Ken Perlin, computer scientist and director of NYU’s Games for Learning Institute, had a problem: Microsoft Word’s default fonts turned into illegible mush at tiny sizes. So he took it upon himself to design a typeface of his own that would remain easily readable on LCD screens even at the microscopic size of just 6 pixels. (Technically speaking, he allotted 5 pixels for the letterforms and 1 pixel of “leading” in between lines.)
The magnified image above shows off the successful result: that’s the first 500 words of the Declaration of Independence, squashed into a 320 x 240 rectangle. Here’s how it looks actual-size:
Sure, it’s not exactly pleasant to read, but it’s easily doable — even when the text is a foot or more away from your face. But Perlin’s handiwork, informally dubbed Tiny Font, even holds its legibility when blown up. For comparison, here’s how Arial looks at an identically minuscule point size:
And here’s Tiny Font again. See how you can actually still, like, read it?
How’d Perlin do it? Like any typographer, he painstakingly designed each Tiny Font letter individually. But he also had algorithmic help: He wrote a program in Java that took normal-sized ASCII characters as input and bitmapped them down to a fixed height of 5 pixels — while mathematically interpolating the variable width of each character to exploit the RGB scan lines on an LCD monitor.
In English, that means that Perlin’s algorithm could squash the width of each letter into the smallest possible space allowable on a monitor’s pixel grid. The “pixels” on an LCD monitor are actually small strips of horizontal dots that repeat: red, green, blue, red, green blue. They’re tiny enough that, when mixed together, we can visually perceive them as any color in the rainbow. Perlin took advantage of that fact in his Java program to make each letterform as visually distinct as possible within those RGB-RGB stripes. (That’s why the letters look strangely multicolored when blown up, but mostly black-and-white at normal resolution.)
It was still a matter of trial and error and personal taste — Perlin twiddled and tweaked the variables in his program to make each letter’s RGB-visibility as sharp as possible to his own eye. But because his algorithm did most of the heavy lifting, he says each letter “took less than a minute” to get right.
Of course, the big question remains: Why? “I just wanted to see if I could do it — I took it as a personal challenge,” Perlin tells Co.Design. “I was also intrigued by questions of improving text legibility on Third World cell phone screens.” After all, the lion’s share of the planet’s cell phones are not equipped with Apple Retina Display technology: they’re stuck with smallish, 320 x 240 QVGA-resolution screens. “If this is the resolution that billions of people will be looking at for the next few years,” says Perlin, “I wanted to know: how good can you make it?”
Thanks to surging interest online, Perlin says he hopes to create a downloadable version of his work — not just Tiny Font itself, but his shrink-ray Java source code as well. Theoretically, that means Perlin’s code could be applied to any typeface that a designer wanted to make micro-readable at a glance. With luck, low-rez screens the world over may soon have a typographic champ in their corner.