Myth: Pixels equate to ink droplets
You may have been told that you have to use a high image resolution (most commonly given as 360 ppi or pixels per inch) in order to get the best inkjet prints. Or perhaps someone has instructed you to use something much lower, maybe even 200 or 180 ppi. Who's right? Is anyone absolutely right?
Very important—we're talking image resolution here, not printer resolution yet.
All seem to get great results, so what's the problem? There are two problems, both related to size and RAM overhead. At 200 ppi, a print from the Nikon D2x 12.4-megapixel sensor would measure 14x21 inches. At 360 ppi, that same file could only produce an 8x12. If you wanted a 16x24-inch print (which this camera will produce beautifully), you'd go from 34.9 MB (the open size of the Nikon D2x image) to 44.1 MB if using a 200 ppi file—manageable for interpolation and easily handled by most digital photographers' computers.
Now, go from a 200 ppi image to a 360 ppi image at 16x24 inches and the file size increases to 142.9 MB! That's a huge jump. This changes how you can work on images in your computer while printing out such a beast. While the number of pixels started the same for both images at 4288 on the long side, they more than double to 8672 for the 360 ppi image, yet only increase to 4288 for the 200 ppi image.
But isn't more better? Doesn't this mean you have more detail? Not necessarily. A printer can only deal with so many pixels. More than it needs to define a detail is wasted information.
I know that some digital printing experts claim they can see the difference with these high numbers, or at least numbers that are a fraction of the 360 ppi number (which also is a myth as will be explained shortly). To get to the truth of the matter, I did an informal test and printed a whole series of prints of the same subject, but with different ppis. I mixed them up and no one at our publication offices could tell them apart or put them into any consistent order of image quality.
Some individuals try to explain why you print with a certain image resolution by math. If a printer has a resolution of, say, 1440 dpi, their reasoning is that a multiple, such as 360, 300, 240 or 180, will give best results. That sounds good, but there's a serious flaw to that reasoning.
Pixels that make up your image and the ink droplets that make up a printer's dpi aren't mapped one-to-one. Even if they were, there's another problem. Inkjets don't put down the same resolution of ink over the paper—it varies depending on the density of the tone and the color. And in another twist, the ink droplets typically change in size, too, depending on the color and density of the image, making their resolution harder to define. Now what “resolution multiple” do you use?
How this mapping of image pixels relates to the way ink is laid down on the paper is one of the most closely guarded secrets of every printer company. There are important patents involved and only those with a “need-to-know” clearance know how this works. If it was so simple and making a direct connection between pixels and droplets was all that was needed, the competition among printer manufacturers wouldn't be as fierce as it is.
I asked a number of engineers in the inkjet industry about this and they all confirmed what you've read so far. Steve Semos at Epson, who's an inkjet printer product manager and an avid photographer, says, “Pixels aren't mapped to ink droplets. Pixels are mapped to a cell that can be made of many droplets of different sizes and resolutions. As far as image resolution goes, you never want to go below an image resolution of 180 ppi as the image will start to break down at that point. There's usually no need to go higher than 300 ppi, however.”
Semos typically prints with 300 ppi image resolution just because it's always safe and easy to remember. According to Semos, a high-quality image will print well even at 180 ppi, but lower-quality images need more than that, especially when they have a lot of noise, which is why he considers 300 ppi “safe.”
Experiment with different values and see for yourself what you think looks best. Following Semos' guidelines will get you in the proper range. Do your own testing and you're sure to generate the best results.