The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 69.0°F | Mostly Cloudy

MIT Proposes National Standard for Digital TV

By Karen Kaplan
Executive Editor

In the year 2000, high definition television sets will be in millions of living rooms across America. In fact, HDTV will revolutionize modern television the way color TV did when it was first introduced in the 1950's.

So says Jae S. Lim '74, a professor of electrical engineering and computer science who works in digital signal processing, speech processing, and image processing. Lim heads the MIT Advanced Television Research Program, and his research group has developed an HDTV system based on digital technology. The group's prototype, which was built by General Instrument in San Diego, is one of five models under consideration for adoption as the official U.S. standard for HDTV.

The MIT-GI model is currently being tested, along with the four other entries, at the Advanced Television Testing Center in Virginia. The ATTC, an industry-sponsored group, will evaluate each system and provide technical results. An advisory committee appointed by the Federal Communications Commission will recommend one of the models for official approval as the nationwide HDTV standard by the middle of next year.

"Each HDTV system has different special features, and MIT's is not the best in every aspect," Lim said modestly. "But I would like to think that MIT has the best record overall in designing a system."

HDTV system models were also submitted by AT&T, Zenith, the Advanced Television Research Consortium, whose members include NBC, the European companies Philips and Thompson, and a Japanese firm. General Instrument also submitted a second entry.


High definition television has "very significant improvements relative to the National Television Systems Committee television service," which is the current standard for televisions, Lim said. These improvements include:

* Increased picture clarity. "The fine details are a lot better," said Lim. "It's an increase of about 400 to 500 percent." With NTSC television, the screen is divided into a grid of 340 by 420 pixels, or picture elements. An HDTV screen's grid is 720 by 1280 pixels, producing a much higher resolution.

* Smoother motion. HDTV runs at 60 frames per second, twice the frame rate of NTSC, so figures move more smoothly on the screen.

* No "ghosting." Lim said HDTV eliminates the "ghost effects" and "snow in the background" that sometimes appears in NTSC television broadcasts. "With the new system, either you see it or you don't see it, and when you do it's spectacularly clear," he added.

* Better aspect ratio. The width to height ratio on current televisions is 4 to 3, and the dimensions were set 40 years ago when motion picture screens were more square. The aspect ratio for HDTV sets is 16 to 9, and the wider screens allow for panoramic filming. Also, because HDTVs have higher resolution, television screens can be bigger.

* Higher audio quality. HDTV is equipped with four channels of CD-quality audio, unlike current sets.

Watching a program on high definition television will be "almost as though you're going to a movie," Lim said.

Although most television watchers probably do not find very much lacking in the current NTSC sets, after seeing an HDTV demonstration, viewers will certainly appreciate the higher quality, Lim said. "Once you see it, you notice it. It's like the car -- before there were cars, it didn't bother you that you didn't have one, but once you have a car, it changes everything," he explained.

Regular color television film can be converted into high-definition film with current technology, so "HDTV won't suffer from a lack of programming," Lim said. Broadcast networks will have to purchase new equipment, and so far they have been reluctant to do so. But Lim predicts they will come around soon: "If they don't do it, someone else will. Then they will, too, because they don't want to be left behind."

Digital vs. analog

The high-definition television standard developed at MIT is different from Japanese HDTV systems in one important respect -- the Japanese system is based on analog technology, while MIT's is digital. By incorporating digital technology, the U.S. has regained the lead in this futuristic television field.

"This is part of a trend," Lim said. "Everything is going digital. Computers are digital. Clocks are digital. Television is just around the corner. Digital telephone services are coming up."

"A digital system erases the noise, in a sense," because digital signals are only zero or one, without a range of values in between, Lim explained. By contrast, an analog signal can be interpreted as having a fractional value. Because this value can range continuously between zero and one, much more noise appears in the picture.

"A digital system gives you either a perfect picture or none at all," Lim explained. "People have designed digital systems so that most of the time they get it right, and even when they don't, they have a nice way of concealing it."

The real advantage of a digital high-definition television system is that it can be integrated with other technologies, like personal computers. This is tougher to do with analog HDTV, Lim said. For this reason, the Japanese analog HDTV entry is generally considered out of the running for adoption as the U.S. standard.

Only American sponsors

In addition to Lim, the Advanced Television Research Program consists of 10 to 15 graduate students and two or three research affiliates. In addition to HDTV, their research has other digital signal, speech, and image processing applications.

The research group has received about $2.5 million in funding over the past three years, all from American sponsors like ABC, PBS, Eastman Kodak, and Motorola. With the growing "Buy American" sentiment, this may boost MIT's chances for selection as the U.S. standard. "I don't think this will hurt us," Lim said. "If everything else is equal, I think (the FCC) should adopt an American system, because every other country does it that way," he added.

If the MIT-GI model is selected, "It will bring us prestige," Lim said. "MIT can also collect royalties based on the patents it holds." Even if the MIT-GI system is not chosen, the FCC "may use some of our technology, so we may still be able to collect some money," he continued.

But Lim was quick to add that he's not in it for the money. "We're in it because we like to contribute technology to the community," he explained. "If, as a byproduct, we collect some money to lower students' tuition, then that's fine too."