The Future of eBooks
When predicting the future of any given technology, I always claim that the easiest way of getting good results is to consider the ideal form factor for a device. In this regard, eBooks are no different.
What, then, is the ideal form factor for an eBook reader? For the last few hundred years, we’ve been using a device for reading that has a lot going for it. It’s cheap to mass-produce, has a long lifespan, doesn’t require batteries, and has a fairly simple user interface. It has some downsides, though, the most important being size. I would estimate that around half of my possessions by mass are books; and, while they have some aesthetic appeal, storing them all is a nontrivial problem—one I solve by lending as many as possible to friends who look like they need something to read.
The biggest advantage eBooks have over their traditional paper counterparts is that eBooks are readily portable. When I was traveling a lot in 2006, I downloaded a few dozen books from Project Gutenberg to my Nokia 770. This device has a lot of limitations, but the screen is absolutely gorgeous. It’s the first device I owned that I didn’t mind using for long periods. Although the reading experience wasn’t as pleasant as that provided by a printed book, the fact that I could carry enough books in my jacket pocket to last for an extended trip was a huge advantage.
How Many Pixels?
Returning to the ideal form factor question, what does an eBook reader lack versus a printed book? In the case of the Nokia 770, the display was inferior. But 2007 saw the commercial exploitation of a technology that has been promised "Real Soon Now" since the mid 1990s: electronic ink. A figure of 300 dots per inch (dpi) typically is thrown around as the minimum for text to be clear and comfortable to read. Current eInk displays provide about 166 dpi. This is lower than the 225 dpi resolution of the Nokia 770 LCD, but more readable due to the fact that it’s reflective rather than emissive. But even this level of readability is still less than that of the printed page.
To keep printing costs down, most books are printed with only one color of ink: black. A technique known as dithering gives the illusion of shades of gray, using patterns of small black dots at a given density to simulate gray. This technique allows printing to sacrifice some resolution in exchange for more colors. (The dot pattern has to be smaller than the eye can discern, so the effective resolution is lower when dithering is used.)
This technique can be applied in reverse on an electronic paper display. Characters in text typically are stored in a vector format. A TrueType font, for example, contains a series of Bézier paths. When a TrueType font is printed or displayed, the curves are drawn over a regular grid of pixels. Some of the pixels are completely covered, and these are colored black. Some are completely uncovered, and these are colored white (that is, not printed on). Others are partially covered. When printing, these partially covered pixels typically are printed if they’re more than 50% covered; otherwise, they’re not printed. Most display systems color the partially covered pixels with a shade of gray, giving a smoother appearance than a black-and-white image at the same resolution. Since most current electronic paper displays support 16 shades of gray, they use this technique to increase their apparent resolution.
The big thing missing from current electronic paper displays is true color. Using 16 shades of gray reminds me of my first laptop, a 386 system, and is incredibly primitive by modern standards. One big challenge for the next year will be getting color displays into an eBook application. Three different types of ink and the same number of levels as are currently available would give 12-bit color. This isn’t as many colors as even a cheap thin film transistor (TFT) LCD offers, but is enough for displaying most images with only a small loss in quality. This capability would make eBook readers convenient for applications such as online newspapers with color photographs.
The other big limitation of current displays is that they take about 0.9 seconds to update—not bad when you consider how long it takes to turn a page in a conventional book, but much slower than most screens, and certainly not fast enough for video. Is video in an eBook reader interesting? Maybe. Once you get the ability to play video, there isn’t a huge amount of difference between an eBook reader and a portable computer. Having a dedicated device for reading is only interesting at the moment because display technologies have widely different sets of advantages and disadvantages. Future display technology, whether based on organic LED, eInk, or some other system, is likely to remove this tradeoff.