So why would anyone ever worry about pixel resolution? Why not take photos with the highest resolution the camera can provide? Why not scan images at insanely high quality?

Well, consider the floor tiles. What if each tile costs a dollar (or a yen or pound or euro)? It doesn’t matter how big or small the tiles are, they all cost 1 unit each.

If you’re rich and don’t care about the cost of the tiles, you can buy as many as you want at the floor care center. But reality says you need to worry about the cost of each pixel. And in terms of resolution, the cost of each pixel is an increase in file size.

Now, I’m not going to get into the mathematics of resolution as it pertains to file size. But suffice it to say, the only thing you need to know is the more pixels in a raster image, the higher “cost” (file size).

Back when designers first started using computers, file size was a very important constraint. We constantly needed to keep the file size low to fit on the small hard drives. But today hard drive space is huge.

So is there any good reason to still worry about the size of print graphics? Large file sizes could slow down a desktop printer that processes your files if they’ve got more resolution than necessary. I’ll cover that later in this chapter.

An example of how higher resolution creates smaller pixels and more details. The 300 pixels per inch image has more details around the bicycle rider and the lamp post. However, the same area is just a bit of a blur in the 72 pixels per inch image.

Linescreen, also known as “frequency” or “lines per inch” (LPI), is not computer jargon—the term has been around for many years. Printing presses print with dots of ink, so when a grayscale image like a Hollywood photo is converted for printing, the different shades of gray in the image must be converted into black or white dots. Dark gray areas are converted into large black dots, close together; light gray areas become small dots spaced farther apart; white areas have no dots.

In a full-color photograph, the cyan, magenta, yellow, and black channels of a color image are converted into dots that will be printed with transparent inks that overlap each other to form all the other colors. The smaller the dots, the higher the linescreen.

Different jobs, such as magazines, books, newspapers, or brochures, are printed with different linescreens. Find the strongest magnifying glass you can, or borrow a loupe from a designer or a print shop (a loupe is a powerful magnifying glass used by printers to see the dot patterns in images). With the magnifying glass or loupe, look at the photos in different printed pieces.

You’ll notice that the size of the dots differs between each piece. Images in newspapers are printed with big, coarse dots—you may not even need a magnifying glass to see them. Images in slick magazines are printed with much finer dots.

As I mentioned, linescreen is measured in lines per inch, or LPI. In an 85-line (LPI) screen there are 85 lines of dots in one inch, both horizontally and vertically; in a 200-line screen, there are 200 lines of dots in one inch. Obviously, the higher the linescreen, the smaller the dot pattern; the smaller the dot pattern, the more detail can be printed. However, linescreen is not the same as output resolution. See the sidebar above for an explanation of this.

This book was printed with a linescreen of 150 LPI. Most magazines are printed with a linescreen of 150 LPI. High-end art books may be printed with linescreens of 200 or higher.

As computers and scanners became more common in printing and production, people started running tests. They looked at how an image printed after it was scanned at various resolutions. They also looked at different linescreens and what different printed resolutions looked like at different linescreens. And they came up with this general rule:

Images tend to look their best on paper when their image resolution is twice the linescreen of the finished, printed piece.

Using a resolution higher than two times the linescreen doesn’t increase the quality of the printed image, but creates an unnecessarily large digital file and can slow down the time it takes to print the image. That’s why I don’t recommend using super-high resolutions for images.

So the following mathematical formula has become the rule to determine the correct resolution.

Image resolution equals twice the linescreen.

Simply multiply the linescreen by 2. The result is the image resolution. If you plan to print an image in a newspaper that uses a linescreen of 85 LPI, you don’t need more image resolution than 170 PPI. If you plan to print an image in a magazine that uses a linescreen of 150 LPI, you don’t need more than 300 PPI. Because so many print jobs are output at 150 LPI, 300 PPI has become a “standard” for resolution.

Yes, this means you must know the linescreen of the final reproduction process before you scan or create an image. How? Ask the right person. Or see the Linescreen and resolution chart on the following page.

The linescreen formula has become a guide for setting resolution; however, it’s not a hard and fast rule. Many times you can get away with lower resolutions, depending on the type of image, the press, the type of paper, and many other considerations. If you want to play it safe, follow the rule.

Sometimes in the frantic pace of the production process, though, you need to resize your image in the page layout program, which essentially changes the resolution. Or you may need to work with images that were scanned for a different linescreen, or perhaps you need to reduce the physical size of an image (like from 4 inches to 3.5 inches).

Well, no one will send you to jail because you didn’t have the perfect resolution. Your images will print. The worst that might happen if you enlarge an image is it will appear jaggy in the final printed piece, and if you reduce an image, the file can take longer than necessary to output.

Take a look back on page 85 at the three images of the building. As you can see the image at 150 PPI doesn’t have all the detail of the first and the one at 75 PPI is a complete mess.

But other images are much more forgiving of the twice the linescreen rule. The clouds in the image on the next page don’t look quite so bad even at 75 PPI. There just isn’t much detail in the sky to lose at the lower resolution.

The clouds and sky of this image don’t need a high resolution to look acceptable.

Use the following chart as a guide for the image resolution of grayscale or color images to be reproduced with different processes. This is just a guide! You really should call the print shop and ask them what they want you to use, especially if you’re setting an advertisement or other project in a newspaper or magazine where there can be a wide range of linescreens.

If your project is going to be reproduced on a commercial printing press, you’ll most likely have your digital files output to an imagesetter, an extremely high-resolution PostScript printer. Even though the final destination is the printing press, the imagesetter is the final output device.

Most imagesetters output at resolutions around 2400 DPI or 2540 DPI. So if you were paying attention in the previous section, you might expect to scan a 1-bit image at a resolution of 2400 PPI or 2540 PPI. Well, it was nice that you were paying attention, but it’s not necessary or even advantageous to use such a high resolution because the imagesetter software throws away any data in the image that exceeds 1200 DPI, so 1200 PPI is the highest image resolution you need.

Let’s go back to our kitchen floor. What if we double the size of the kitchen, but we don’t have more tiles to cover the space. We might have to stretch each of those tiles to twice its size. (I know it’s hard to stretch tiles, but stay with me.) The pattern of the tiles enlarges to twice its size. But does the pattern have any more detail? No, the tiles are bigger, but the details stay the same. An image that looked fine at its original size doesn’t look so good when stretched; it loses detail.

That’s what happens when you enlarge a raster image, whether you do it in the image editing program or in the page layout application. When you originally scan or create a new file, it creates a set number of pixels. If you later increase the physical dimensions of the file (like the rubber being stretched), the pixels have to stretch and the image looks bad.

But what if you could add more pixels instead of stretching them as you scale an image? In Photoshop, the term for changing the number of pixels as you change the size of an image is called resampling. When you resample up, the resolution stays constant as you increase the dimensions of the file.

You might see this resampling feature and say, “Aha! I have found a way to avoid the jaggies! I’ll just resample the image as I increase the dimensions and thus it will keep the detail and look great.” Unfortunately, resampling up doesn’t work as well as getting the right resolution to begin with.

The program doesn’t know what detail it’s supposed to insert as it creates the new pixels. So it guesses. Unfortunately, its guesses result in a fuzzy image rather than a detailed one. If you understand what’s going to happen when you increase the dimensions of an image, you can sometimes get away with it, but it’s never a good idea to double or triple the physical size of a raster image.

If it’s a scanned image, rescan the original art at a larger size as discussed in the next section. If it’s an image from a digital camera that can’t be retaken, try the sharpening technique described later in this chapter.

Let’s say you have an old high-school yearbook photo that’s only 1 inch wide. But you want to print it at 8 inches wide. How do you scan the image so that you get the correct resolution? The simple formula for scanning with “extra” resolution is this:

Divide the final width by the width of the original image and then multiply it by the desired final resolution. The result is the resolution at which you should scan the image.

In the case of the yearbook photo this becomes (8/1) × 300 PPI = 2400 PPI. This gives you enough pixels to resize the image to the correct resolution. (There are also controls in the scanning software that will do all the necessary math for you.)

But what if you just want to scale something ever so slightly? Like you’re putting in a photograph in your brochure and you need the photo just a bit bigger. Do you really have to rescan the original?

No, says the ever-practical person. (We won’t ask the purists.) For continuous-tone images like photographs, that tiny little bit of scaling won’t make much of a difference at all. In fact, it’s hard to see any difference in an image that is scaled less than 10 percent.

Obviously there are problems with increasing the dimensions of a raster image, but what about decreasing the dimensions? It’s no problem at all if you do it in the image editing program.

But if you place a raster image into a page layout program and then reduce its dimensions on the page, you do have a potential problem because even though you resize the dimensions, the file size stays the same. This means the computer may still send all the image data to the printer.

For instance, let’s say you place a 4-inch wide photo on your newsletter page and realize you really need it 2 inches wide. The 4-inch photo is 2.3 MB; when you resize it on the page to 2 inches, there’s still 2.3 MB of information in the file.

Unless you set the print controls in the page layout program to throw away any extra pixels, the computer will send 2.3 MB of data to the print processor. If you’re sending the job to the office desktop printer, this could tie the machine up for a while. Fortunately, the professional print processors at large print shops can usually handle this type of excess information.

A program like Adobe Photoshop or Photoshop Elements can help you enlarge an image that doesn’t have enough resolution to begin with. Use the option to Resample Image and then choose the Bicubic Smoother (best for enlargements). You can then enter the settings for a new document size.

As shown below, Bicubic Smoother does avoid the jaggies associated with enlarging images, but it leaves your image a bit soft or blurry. The Nearest Neighbor option results in jagged edges. For images that require lots of details, resampling up is not a good option. But if it’s just a simple photo of the family Christmas dinner, a little resampling up won’t annoy Uncle Nathan.

An example of what happens when the small image is resampled up. In this case the small image of the pocket watch was scaled up 400%. In the middle image, the resampling option was set for Bicubic Smoother. This added the soft blur around the edges. In the far right image, the resampling option was set for Nearest Neighbor. This creates jagged edges in the image.

You can fix the softening of an image by applying a filter such as the Unsharp Mask in Photoshop. This gives the illusion of better details by sharpening the edges between the dark and light pixels. However, sharpening can’t add details that aren’t there to begin with. One of the ways that I was taught to use sharpening was whenever I was about to increase the contrast on an image, I should switch to apply a small amount of sharpening instead.

An example of how the Unsharp Mask filter can enhance details in a “soft” image. Here enlarging the watch on the left has blurred some of the details. A slight sharpening on the right gives the illusion of more detail.