Image Editing Basics
THE IMAGE WINDOW OPENS BEFORE YOU FOR THE FIRST TIME, and you take a deep breath as you look it over and realize all the little things that need to be fixed before the image can be unleashed upon the public. What to do first? How to turn this into a masterpiece? Many images need only tone and color corrections plus a little sharpening before they're ready to go out the door. Other images need a few physical adjustments such as cropping and straightening, or may suffer from lens distortion that you'd like to remove. For these images, it's good to take care of physical corrections before you start adding layers and elements that depend on where things are in the image.
Before we get into the tone and color corrections in the following chapter, we want to spend a little time covering the ins and outs of image editing basics: resolution, transformations, and selecting areas and objects. Once you get a handle on these methods, it becomes much easier to apply tone and color corrections exactly where you want them.
All About Resolution
Don't Forget Your Resolutions
In the strict sense, resolution is a measure of how fine the image dots are, by dividing the number of pixels by the physical size (as in "300 dots per inch"). But the exact meaning of that number does change depending on where and how an image is used, and where it is in the capture-to-output workflow. Knowing your resolutions is important when you're communicating with someone else. Here's how we've boiled it down ( Figure 3.1 ):
- Sampling resolution. If you're using a scanner, this is the resolution at which a scanner samples the original. This number describes, in samples per inch (spi), how precisely an image was scanned. You control sampling resolution using your scanning software. For digital cameras, sampling resolution doesn't really apply: You could be taking a photo of a landscape 3 miles wide or a close-up of a 3-inch-wide flower, making "dots per inch" quite arbitrary.
- File resolution. When you save an image from Photoshop or other software, the software writes an image resolution into the file. File resolution doesn't mean much unless you print the image, because it doesn't refer to the number of pixels in the file. Some digital cameras write 72 ppi into a file just to put a number in that field.
- Effective resolution. This is the resolution of an image at its final output dimensions. For example, a 1600x1200-pixel digital camera photo scaled to 2.5 inches wide on a layout, creates an effective resolution of 1,600 pixels ÷ 2.5 inches or 640 ppi. If you scale up an image too far, its effective resolution drops below the value required for good quality.
- Device resolution. This number isn't part of the image at all, but is the resolution of the output device. For example, a monitor might be 96 dpi, and a platemaker might be 2,400 dpi. For print, an image doesn't need to match the device resolution. For example, you don't need to send a 5,760 ppi image to a 5,760 dpi inkjet! A 360 ppi image prints very well.
Figure 3.1 How the meaning of resolution changes as you move an image from capture to output.
If you aren't printing an image, don't worry about dpi at all. Just pay attention to the image dimensions in pixels (e.g., 1600x1200 pixels). That's how image sizes are measured on monitors.
Finding Out the File Resolution of an Image
I do a lot of print work, so I do need to be aware of the effective resolution of my images. Our publication requires images to have a resolution of 300dpi or higher. How do I use Photoshop to find out the resolution of an image?
You may not even need to use Photoshop! If you view an image in Adobe Bridge and display the Metadata panel (View > Metadata Panel), you'll see Resolution listed there. If you've already got the image open in Photoshop, you've got options ( Figure 3.2 ):
- To display the file resolution in the document window status bar, click the black triangle next to the status bar and choose Document Dimensions. The status bar displays the file resolution after the dimensions, in parentheses.
- To display the file resolution in the Info palette, choose Palette Options from the Info palette menu and enable Document Dimensions.
- Choose Image > Image Size, and you'll find the Resolution option in the Document Size section. What happens when you change the Resolution value depends on if you enabled the Resample checkbox; see the topic "To Resample or Not to Resample" on the next page.
Figure 3.2 You can view file resolution in the Info palette (top left), document window status bar (bottom left), and in the Image Size dialog box (top right).
Kenneth, What's the Effective Resolution?
I sent a digital photo to a magazine editor, but he told me that its resolution was too low. The magazine requires 300 ppi, and he said my image wasn't acceptable because it was 180 ppi. I was told that a 6-megapixel camera can produce quite good quality for an 8x10 image, and I didn't crop my image or change its resolution before I sent it to him. So why does he think the resolution is too low?
The editor is making a classic, common mistake: evaluating file resolution when he needs to be looking at effective resolution instead. In other words, he didn't take the final output size into account when looking at your image.
Let's examine at a 6-megapixel image and look at its resolution properly. One popular 6-megapixel digital SLR produces an image with dimensions of 2000x3008 pixels. The resolution means nothing until you state the size at which it will be printed, so let's state some sizes. Use the Image Size dialog box in Photoshop to do the math for you:
- With the image open in Photoshop, choose Image > Image Size (
Figure 3.3 To determine the effective resolution of an image in the Image Size dialog box, disable the Resample Image checkbox and change Width and Height.
- Make sure the Resample Image checkbox is disabled.
- In the Document Size section, enter the width or height at which you want to output the image.
- The Resolution value now tells you the resolution the image can inherently provide at that size.
The 2:3 frame proportions of the image don't match the 3:4 proportions of an 8x10-inch image, so we try entering 8 inches wide and Photoshop tells us it'll be 12 inches tall and 250 pixels per inch. A little low, but most magazines aren't 12 inches tall.
That still leaves one big question: Why does the editor see the image as 180 ppi? Digital cameras have no inherent resolution, but some would find it odd to have an empty Resolution field, so digital cameras tend to embed an arbitrary ppi value into their files. You could pull 3 different 6-megapixel cameras off the shelf, take a picture with each one, and compare the dpi values in the images. The image from one camera might claim 180 ppi, the second 300 ppi, and the third 350 ppi—even though all three actually produced the same number of pixels.
To Resample or Not to Resample. . .
That little Resample Image checkbox at the bottom of the Image Size dialog box ( Figure 3.3 ) is a frequent source of confusion.
When Resample Image is on, pixels are added or subtracted during resizing ( Figure 3.4 ). Resampling is usually a good idea when you make an image smaller, and occasionally a good idea when you enlarge an image. Resampling during reduction can remove unnecessary image data so that the file size doesn't take up too much disk space. Resampling during enlargement can help avoid jaggies in the image, but only to a point. Even with the best resampling methods, you can enlarge an image only so far—typically less than 200 percent—before it no longer looks sharp. Resampling carefully can help keep an image within that happy medium between too little detail and an unnecessarily dense image file.
Figure 3.4 When the original image (top) is enlarged with resampling on (top right), Photoshop creates more pixels to keep the resolution constant. When enlarged with resampling off (bottom right), the number of pixels must stay constant, causing the resolution to drop—the pixels get bigger.
You can experiment with resampling if you have an image open in Photoshop. First choose Image > Image Size. When you enter new values for Width and Height under Pixel Dimensions and Resample Image is turned on, the file size changes but Document Size stays constant.
There isn't one best method for all situations, so Adobe provides five ways to resample ( Figure 3.5 ):
- Nearest Neighbor duplicates a sample by simply copying a pixel next to it. This doesn't make photographic images look good, but is often better for some 1-bit-perchannel or indexed-color images.
- Bilinear duplicates a sample by averaging the pixels on either side of it. It's faster than the Bicubic options, but doesn't look as good.
- Bicubic looks at the samples on all four sides, averages them, and applies that value to the new sample point. Bicubic is slower than Bilinear or Nearest Neighbor, but yields the best quality with continuous tone images.
- Bicubic Smoother is a version of Bicubic that often works better when enlarging an image.
- Bicubic Sharper is a version of Bicubic that often works better when reducing an image.
Figure 3.5 The resampling choices in the Image Size dialog box.
The bottom line? When working with photos, use one of the Bicubic options. Because actual results depend on the image content, you might want to try each option when you're working with critical images.
Image Size vs. Canvas Size
The simplest explanation is that the Image Size command (Image > Image Size) affects the image, and Canvas Size (Image > Canvas Size) affects the document. If you enlarge the Canvas Size, you end up with more space around the image. If you reduce the Canvas Size, part of the image is chopped off.
A common use for Canvas Size is adding more area to a document while preserving the existing image size ( Figure 3.6 ). A great example of this is when you're assembling several photos into one big document. You can start with one photo and enlarge the Canvas Size to make room for adding and arranging other photos.
Figure 3.6 Original image (top), Canvas Size dialog box with new document dimensions entered (center), and the resized document (bottom).
Resizing a Batch of Variously Sized Images
I've got a folder full of images that my boss wants me to put up on our web site. I've got two problems here. One, the images are print-sized, so I have to reduce them for the Web. Two, they're all different sizes, so the Image Size command won't work right in a batch action. What can I do?
Use Fit Image (File > Automate > Fit Image) instead of the Image Size command. You can set a width and height within which each image is resized to fit. And it works great as part of a Photoshop action.