmed to decode digital and give you "Video On Demand", etc; it won't be programmed to record shows, not unless you ask for a DVR for additional $$ each month.

♣Cable QAM Modulation♣

"Television Has Proved That People Will Look At Anything Rather Than Each Other."   -Ann Landers

Engineers have created many different kinds of modulation. One of the first modulation techniques was AM. An AM ("Amplitude Modulation") transmitter changes the amplitude of the carrier wave according to "baseband" voice and music, mixed together with the carrier.

Cable providers in the US use a more sophisticated modulation technique called QAM; QAM is used by US digital cable TV facilities. QAM ("Quadrature Amplitude Modulation") works on the same general principles as AM modulation. (In fact, ordinary AM is a type of QAM, a "subset".)

In Summary→ Over-the-air ATSC digital is broadcast using 8-VSB modulation, and cable TV is broadcast along cable using QAM modulation.

Why not just use plain old AM to carry cable TV through the fiber optic and coaxial cable network? Cable has a lot to carry, lots of data on lots of channels, lots of bandwidth. As we said, a cable provider often has as many as 125 physical RF channels, ranging from 54 MHz to 800 MHz.

Luckily, cable signals are strong, and if they weaken, they can be re-amplified; and with cable, errors in transmission are rarely a problem. The goal of the QAM modulation technique is to send as much data, as many physical channels, down the "pipeline" as possible.

It really came down to picking a modulation method for cable TV that allowed as high a data rate as possible. And when it came down to AM vs. QAM, QAM simply could carry more data over a given bandwidth. QAM is used for transmitting digital cable. And, of course, there are various flavors of QAM.

256-QAM is the most common modulation technique among US cable providers today (2010). Each of perhaps 125 physical RF cable channels has its own RF "carrier wave"; and each of these carriers is QAM modulated. Thus a cable Head End could require as many as 125 QAM modulators.

Why 256-QAM? The best answer is to look at Constellation Diagrams. (A little later, we'll take a peek at some constellation diagrams.) If we take a look at a Constellation Diagram for 256-QAM, we can see that there are 256 points on the diagram, and the points are quite close together. This means that the signals have to be strong, with very few errors. But if the points are strong and well-defined (and they are), 256-QAM can carry lots of cable TV data.

Imagine that 256-QAM is a track for runners at the Olympics, with lots of narrow lanes. But our runners are strong and skilled, and they almost never step outside of their narrow lanes. Lots of lanes, lots of runners who don't make errors. That's 256-QAM. QAM doesn't quite double the maximum AM data rate, but it gets closer to the Shannon-Hartley maximum than AM does.

Over the last 100 years, new modulation techniques have been developed; many of these techniques work with all of their might to use as much of that "Shannon-Hartley" (recall them?) maximum theoretical data rate as possible. (No scheme has yet reached 100% of its theoretical Shannon-Hartley maximum, but some are close.)

QAM, instead of simple AM, is used for transmitting cable TV simply because QAM can carry a higher data rate (more runners in narrow lanes) than the "old-fashioned" pure AM modulation, dating from the early 1900's; QAM uses more of the available bandwidth (same width Olympic track, but more runners on the field at once), more of the fiber-optic and coaxial cable pathways that comprise today's cable systems.

SO... What is this QAM modulation anyway? Recall how we sent I and Q color signals in NTSC analog color TV? Recall that we amplitude modulated two RF carrier waves that were 90 degrees out of phase? Recall that we were able to get two signals in effect, I and Q, from a single carrier? The QAM modulation that cable TV uses is very similar to the NTSC analog color modulation technique.

Each of up to 125 cable channels is 6 MHz wide; and any modulation system for cable needs to pack data densely into the smallest possible bandwidth. Unlike satellite, the signals running along the cable network are strong, and errors are rare. Strong signals mean less noise. And so many channels are modulated at the Head End and run along the same fiber-optic and coaxial cable, side by side. But the modulation system for cable does need to worry about inter-channel interference.

Under The Covers→ As with all modulation techniques, QAM sends its data (called by engineers its "baseband signal") by changing some aspect (or aspects) of a radio wave. In QAM, the amplitudes of two radio waves, each 90 degrees out-of-phase (two carrier waves in "quadrature", as engineers would say), are changed. In other words, in QAM, we amplitude modulate (AM) two carrier waves that are 90 degrees out of phase with one another. Just like NTSC color TV.

Because the two QAM carrier waves are separated by 90 degrees, the amplitude modulation of each of the two can be recovered separately at the receiver. How? Well, think back a moment to high school trig→ Multiplying the carrier with a cosine wave (and using low-pass filters to remove higher frequency terms) gives us the in-phase signal (the one called "I" in NTSC color). And following exactly the same procedure, but multiplying instead by a sine wave gives the out of phase signal (the one called "Q" in NTSC color). So QAM allows two sets of information to be amplitude modulated onto one carrier wave, without having the two interfere.

You also can think of QAM as two streams of data coming in, the first one modulating the amplitude of the carrier, the second one modulating the phase of the carrier. (A single carrier being amplitude and phase modulated is identical to two carrier waves 90 degrees apart, each being amplitude modulated; trust us on this, it's simply a "trigonometric identity".)

You may think of QAM correctly in either way; it's just two ways of looking at the same thing, take your pick. The two modulating signals are, in effect, merged in modulating the amplitude and phase of the carrier. And ordinary AM can be throught of as a subset of QAM where the phase variation of the carrier is simply zero.

In digital QAM, the flow of bits is divided into two equal parts, giving us two independent signals to be transmitted. One bit stream is then multiplied by a cosine carrier (the "I" signal), while the other channel is multiplied by a sine carrier (the "Q" signal). In this way, there is always a phase difference of 90° between them (HS trig again). And then the two carriers simply are added and sent through the cable network as one carrier. Q.E.D. (Quite easily done.)

Deeper Into QAM→ Instead of 256-QAM, 64-QAM is often used in modulation for digital cable, especially when it's desired to send a large number of virtual channels (programs) along a modulated carrier.

Incoming 6-bit data words are split into two three-bit "words" (or "symbols"), and each of these three-bit words is used to modulate one of the two quadrature carriers. The the sum of the two quadrature signals is 64 combinations of amplitude and phase. (Three-bits can code eight different symbols→ Recall? Three different Giant SuperMarts, each one open or closed, can be in eight overall states.)

In a QAM receiver, the waveform is sampled twice per cycle, in phase with the two original carriers, so that each sample will represent the amplitude of only one of the two carriers. (The other carrier, at the sample time, is always zero.) A sampler of this kind is effectively a phase sensitive rectifier, and it simultaneously de-modulates and de-multiplexes the QAM signal back into two 8-level signals.

See the similarity between QAM and the color carrier used in NTSC analog television, in which the two color "difference" signals ("I" and "Q") are encoded on in-phase ("I") and quadrature ("Q") carriers? To visualize how 64-QAM works, you can simply replace the analog "I" and "Q" signals of NTSC color with the pair of eight level signals. 8 × 8 = 64.

The result will be 64 possible vectors. 6-bit data packets come in, we split them in half into two three-bit symbols, and we use each of these three-bit symbols to modulate a pair of carrier waves that differ in phase by 90 degrees. Simple stuff, this QAM be.

You may have noticed that we've been assuming all along that the phase of the received signal is known at the receiver, because if the demodulating phase is even a little bit off, it results in crosstalk between the modulated signals (I and\ Q). The phase reference must be known. Recall that in NTSC color TV, we transmitted a "colorburst" so that we could sync up the phase of the transmitter and the phase of the oscillator at the receiver to retrieve I and Q.

What do we do in QAM, dog Wolf? The receiver must somehow synchronize its own local oscillator so that it matches that of the transmitter's at the Head End. This is necessary, because the signal undergoes an unknown time delay in the channel, and it arrives at the receiver with an unknown phase offset with respect to the receiver's own local oscillator. Or as Dr. Steve likes to say at dinner parties→ “The receiver must synchronize to the carrier phases, because the carriers only can be demodulated using coherent demodulation.” (We could not agree more.)

The QAM "receiver" is usually→ The cable provider's set-top box or the TV set itself (in "cable-ready" sets) or a TV set with a slot for a CableCARD (which is usually an HDTV set). The set-top box (or the TV set) connected to the cable provider is the guy who has to sync himself to the QAM carrier phase, on every physical RF channel.

This issue of carrier synchronization at the receiver must be handled somehow in QAM systems. Yes, at the receiver, the two separate signals, "I" and "Q", can be recovered using a circuit called a coherent demodulator; but the coherent demodulator in the receiver really needs to be exactly in phase with the received signal, or otherwise, the two modulated signals cannot be independently received.

Clearly, this is a problem that must be solved whenever QAM is used. For example, in the NTSC analog color television systems with I and Q, a colorburst of 8+ cycles is transmitted from the TV station for 2.5 microseconds and inserted after each horizontal sync pulse (remember him?) at the start of each line, for phase reference.

Even Worse→ This sync between the transmitter and the de-modulator in the receiver is also a potential problem with "QPSK" modulation that is used in satellite TV broadcasting, so with both cable and satellite TV at stake, we probably should come up with an answer to how the receiver knows the phase of the carrier back at the transmitter.

The Solution→ For analog QAM modulation, QAM transmits an additional component of the carrier along with "I" and "Q"; and from this additional carrier component, the timing and phase can be recovered in the receiver.

In digital QAM, this additional carrier component is not necessary; periodic "sync" signals carry the necessary phase information. These signals are easily recognized by the receiver.

Got this now? In analog QAM, there is in effect a third carrier sent along with I and Q, which the receiver (set-top box or TV set) uses to sync itself with the transmitter at the Head End. And in digital QAM, special sync signals are sent peiodically to sync the receiver with the transmitter. These are QAM's version of the colorburst; both have the same need, to sync the de-modulator in the receiver with the modulator in the transmitter. And both work.

In a typical cable TV application, the data is randomized before being fed to the modulator. The use of randomizing (and error correction) is not part of QAM modulation. (But it sure sounds like stuff that happens in the "Exciter" for over-the-air broadcasts.)

In other words, real-life systems need additional processes. The error correction coding is designed to overcome a reasonable bit error rate, so that the transmitted data is essentially error free. This is important for digital cable television (modulated by digital QAM), because MPEG compressed video data is more sensitive to bit errors.

A Topic Revisited→ Many HDTV sets contain an integrated QAM tuner; this allows the reception of digital programming sent "in the clear" (unscrambled) by cable providers at no charge, usually your local broadcast stations. However, most digital channels are scrambled, because the cable providers consider them to be extra-cost options and not part of the "basic cable" package.

Exactly which channels are scrambled varies by location, and can also change over time. In the US, a television that is labeled "digital cable ready" can have a CableCARD installed by the cable provider to unscramble the protected channels, allowing subscribers to receive all authorized digital channels without using a set-top box.

Interestingly, the availability of this QAM HD digital programming is rarely described or publicized in cable company product literature. These QAM "services" are not included in guide information on devices like TiVo DVRs, and can be unexpectedly moved from RF channel to RF channel.

This makes watching QAM channels frustrating for the casual viewer, encouraging them to purchase a "digital cable package", which includes a set top box and access to TV Guide data. Other folks enjoy the challenge of hunting for QAM services; such folks often use scanning devices that find "things" hidden in physical RF cable channels. Whatever.

You need a different tuner in your HDTV set for QAM signals than you need to receive 8-VSB over-the-air transmissions, although many TVs include tuners that can tune both 8-VSB and QAM, and demodulate both. The electronics required is quite similar for the two. The set-top box provided by the cable company for digital cable contains a decoder to unscramble encoded digital programs in addition to one or two QAM de-modulators. ¿Está claro? Great.

QAM And Set-Top Boxes And DVR's

♣VERIZON FiOS TV - The New Kid On The Block♣

"When One Of Your Dreams Comes True, You Begin To Look At The Others More Carefully."   -Dawn McGatney

Verizon's FiOS (short for "Fiber Optic Service") is a TV provider (and a phone and an Internet service) that is offered to about 10 million households in the US. FiOS has grown rapidly; as of May 2009, television subscribers totaled 2.2 million households, including 300,000 households added in first quarter 2009.

So what's this FiOS TV anyway? Yes, we know, fiber-optic cable; but aren't most large cable systems fiber-optic? What makes FiOS different?

Answer→ True, a lot of cable providers rely on fiber-optic cables. The digital QAM (that you now know all about) usually runs along fiber-optic cables, from the Head End (where the cable provider collects all its programs, encrypts a lot of them, assigns virtual channel numbers, etc) to the neighborhoods it serves, and then blat; the fiber connects to metal coaxial cables to traverse the streets of our neighborhoods.

Fios is similar to cable in that it employs QAM modulation; and it's like satellite in that it's 100% digital. But when traversing the streets of our neighborhoods, FiOS remains fiber-optic. FiOS employs a fiber-optic delivery system called FTTP ("Fiber To the Premises"). FiOS runs from the street in front of customers' homes to the box attached on the outside wall of their homes still using fiber-optic cable.

As a consequence, FiOS has a wide bandwidth. FiOS does not need to compress its TV channels. And so FiOS TV pictures look better. We watched programming from FiOS on a friend's 42" HDTV, and we thought we were watching HD; but it was actually ABC at 480i30 (SD). And FiOS is fast; and so it makes for a good connection to the Internet.

Now, fiber-optic cable is not new; fiber-optic cable has been around for decades. But only in the last few years has Verizon has been running fiber directly to customers' doorsteps. Verizon (and other telephone companies) have been replacing that "last mile" of copper cable running to homes with superfast fiber-optic cable, finally bridging the gap between homes and the fiber cable that runs to neighborhoods. Fiber-optic cable has the potential not to be just a little faster than metal coaxial cable, but hundreds of times faster.

Mile for mile, fiber is cheaper than an equivalent length of metal coaxial cable. Optical fibers are thinner than copper wires, and so more fibers can be bundled into a cable of a given diameter. This allows more signals to go over a fiber cable of the same size as coax; and it permits more channels to come through the cable going into your cable TV box.

Unlike electrical signals in copper wires, light signals running along fiber do not interfere with signals in other fibers in the same cable, meaning clearer TV reception. Currently, FiOS offers more than 100 uncompressed HD channels.

A fiber-optic cable is made up of thousands of fiber-optic strands of nearly pure glass, bundled together, each about the width of a human hair. Each fiber strand is composed of a glass core through which light travels. And surrounding each glass core is "cladding", an optical material that reflects light back into the core whenever it hits the edge of the core at the cladding; the light can't escape. Light in a fiber-optic strand travels through the core, constantly bouncing off the cladding, using a principle called total internal reflection.

There are single-mode fibers and multi-mode fibers. Single-mode fibers have 9 micron cores (a micron is about 1/25,000th of an inch), and they carry infrared laser light with a wavelength of 1300 to 1550 nanometers (a nanometer (nm) is a millionth of a millimeter). Multi-mode fibers have 63 micron cores and carry shorter infrared light from LEDs (light emitting diodes) with a wavelength of 850 to 1300 nm. (Visible light is between 400 and 700 nm.)

Fios runs Single-Mode fibers to customers' homes.

The cladding surrounding the glass core absorbs no light, and so a fiber-optic strand can carry light quite a distance. Some loss does occur when the light is transmitted through the fiber's glass core, due to impurities in the glass, especially over distances of more than half a mile. And so, one or more optical regenerators is spliced along the cable to boost weakened light signals.

An optical regenerator consists of optical fibers with a special coating (doping). This doped portion is "pumped" by a laser. When the weakened signal enters the glass with the doped coating, the energy from the laser allows the doped molecules to become, in effect, their own tiny lasers. The doped molecules emit a new, stronger light signal, with exactly the same outgoing light signal as the incoming weaker light signal. In other words, the optical regenerator acts as a laser amplifier for the incoming light signal.

Relatively few optical regenrators are required for FiOS optical cables, even over long distances. In early fiber-optic cable systems, the number of amplifiers between the Head End and the customer was around six, compared to 30 or 40 in coaxial cable systems. In newer fiber-optic systems (those constructed after 1988), the number of fiber-optic regenerators has been reduced further, to the point that today (2010) only one or two amplifiers are required to provide most homes with a strong signal.

Verizon FiOS TV is delivered over a "fiber to the premises" (FTTP) network, using passive optical network (PON) technology. Phone, TV, and Internet data travel over three different wavelengths of light, all three in the infrared spectrum (infrared is light waves that are too long, too "red", to be visible to the human eye).

Passive Optical Network? HUH?

A "Passive Optical Network" is one which does not use electrically powered components to split a light signal. Instead, the light signal is simply distributed using beam splitters. Each beam splitter can split a fiber into 16, 32, or 64 other fibers. "No Power" means that a beam splitter can't provide any switching or buffering. The resulting connection is called a point-to-multipoint link. (Meaning→ One point, the Head End, is linked to many points, lots of homes.)

Since we have no switching capabilities in a PON, each signal leaving the Head End must be broadcast to every user served by that splitter (including those users for which the signal is not intended). It's therefore up to the Optical Network Terminal (ONT) on the outside wall of the subscriber's home to filter out any signals intended for other customers.

And since Beam Splitters can't do buffering, each individual Optical Network Terminal must be time-division multiplexed to prevent signals that are leaving multiple customers' homes from "colliding at an intersection". Thus, customers must "take turns" transmitting information.

The "Passive Optical Network" employed by FiOS has advantages and disadvantages. They avoid the problems involved in keeping electronic equipment outdoors. And they can simplify the delivery of analog television. However, because each signal must be pushed out to everyone who is served by the splitter (rather than just to a switching device), the central office must be equipped with a powerful transmitter, a transmitter called an Optical Line Terminal (OLT).

And because each customer's Optical Network Terminal (ONT) must transmit all the way back to the central office (rather than simply to the nearest switching device), customers can't be as far from the central office as they could be using active optical networks.

Once Again, Please→ To serve a home with FiOS, a single-mode optical fiber extends from an (OLT), located at a FiOS Central Office or Head End, to each neighborhood serviced, where an Optical Splitter fans out the same signal on up to 32 fiber-optic cables, thus serving up to 32 FiOS subscribers. At the subscriber's home, an Optical Network Terminal (ONT) transfers data onto the copper wires inside the home for phone, TV, and Internet access. How?

One of three wavelengths is dedicated to carrying television channels, channels that are compatible with ordinary cable television equipment. The other two wavelengths are devoted to carrying data, one for outbound data, the other for inbound data.

1. 1550 nm for TV video with 870 MHz of bandwidth.

2. 1490 nm for downstream data at 622 Mbit/sec (2.4 Gbit/s with GPON (A Gigabit Passive Optical Network) ).

3. 1310 nm for upstream data at 155 Mbit/sec (1.2 Gbit/s with GPON).

Note→ Verizon's TV service is not video over IP (Internet). Well, Ok... Video On Demand and interactive FiOS features such as "Widgets" and Programing Guide Data are delivered over IP (the Internet). But most TV content, including Pay Per View (PPV), is provided over the video signal, which carries digital content up to 870 MHz in bandwidth. (As of 2009, all FiOS is 100% digital.)

The FiOS broadcast, with digital QAM channels, originates from a traditional Cable Head End and travels over various SONET networks, until it eventually arrives at a Local Serving Office. ("SONET" is "Synchronous Optical Networking", a multiplexing protocol for transferring multiple digital bit streams over the same fiber-optic cable.)

The FiOS RF signal occupies 870 MHz, and it is modulated onto the 1550 nm infrared wavelength light beam. The optical Video signal at 1550 nm is then combined with the IP Data signal at 1490 nm via the use of a Wave Division Multiplexer (WDM) and is sent out to the PON (to the Passive Optical Network, no active components, recall?).

The Multiplexer also directs the incoming 1310 nm upstream data from the ONT (the Optical Network Terminal... the guy who filters out other customers from entering your home) back to the OLT (the Optical Line Terminal... recall, the powerful transmitter that we need at the Head End.)

At the ONT, often located on an exterior wall at the subscriber's home, the RF video is transformed from light on a fiber optic cable to radio frequency sent over a coaxial cable into the home... and then most commonly to a FiOS Hybrid Set-top-box that handles both TV RF and IPTV (TV over Internet) video. (No, this is surely not our typical cable TV arrangement.)

FiOS TV from Verizon uses a new Home Networking technology based on MoCA (Multimedia over Coax Alliance), a standard defined by a consortium of companies. (The Multimedia over Coax Alliance develops specifications for home networking over coaxial cable; such cable is commonly used for antenna connections to TVs and radios, and for cable TV.)

This MoCA technology converts regular Ethernet traffic to a signal that can traverse the coax wiring already in place in many homes, without affecting Cable, Satellite, or other FiOS TV signals. Multiple Nodes/Devices can be hooked to the same coaxial cable, and they can communicate on the shared medium.

In order to be able to make multiple devices communicate over the same medium and maintain Quality Of Service by avoiding packet collisions, one of the nodes elects himself as "boss" of the network and tells everybody when they can transmit and when they can't. In order to increase the throughput of the network, each node can talk with the others directly, avoiding wasting bandwidth by going back and forth to/from a centralized access point.

A 50 MHz portion of the spectrum in the 950 MHz (MoCA WAN) and 1150 MHz (MoCA LAN) vicinity is reserved for communication among the different MoCA nodes. The 50 MHz channel is divided into 256 sub-channels. Each sub-channel communicates using the best modulation possible, starting from BPSK (low data, used when errors are possible) up to 256-QAM (strong, error-free signals, lots of data). Every node will negotiate the best modulation profile that can be achieved, while maintaining a low Packet error rate for each of the nodes.

In order to effectively Multicast or BroadCast packets destined to more than one node, the lowest common denominator profile is used, so that all the nodes can receive the communication correctly.

Up to 8 networks (50 MHz each) can co-exist on the same wiring infrastructure without interference. The result is a reliable backbone for Home Networking, which at the current state of the art, can support a total throughput of around 150 Mbps.

Verizon uses ordinary Cable TV channels to broadcast TV to all users of FiOS TV which are hooked to its Fiber Network. The architecture for sending broadcast TV works similarly to current Hybrid Fiber Coaxial Cable networks, but the fiber reaches the home of the end user; and the signal is converted back to coax directly inside the home.

So, why does Verizon use MoCA technology in the home?

While Broadcast Television does not need to be transported over Ethernet, Video on Demand requires both→

1. The ability for a user to request a particular program.

2. The ability for Verizon to send a program to a specific user. So, for VOD, Verizon uses this backbone Home Networking technology to carry the traffic associated with the VOD request and the VOD stream itself.

And what are the advantages of this technology?

1. Coaxial Cables are often already installed in the Home.

2. Coaxial Cables are already used to carry the TV signal to the TV sets and Set Top Boxes in the home.

3. Precise Networking Solutions loves to install coaxial cable in folks' homes.

Verizon originally had planned to use MoCA technology for the delivery of LAN traffic to all the Set Top Boxes in the house, but now two MoCA networks can be established in a FiOS installation in order to minimize installation costs by reusing existing wiring as much as possible.

The result is that a MoCA domain is established between the Router and the different Set Top Boxes (MoCA LAN), and another domain (MoCA WAN) is established between the router and the ONT (which is the box that converts the optical signal coming in the home from light on fiber to regular Ethernet signals on coaxial cable.)

And that's FiOS.

♣QPSK Modulation And Satellite TV♣

"Imagine What It Would Be Like If TV Actually Were Good.
It Would Be The End Of Everything We Know."   -Marvin Minsky, MIT

As We Said→ Direct Broadcast Satellite (DBS) TV transmissions, from satellites in space down to the small 18" dishes that DirecTV and DISH provide, are 100% digital and have been for some time.

Now, US Satellite TV transmissions employ various flavors of Phase Shift Keying (PSK) modulation. And QPSK (Please Don't Scream... Quadrature Phase Shift Keying) is the most commonly used flavor of PSK.

Why, dog Wolf?

Well, noise is the enemy of satellite transmissions (these transmissions from outer space are very weak), and QPSK modulation works fairly well, because it's highly noise resistant.

Let's return to our analogy of a track at the Olympics. Unlike cable, where we had many, many skilled runners (lots of data), who stayed in their lanes (few errors), for satellite, we need wide lanes for relatively few runners. The "satellite runners" are prone to errors... they are notorious for sliding into the lanes of adjacent runners... and so we need a small number of wide lanes. QPSK modulation provides a small number of wide lanes.

So, summing our TV modulation techniques, we now have, to date→

• NTSC VSB for over-the-air analog TV broadcasts.

• ATSC (E)8-VSB for over-the-air digital TV broadcasts.

• QAM for digital cable/ FiOS TV broadcasts.

• QPSK for digital satellite TV broadcasts.

(And the winner is...) Yes, there is a lot of TV modulation going on in the US (and in outer space).

Cypher→ We're going to kill QPSK, do you understand that?
Trinity→ Morpheus believes that QPSK is The One.
Cypher→ Do you? I'm going nuts with all of these modulation techniques.
Trinity→ It doesn't matter what I believe. The Oracle believes in QPSK.
          . . . . .
Morpheus→ Quadrature Phase Shift Keying is The One.

Morpheus, Neo, And Analog TV

(Sorry, we forgot to use QPSK modulation here, and we had cross-channel errors between HD and "The Matrix".)

First let's simply define phase shift modulation, and then let's see if there is some common ground... some way that we might relate all of the digital modulation techniques to the amount of data that they can send... and to their error rates.

QPSK is a modulation technique that doesn't change either the amplitude (AM) or the frequency (FM) of a carrier wave; instead, QPSK changes the phase. QPSK shifts the phase of a carrier.

A Phase Shift

The two waves above both have the same amplitude; and they both have the same frequency. But the blue wave has shifted its phase by 45º, relative to the red reference wave. Now imagine the blue wave shifting farther and farther to the right, relative to the red reference wave, possibly in response to bits of music or bits in a TV video signal. That would be digital phase shift modulation. The phase of the carrier wave shifts, relative to some reference, in response to "baseband" information in the form of bit patterns. It is this phase shift that carries information, that carries our digital satellite TV.

Now, The Unified Field Theory→

There is a kind of diagram, one that we mentioned briefly, that is used in digital broadcast engineering; it's called a Constellation Diagram. It displays a signal that has been modulated by some digital modulation technique. It is a plane, with an x- and a y-axis. And it has small circles on it. Each small circle represents a possible symbol; that is, some legal pattern of bits, that may be retrieved at the instant a receiver samples our modulated carrier wave.

The name of a digital modulation technique usually begins with a number→ 8-VSB, 64-QAM, 256-QAM, etc. In "8-VSB", the 8 simply means that the sampling at the receiver might retrieve one of 8 possible symbols... one of 8 possible bit patterns; and anything other than these 8 would be an error.

Any digital modulation scheme uses a finite number of distinct signals (called "symbols") to represent digital data. For example, QPSK uses a finite number of phases, and each of these has a unique pattern of bits.

Recall that analog TV receivers can legally retrieve almost anything between 12% and 100% signal strength, and it's (potentially) Kosher. But digital can retrieve only a limited number of bit patterns. Digital signals are in a finite number of discrete states. And because of this limitation, a digital receiver can better tell the real data from the noise.

8-PSK Constellation Digram

The diagram above shows a carrier with 8-PSK modulation. The axes are labeled with our old friends, I and Q. The small circles are very sharp, very well-defined, meaning very few errors. And we get a different 3-bit symbol depending on the phase shift of the carrier. The possible legitimate phases are each 45º apart. The receiver selects one of these possible 8 symbols as its estimate of what was actually transmitted; in other words, the receiver selects a small circle on the Constellation Diagram, the one which is closest to that of the received symbol.

Eight possible symbols is not a large amount of data, but when errors are a problem, as in satellite reception, we want our little circles to be far apart. Our Olympic track here has eight wide lanes. This is one modulation technique that can be used by satellite TV.

Notice that the if we were to start increasing the possible number of "Kosher" phases, two things would occur→

1.) We could send more data.

2.) If the "circles" we received became more fuzzy (there were noise with the data), the digital receiver would be less likely to "guess" the correct circle (the correct symbol, the correct bit pattern). The fuzzier the circle, the less likely that the receiver will make a correct guess.

Now take a look at this Constellation Diagram→

BPSK

Even though a Constellation Diagram is a plane with an I axis and a Q axis, the little circles may not have both of these co-ordinates. The above modulation method is called BPSK, and we merely present it as an example. We didn't do any fancy I and Q modulation here, like NTSC analog color TV or QAM does (no quadrature stuff). And so even though we still display our little circles on a plane for our BPSK Constellation Diagram, the Q-axis (the vertical "y-axis") doesn't convey any information.

We can see from this diagram that→

1. We're sending a lot less data (we're receiving only a "0" or a "1" at any sampling interval).

2. If our circles are very fuzzy, our receiver still can make a good guess as to what the transmitter intended. (And many circuits in the receiver can be simpler, because the receiver doesn't have to separate the I and Q components.)

BPSK (Binary Phase Shift Keying)→ Because of its simplicity, BPSK is used in low-cost transmitters, and it is used in transmitting credit card data, such as American Express' ExpressPay, verification of credit cards at gasoline (petrol) pumps, and many other passive applications.

8-VSB (used for digital TV over-the-air) might look similar to BPSK, but with four small circles on each side of the Q-axis; again, with 8-VSB there is no Q component, simplifying circuits in our HDTV sets a bit.

8-VSB For Engineers→ In spite of our comment above, 8-VSB modulation (the guy used to modulate digital TV over-the-air) actually does have both an I and a Q component. However, there is no additional information of any value in the Q component, so the HDTV set doesn't have to bother separating I and Q.

Because 8-VSB modulation essentially transmits only the upper sideband, we have, in effect, added together two carriers that are offset by 90° (a sine and a cosine). The cosine carrier is modulated by the I signal. The I components also are shifted by 90° to modulate a sine carrier, producing the Q signal. When these cosine and sine carriers are combined, one of the sidebands is eliminated through phase cancellation, and we have achieved vestigial sideband modulation (VSB), and the resultant signal does contain both I and Q components. The following modified Constellation Diagram shows an actual 8-VSB transmission over an interval of time→

Actual 8-VSB Diagram

The vertical lines on the modified Constellation Diagram represent the eight legal amplitude levels in 8-VSB. If our signal falls anywhere along one of these eight vertical lines, it indicates that the I component is equal to the corresponding 8-VSB symbol. And we never worry about the Q values; they have no additional information. (Eight moderately wide lanes on our Olympic track, some potential errors by our runners.)

Ok, what does a QPSK (used by satellite) Constellation Diagram look like?→

QPSK (or "4-PSK")

In QPSK (4-PSK), the carrier can be in one of four phase states... 45, 135, 225, and 315 degrees shifted from the reference phase (the four phase shifts are each 90º apart). Four wide lanes on our Olympic track.

OK... Not bad. It carries a reasonable amount of data; and there is a decent separation between the phases (in this case, phase again is plotted on a circle), so that if things get noisy (and we have weak signals coming in from space), we can usually make a correct guess at the intended bit pattern at the receiving dish. And since we use both the I and Q axis for our phase modulations (plural), the receiver needs to be more complex to separate I and Q (just as our old color TVs had to be a bit more complex to separate I and Q). ¿Está claro?

And finally, just for the sake of completeness, let's look at a constellation diagram for QAM (used by FiOS and digital cable). Here is just 16-QAM, but you will get the picture; lots of data can be transmitted because we have a low error rate→

16-QAM

TIME NOW TO LOOK AT SOME ACTUAL
DIGITAL TV SETS

The Holiday Season, December 2009→ Installing HDTV sets and Home-Theatre sound systems and seating, along with Blu-ray Disc Players, occupies a major part of Precise Networking Solutions. We didn't plan it that way, it simply happened; the word spread. (PNS is currently (1 February 2010) booked to roughly 1 May for projection systems with "The Classic" sound system and five seat theatres with full home networking. (DAwn is actually on the waiting list for an HDTV set in her condo.)  )

At the bottom of