Last time I wrote about how opinions on quality in digital image capture changed over the last fifteen years. Now I’d like to do the same for inkjet printing. If you asked the gurus at the imaging engineering conferences of the early nineties what it would take to get photographic output from a drop-on-demand inkjet printer, they’d tell you, in chorus, “One picoliter drop size.”
How small is a picoliter? Imagine standing in your kitchen with a quart of milk in your hand. Imagine one million tiny glasses in front of you. Now pour just enough milk in each glass so that they’re all equally full. Now dump out all the glasses but one, and distribute the milk from that glass equally into all the other glasses. Neglecting evaporation, you now have about a picoliter, one trillionth of a liter, in each glass. If you had a trillion glasses lined up, and each glass were an inch across, the line would wrap around the earth more than 600 times. It’s hard to imagine.
Those of you who have been following the recent progress of inkjet printing technology know that we reached photographic quality with droplets of larger than 5 picoliters. How come we were so pessimistic? It’s not that our calculations were off, we were thinking inside the box. We had made an assumption that turned out not to be true: that we’d stick with four-ink color printing.
Drop-on-demand inkjet printers, color laser printers, and offset presses are all binary printers: at any given place on the paper, they can either put ink down or not put ink down, but they can’t put down shades of ink. We researchers had known for a long time that when you’re going for quality in binary printing, adding additional inks with dilute colors really helps. Thus, if you’re printing black and white images, adding one or two grays will radically improve image quality when it becomes prohibitively expensive to make the dots smaller. If you’re printing color, add a light cyan and a light magenta (you don’t need a light yellow because it’s already pretty light). However, the folks who made printing presses, and maybe the folks who bought printing presses, didn’t seem to want to go down that road. So, we figured it was kind of like having more than three kinds of color sensors in a camera¬—something that we researchers just knew was the way to go, but for some reason the manufacturers and/or the marketplace weren’t interested—and we didn’t think we’d ever get the printer manufacturers to go down that road.
Well, we researchers were wrong, at least in the case of inkjet printers (Even today, when printing presses stray beyond the traditional four colors, it’s to get more saturation by adding colors like green and orange). Light cyan and light magenta popped up on inkjet printers about five years ago, and were joined by light black (now, there’s an oxymoron!) two years ago. The improvement in quality was phenomenal, the printers sold like hotcakes, the customers using the printers for photography were ecstatic, and our predictions turned out to be junk.
What went wrong? With image capture, we didn’t understand the nature of photographic quality. With printing, we didn’t understand the nature of the customer. A printing press can cost millions of dollars. Printing operators are notoriously conservative. Printing customers value predictability exceedingly highly. All that adds up to extreme resistance to change. In the inkjet world, the cost of the printer is (artificially) low compared to the cost of consumables such as ink and paper, and the consequences of a purchasing mistake are small, so customers are more adventurous. Photographers in particular have demonstrated their interest in achieving high image quality, no matter what the risks.