Defeating the computer's arch enemy: unreadability
Monday, March 5, 2001
People miss planes, burn dinner, and stay up way past bedtime just to read one more page of a good book.
But it's not just the quality of the prose that causes the worm to burrow so deeply into a book. As typographers have long known, the aesthetics of print have a lot to do with keeping the eye on the page.
Since Gutenberg put together the first printing press with movable type more than 500 years ago, typesetters have agonized over the optimum point size of the letters, whether they should have those curlicues on the ends called serifs, what style of font to choose, and the precise amount of white space needed between lines of text to make the words stand out.
Printing got so good that readers found it easier to immerse themselves in a book. But with the advent of the computer, reading became infinitely more difficult. The biggest problem was resolution -- the clarity of words and images on the screen. A big stumbling block in computer evolution was fuzzy letters arranged on glaring screens that left the reader with a blinding headache. Readability is so poor that, according to Microsoft, the average reader hits the print button after just three paragraphs.
As the rest of the world fiddled with new technologies that would provide higher resolution, Microsoft was working on new fonts that would make it easier for the eye to focus on a computer screen. But Bill Hill, a Scot hired in 1994 to head Microsoft's typography section, wanted to know what was going on in readers' brains when reading a book. He was convinced people would switch from printed page to screen if he could duplicate the experience.
"The magical thing about the book is it disappears when you read it. You're not even aware of the book because the real book is going on in your head. How does it do it?" Hill says on the Microsoft Web site.
The result? He decided the brain works on two levels as we read: It subconsciously recognizes words as patterns, which frees the unconscious to concentrate on the words and their meaning. So typographical elements -- the shape and thickness of the characters, the space between the letters and words -- help us recognize the patterns. If the proportion of those elements is radically different, as it is on a computer screen, it interrupts the reading process. The brain has to make a conscious effort to identify the patterns, which takes away from the comprehension.
Resolution is measured by dpi, or dots per inch, which refers to the number of pixels on a computer screen, a number usually between 72 and 120. At about 300 dpi, the eye can sail through pages of paper, said Microsoft product manager Mario Juarez.
According to the wisdom of Hill, recorded for posterity at the Microsoft Museum's display on ClearType: "Trying to portray type on screens with today's resolutions is like somebody actually painting a picture of the Mona Lisa with a paint roller. The problem is you need a smaller size paint roller or even a brush."
Look closely at a computer terminal and you will see a grid in the background made up of very small circles (on a screen lit by cathode-ray tube) or squares (on a liquid-crystal display).
Microsoft works with LCDs because they are the screens used in laptops and e-books. Hill, who was asked to join Microsoft's eMerging technologies group in 1998, decided to look more closely at the pixels, the building blocks used to construct words and images, which are black or white. Stacking blocks to make a capital "A," for example, creates jagged edges, something the eye trips over.
Hill and his team found a way to divide pixels into equal-sized vertical stripes of red, green, and blue lights that form the basis for colour on an LCD screen. Their software, called ClearType, mathematically manipulates the subpixels, tricking the eye into seeing smooth edges where there are nothing but snaggle-toothed letters. The impact on resolution is immense: It is the equivalent of reading 300 dpi.
Microsoft won't yet reveal exactly how it works, but Hill believes his group has invented something to rival Gutenberg's movable type, something that will allow electronic books to challenge the primacy of paper.
"The next five to 10 years will probably change society as much as the 200 or 300 years after Gutenberg," according to Hill. ". . . We're on the verge of something."
In search of a clearer electronic display
While letterforms can be printed with smooth outlines on paper, a computer screen displays images using small blocks called pixels.
On most screens, there are only 72 pixels per inch, making type at small sizes very blocky and hard to read. Creating the appearance of curves using square blocks requires an optical illusion. A number have been developed to achieve this. (These letterforms are magnified to show more clearly what distortions are needed to trick the eye.)
This is a method of rearranging pixels in order to create the best possible illusion of the letter's real shape. Different mathematical instructions are calculated to suit different type sizes.
This method takes the illusion a step further by shading some pixels grey to improve the appearance of smoothness at the edges.
Microsoft's new ClearType process attempts to take the process even further by using the three colour components within each pixel.
Every pixel is made up of red, green and blue light. ClearType varies the intensity of each colour to blur a letter's jagged edges even more.