Where Science and Serendipity Converge

Redmond, Wash., Jan. 25, 1999 — Greg Hitchcock wasn’t expecting much when he opened Microsoft Paint one day and rendered a set of random letters on a low-resolution computer display.

He was trying out some new code he’d written as part of a promising experiment to make text easier to read on an LCD. His first attempt failed, drawing the characters in a reverse image, like a photograph negative. But after a quick fix, Hitchcock tried the code again and this time it worked. He was stunned. The letters, while still not perfect, were almost unbelievably clear.

Hitchcock, a Microsoft employee, had for several intense weeks been part of a small team of researchers and developers attempting to tackle one of computing’s most vexing problems: the poor quality of typefaces, particularly on devices such as laptops that use LCDs. The group had devised an approach for manipulating the pixels on an LCD, and Hitchcock had just finished working out a preliminary set of algorithms.

“At this point it was just a theory,” Hitchcock recalled. “Then came the moment where I tried it out.”

The team had hit a major, albeit unplanned, milestone in the development of a breakthrough technology, soon to be dubbed ClearType, a classic example of how research and development can meld to produce significant innovation.

The remarkable improvement in font display resolution is only part of ClearType’s appeal. Just as valuable is the fact that it works on existing machines without requiring hardware modifications or upgrades. A software solution integrated into the operating system, ClearType will be available in Windows products as early as this year. ClearType represents the convergence of the expertise of a group of individuals who, while working separately, had each long been focused on a common goal: improving the readability of computer displays. Their efforts came together in an intense period of innovation and discovery that started in May of last year and culminated in the unveiling of ClearType at COMDEX in Las Vegas on Nov. 15.

Clearly pleased, three of the masterminds behind ClearType are now focused on integrating the breakthrough technology into Microsoft products: (left to right) Greg Hitchcock, Dick Brass, and Bill Hill (not shown: Bert Keely and Mike Duggan).

From Vision to Reality

The story goes that the single most important event in Bill Hill’s life was learning to read at the age of three.

The son of a steelworker in the slums of Glasgow, Scotland, Hill was taught early that education was the ticket to a better life. He earned scholarships to good schools and eventually became a successful national journalist. But by the early 1980s, his interest migrated to computers-desktop publishing in particular.

A striking figure with a broad beard and Scottish brogue, Hill headed Microsoft’s Typography group for two years. But he grew frustrated with the limitations of CRTs and LCDs, and sensed that no amount of typography development could make computer screens truly readable without a breakthrough in display technology.

In May of 1998, Hill joined Microsoft’s Bookmaker project, an effort to develop electronic books. Working long into summer evenings, and brainstorming constantly, the team members took their knowledge of how the eye processes the printed word and meshed it with expertise in the physics of LCDs. They came up with an idea for manipulating LCD pixels in a set of new ways. It succeeded sooner than any of them had dreamed.

Pondering the Problems of the Pixel

To begin to understand how ClearType works, take a close look at the LCD on a laptop or handheld PC (use a magnifying glass). You’ll see that letters and pictures are composed of clusters of tiny, rectangular pixels. Look closer and you’ll see that each pixel is actually made up of three smaller sub-pixels of red, green, and blue.

On an LCD, a florescent light is typically projected through a screen of pixels, each of which contains three microtransistors, one for each sub-pixel. The sub-pixels act essentially as colored shutters and the transistors open them in combination to produce a given color in the pixel.

The problem with the LCD is that the pixels are so large, and the screen area so small, that it’s long been impossible to provide resolution adequate for comfortable reading. “The typical laptop has a resolution of about 90 DPI (dots per inch),” Hitchcock said. “You get about 100 DPI on some of the laptops as well. But by contrast, 300 DPI now is considered low resolution for the standard printer. We’re still three times short of that with an LCD.”

The result, Hill said, is that the eye constantly struggles to resolve the jagged edges and blurry lines generated by coarse pixels.

“Typefaces have very subtle features that have developed over thousands of years to make words easily recognizable,” Hill said. “On the computer screen, we don’t have the resolution to show those features.”

Consequently, on a document of any significant length, “you tend to hit the print button.” Hill and the team recognized that the only way to truly render fonts was to address the area beyond the traditional pixel boundary and find new ways to work with sub-pixels. “There’s not just one thing happening with the code in this product,” Hill said. “There is actually a whole set of different algorithms at work in ClearType.”

The technology produces letter shapes and character spacing that appears more like a printed page than any conventional approach can produce. “ClearType makes inexpensive screens look as good as the finest displays and makes the finest displays look almost as good as paper,” said Dick Brass, another member of the ClearType team.

ClearType and the Electronic Book

Team members regard ClearType as only the leading edge of a wave of innovations. Most important, they see it as the critical step in bringing to life a vision of electronic books that will finally render paper obsolete.

“It has much broader implications than just improving the display on an LCD,” Hill said. “We always viewed that readability was something that absolutely had to be solved in order to make electronic books work. But if you solve the book, which is an ‘extreme case’ of reading, you solve everything underneath it. So you make business documents, spreadsheets, database information, and the Web all more readable.”

Brass echoed that idea, describing ClearType as “the fundamental enabling technology for e-books right now.”

“You also need LCD displays, low-power processors, long-lasting batteries,” Brass said. “But none of them would have allowed us to make e-books that people would want to read at prices they could afford to pay now without the 300 percent improvement in resolution that ClearType gives us.”

Brass envisions a day when electronic books in a wide range of form factors will supplant paper-based reading. “We all know that wood is only going to take us so far as an information distribution medium, the way wood vehicles and wood buildings and wooden weapons only took us so far. And now we’re getting ready to move past wood. We’re at the edge of a huge transformation in our society.”

While looking forward, the team also paused briefly to reflect on the serendipity that led to the creation of ClearType.

“There was a lot of experience combined between the members of our team -and probably getting on to 40 years of experience in different aspects of the technology,” Hill said. “The magic was that we all came together and brought different pieces of the puzzle, and they fit.”

“ClearType happened by getting smart people together, and smart people make things happen. There’s also some magic to it,” Hitchcock said. “And, you know, it makes you wonder how many more of these breakthroughs are still out there, just waiting to happen.”

Related Posts