The Digital Television Revolution

The 2012 London Olympics really brought home to me just what a massive technological jump in digital media has occurred during the last few years. There has been significant advances in digital compression and transmission.

This year, in addition to high definition broadcast, which made its appearance in the 2008 Beijing games, 3D television was also added to the line up, offering more channels and choices. With analogue television broadcast almost becoming extinct, digital televisions promise of delivering more for less has become a reality. Now, how did we arrive at this point and what does the future hold for digital multimedia?

Prior to the digital switchover, analogue television was resource hungry in terms of the amount of bandwidth required to carry a single channel. This is typically between 6 – 8 MHz depending on the type of video standard being used. This limited the number of channels which could be transmitted, since there is a finite amount of spectrum that must be shared with other services such as mobile, radio and two way communications.

What the digital standards of ATSC (North America) and DVB (Rest of the World) provided was the ability to reuse the existing analogue spectrum more efficiently. This meant a typical 8 MHz carrier used for analogue broadcast could be converted to DVB-T (Digital Video – Terrestrial) making it possible to carry 9 standard definition channels or 3 HD channels plus one SD channel for the same amount of bandwidth.

It would have required in excess of 70 MHz of frequency spectrum to achieve this with the old analogue standard. In addition to squeezing more channels into less space, digital television is much clearer and doesn’t suffer from ghosting or other artifacts which troubled analogue systems. Being digital also allows other features like improved digital sound, electronic program guide and subtitle support to be included.

Televisions are sold with the digital decoder integrated and older televisions can use a separate set top box. As technology advances, we will also see improvements in the compression techniques used, which means even more content for digital media, already this has enabled 3D broadcasts for some events such as the Olympics.

The Future

Eventually as fibre to the home is deployed worldwide, the all IP enabled set top box will replace the DVB standard, since the IP set top box has a distinct advantage over digital broadcast technologies, specifically multicast join requests. Unlike DVB-T or DVB-S, IP multicast allows the receiver to send a join message to the network for the desired channel then if the request is successful the broadcast is routed to the receiver, only the bandwidth for the requested channel is used. With the DVB standard, all available channels are being broadcast simultaneously, and the channel count is limited by the finite amount of channel bandwidth regardless of the compression techniques being used.

The IP set top box can support both selective multicast (one to many) and on demand unicast (one to one) broadcast, this allows for virtually unlimited amount of content. However, unlike DVB, IP set top boxes have to worry about latency and QOS, since there is traffic contention with both residential broadband and IP Telephony. A poorly implemented IPTV deployment can behave like analogue television in an over subscribed service provider network, unless the correct traffic management is in place.

High Definition

Today HD is regarded as premium content by most operators and is charged at a higher rate than SD (Standard Definition). However, over time this will change as people upgrade their televisions to HD models. Today there are two standards for digital HD broadcast, 720p and 1080i. The ‘p’ means progressive and the ‘i’ means interlaced. In 720p broadcast, the picture is made of 720 horizontal scan lines and a vertical resolution of 1280 pixels, which has the advantage that one frame represents a complete image.

In 1080i broadcast, the picture is made from two 540 horizontal scan images which when combined make 1080 lines. The vertical resolution becomes 1920 pixels. Most modern televisions support playback of 1080p, which is definitely more desirable than 1080i especially in fast moving sequences where motion blur can be experienced. However, on modern televisions the difference is barely discernible.

Initially the public uptake of HD was slow, the receivers were expensive and the available content was limited. HD television has really been an evolution rather than a revolutionary change for most of us and this is also true of digital television in general. As digital switchover continues worldwide and consumers replace their televisions, digital will become the new standard. However, it is unfortunate that technology won’t help to improve the content.

The Art of Selling a Spectrum Game

Let’s face it, graphics were not the selling point of a ZX Spectrum game. It wasn’t often someone would pick up a cassette box and shout out “WOW, look at the graphics on this game!!” – Spectrum gamers knew what kind of graphics they were probably going to get even before flipping to the back of the box.

What made it worse was that a lot of the times on the back of a box the publishers had provided screenshots of not only the Spectrum version, but alongside them screenshots from the rival Commodore 64 version, and even the Atari ST and Amiga versions which were streets ahead with graphics capabilities. Some cassette inlay’s took it a step further with a complete disregard for false advertising as they showed screenshots from a completely different system (one of the ones with the much better graphics) and decided not to show any Speccy screenshots at all! Admittedly, there were times I would look at these comparison screenshots and think “Why can’t my game look like *that*”. I’d still buy the game anyway, because I knew what to expect and of course I could always use my imagination to make the game better. No matter what version of screenshots I was shown, I had a feeling of what was going to be fun. But what made the Spectrum owner pick up the box in the first place?

In a time without YouTube or the internet, and television advertising for games was unheard of; it was the cover art that had to grab your attention. Yes there were Spectrum magazines filled with screenshots and reviews but when you turned the page to reveal a full page colour advert for a game, it was dominated by incredible game cover art, and only a few small screenshots of the game (if any at all) usually subtly placed at the bottom with the other unimportant stuff.

When I’m talking cover art, this was not computer designed 3D CGI at the standard seen these days; these were beautifully drawn or hand painted – this was real talent, and time and effort spent – nothing computer aided or digital. In some cases you could see the felt tip pen strokes, brush marks or pencil lines. This was real art. Walking in to a computer shop and looking across the shelves at a sea of cassette boxes, each one with their own cartoon cover, painted hero scenes, or movie poster style art – you knew you were in for a treat, even if the treat was the time you spent in the shop looking at them. There were titles you had never heard of, title’s that didn’t even show a single screenshot on the back of the box! But this added a mystique to the choice of this week’s game purchase. Even without screenshots, the cover art told you it was worth taking the gamble as you stared at the picture on the front of the box on your bus journey home (..sometimes the gamble didn’t always pay off, though.)

These sometimes breathtaking illustrations would pull you in, and they tempted you. Like the art on the cover of a book, you wanted to open the pages and dive in to the story to be the character emblazoned on the front; the cover set the tone for the incredible adventure you were about to embark on… which of course ended up being a number of basic looking pixelated shapes awkwardly moving around a screen to the soundtrack of a few bleeps and white noise, but that’s not the point.

Today graphic artists could simply take a frame from the photo-real texture mapped game sprite and place them in any position or pose, and that alone would be enough to sell the game. However, in the days of the Spectrum, in it’s place would stand an actor in action poses dressed up in full costume as characters from the game! I, of course, refer to the very memorable cover of “Barbarian”. It gave it an extra dimension of realism to the point of sale rarely seen today – oh, and boobs. Protesters focused so much on the risqué (although not by today’s standards) cover art, that nobody pointed out that in the game you chop peoples head’s clean off with a sword, for it to be then kicked across the screen! To be fair, the type of person to make complaints about a girl wearing a bikini on the front of a computer game box, probably didn’t know how to load the game up in order to be outraged by the beheading.

Grand gestures and attention grabbers were needed in the early days of computing, of course this was mostly to counter the incredibly unrealistic game and sometimes pitiful gameplay of a title – usually the movie-licenced ones, to be fair.

If a movie was a big hit, any kind of game of any standard would do – sometimes with no actual relevance to the plot of the movie, and forget screenshots – not needed! 
Get the license to publish a game of the worldwide epic movie “Jaws”, put the famous Shark on the front emerging up toward the swimming girl; then it’s going to shift a considerable amount of units. Oh wait, what about the game? OK just swap the X’s and O’s for Shark Fin’s and Girls Face’s in a game of sharky Tic-Tac-Toe – that should do it! (That wasn’t the game version of Jaws, by the way, I just made that up for an extreme example – the actual game was *much* less relevant to the plot). The point being, as long as it had the big Hollywood cover art, then it was going to sell by the bucket load no matter what. Gamer’s did feel let down however, and through the years would become wise and double check the screenshots and reviews of movie-licensed games, just to make sure they weren’t being conned.

There were good games, and bad games, correct screenshots, deceiving ones, and no screenshots at all; but one thing was certain when you purchased a Spectrum game – you were going to have a new experience (good or bad) that started the moment you set eyes on the cover art.

Through the Looking Glass, Our Vanishing Spectrum – Part II

In the last blog post we identified the looming potential problem of total consumption of the radio frequency spectrum. If we want to head this off, we need to begin with some guidance about how we should rationally allocate and use the spectrum — for what sorts of things should it be used, for what other sorts of activities should it not be used? What should we be doing with it today and what should we not be doing?

We will start with a living chap who, although not particularly well known as such, may be one of the most significant telecommunications visionaries that this country has produced. The curious thing is that he is not an engineer or a scientist; he was trained as an architect. His name is Nicholas Negroponte. (Yes, you have heard that family name previously: older brother John Negroponte was a US ambassador to the United Nations). Today, Nicholas is one of the guiding forces behind the “One Laptop per Child” movement which seeks to provide a $100 (maximum cost) functioning laptop computer to each school child in Third World nations.

Nicholas has held a number of significant assignments, but the one that will interest us here is his serving as a founder and the first Director of MIT’s Media Laboratory, back in the 1980s and 1990s. The Media Lab was (and still is) a place where many new and revolutionary ideas about the mass media and telecommunications were/are proposed, debated, and in some cases reduced to practice. A book by Stuart Brand with the name “The Media Lab” documents much of the early history of the place, and although now dated it is still worth a reading. You may find surprising the percentage of the Lab’s predictions from those early days of communications and computing that have now come into commonplace use.

Among many other novel ideas flowing from Negroponte and the Media Lab, he proposed a fundamental communications concept which, to the Curmudgeon’s way of thinking, is simply brilliant in its simplicity and universality. In just a few words it lays out the basic guidance for the rational uses to which the radio frequency spectrum ought to be put. By rights it should be known as Negroponte’s Law, although the Curmudgeon is not certain that this identification has been designated. But it has been called the Negroponte Switch (switch as in “change-over,” not switch as in “make and break circuit element”).

At the fundamental, bedrock level, as paraphrased here, Negroponte teaches:

Telecommunications services which are FIXED in nature (i.e., where the receiver is at a stationary location) should preferentially be transmitted by wired circuits; telecommunications services which are MOBILE in nature need to be transmitted by wireless circuits (i.e., by radio).

What an elegant idea! It’s one of those simple little concepts for which it might be said, “anyone could have thought of it!” And nobody except Negroponte actually did.

It says that, in a world of limited spectral resources, let’s use the radio spectrum for transmission only where and when we need it, and avoid using it where other choices are available. That principle is, truly, the Ursprung of spectrum conservation!

So, as Negroponte did, let’s apply it toward rationalizing today’s telecommunications world and, noting some caveats along the way (for there are some valid practical exceptions that need to be made), let’s see how well we’ve done at meeting the target.

Let’s begin with broadcasting, but we’ll have to be a bit more specific: we’ll treat aural and video broadcasting separately, and leave text broadcasting for a later time. Video broadcasting — it is a fixed or a mobile service? Viewers generally sit in rooms in front of stationary receivers. And the Negroponte Law choice for television? To a huge extent, it’s a fixed service, so it goes to the wires for transmission! We don’t need to tie up huge chunks of radio spectrum with basic television distribution; there’s enough wire line capacity in most places to handle that chore. In fact, the only reason that television broadcasting has spectrum allocations is an historical one: after World War II, when television broadcasting was first authorized, the use of the radio spectrum was the ONLY way in which it could be easily and cheaply mass-distributed! That’s no longer the case.

Now for the exceptions. There are some rural, sparsely-populated areas of the country where it is economically infeasible to “hang cable.” These may still need RF distribution, either from local translators or by limited satellite service. And there is some developing interest in broadcasting to cellular telephones, although the acceptance of this by the public is not yet really known. But this is a special, bandwidth-limited television signal. Consequently some bandwidth-limited frequency allocations may be needed for “cellular television.” Overall, though, the traditional over-the-air broadcast television distribution system needs to be “sunsetted” and eventually removed from the radio frequency spectrum! To what purposes the liberated spectral resources might be allocated is quite a different matter, one which will be considered later.

Aural broadcasting: Most people listen to broadcast radio primarily while driving automobiles or using public transit services, or on small portable receivers while doing other things such as jogging, gardening, or working around their homes. Radio certainly is a mobile service and it should keep its access to the radio frequency spectrum. But there is also some true fixed broadcast usage: “Internet radio” is expanding. It consists of both live and previously-broadcast program streams from on-air broadcasters and also the independent audio streams from Internet-only program distributors. As is the bulk of all Internet traffic, Internet radio is primarily distributed by wire transmission. Since, in a general sense, wire line transmission is not capacity constrained as is the RF spectrum, the Internet can host a very large number of “radio” stations without jeopardizing a national resource. Thus it is a “good” place for the expansion of aural broadcasting, although there will always be a need for radio frequency transmission as well.

One additional caveat: per the discussion in an earlier blog posting, the United States has parallel broadcast radio services, on “AM” and “FM” bands. There is no longer a need for both services, and one of them should relinquish its spectral allocation (a mandate which is easy for a telecommunications theorist to deliver, but there are practical accommodations that will need to be implemented as well.) VHF FM radio, especially as it gradually transitions into full digital broadcasting, has the clear technological lead. It also has the majority of the listeners, and it has the capacity of absorbing much of the current content of AM radio as well. Thus, AM radio should be “sunsetted,” and its spectrum refarmed.

Print space is short, and we will continue the Negroponte’s Law examination of other radio spectrum users in subsequent posts. For now, though, once you see how the game is played, you might want to try it by yourself.