I work with lots of photographers, most of them professionals. When I first begin to work with a new photographer, one of the first questions I ask is how she handles sharpening in her workflow. That’s because sharpening is one of the most misunderstood aspects of the postproduction workflow. When it’s done incorrectly, it can have a detrimental effect on the final image. If someone is making this mistake, I want her to know before moving on to other things.

Digital photo sharpening is nothing more than enhanced edge contrast. Photoshop tricks you into thinking a photo looks sharper by isolating edge detail and enhancing contrast along those edges. One side of the edge is lightened while the other side is darkened. The enhanced edge contrast is referred to as haloing because of the effect it causes along these edges.

There is no magic formula for sharpening because the amount of sharpening for a particular image depends on two very different things — the content of the image and its overall dimensions. Images with lots of edge detail, like the bowl of silver rings shown below, can handle more sharpening than images with fewer hard edges, such as photos of people or a photo of a landscape on a foggy morning. This is because lots of sharpening adds to the feel of the ring photo, while it would detract from the softer feeling of the portrait or foggy landscape. Additionally, a smaller print of this shot doesn’t need as much sharpening as a larger version would require.

© Jordan Sleeth

At first thought, sharpening seems like a no-brainer. Who wouldn’t want their photos to be sharp? However, the subject is much more complicated. The main reason is that there are three different kinds of sharpening. Each type of sharpening needs to be applied at the appropriate time.

• Capture sharpening. By their very nature, digital image files need to have some sharpening applied. This sharpening pass is considered baseline sharpening. When you shoot raw this is addressed during conversion with Adobe Camera Raw. When you shoot JPEG, this baseline sharpening is applied by the camera.
  
  Caution: Most cameras have a sharpening setting that allows you to apply different sharpening presets to your photos as you capture them. If you’re shooting raw, this setting has no permanent effect on the image. However, if you shoot JPEG, it does have a permanent effect. I always recommend turning off all in-camera sharpening when shooting JPEG because it’s easy to add sharpening later, but impossible to remove it after the original file is created.
  
• Creative sharpening. This type is used to fine-tune an image creatively by selectively modifying the sharpness of selected areas of the image using selections or masks. When I use the term sharpening here, I am also referring to its opposite, blurring, which is the lack of sharpness. What this means is that creative sharpening can be used in the same image to sharpen something of interest, such as someone’s eyes, and to blur something else, like the background around the subject. The image does not need to be resized before this sharpening is carried out.
  
  The main thing to understand about creative sharpening is that its effect is relative to the rest of the image. The goal isn’t to make part of the image perfectly sharp. The goal is to make part of the image stand out from its surroundings by sharpening it or blurring the detail surrounding it.
• Output sharpening. This is overall sharpening that’s designed to prepare an image for final output, such as printing or onscreen viewing. This sharpening is applied to the entire image with the intent of getting it ready for a particular output option. One of the things to understand here is that size matters. A file that’s being prepared for printing as a 5 × 7 requires a completely different sharpening scenario than the same fi le being prepared for a 16 × 20. Otherwise, if the sharpening on the 5 × 7 looks great, the 16 × 20 will not be sharp enough.

The reason sharpening is divided into these three areas is because oversharpening degrades the quality of a photo by introducing unwanted artifacts. Oversharpening occurs when output sharpening settings, which are stronger, are used for capture sharpening. Later when the image is sized and prepared for output, it’s necessary to sharpen again because the resizing affects the first sharpening pass. When the second sharpening pass is carried out on a previously over-sharpened file, it can adversely affect the quality of the image.

This also means that any creative sharpening will be further sharpened during the output sharpening process. Knowing this is important because it means you need to avoid overdoing the amount of creative sharpening.

Something you especially want to avoid when possible is sharpening an image for output, changing its size, and then resharpening for a new output size. Sharpening on top of previous sharpening adversely affects the image by introducing unwanted artifacting — distortions introduced by the digital process — causing image details to look “crunchy” instead of smooth.

On more thing to keep in mind about sharpening is that it’s not used to fix severely blurred images. It’s used to compensate for some of the effects of digital capture — whether by camera or scanner. If your photo is a little soft in focus, then you may be able to help it with sharpening. But you’ll never be able to make it look the way it would have looked if it had been shot in focus.

No matter if you’re doing creative sharpening or output sharpening, you’ll use one of Photoshop’s two sharpening tools: the Unsharp Mask (USM) and the Smart Sharpen filters. These filters are very similar to one another. I’ll compare and contrast them as you explore how they’re used.

Using the Unsharp Mask Filter

People are often confused when it comes to using the USM filter because the name is totally counterintuitive. Why would someone want to use something named unsharp to sharpen an image? The reason this sharpening filter has such an odd name is that it refers to a sharpening method that was used with film before digital editing was an option. In that method, a negative that needed to be sharpened was duplicated. The duplicate negative was intentionally created just a bit out of focus. The two negatives — the original and the new one — were then sandwiched together slightly out of registration and then printed. The effect increased contrast around edge detail and made the resulting print look sharper. This is the same way the USM and the Smart Sharpen filters work.

The Unsharp Mask dialog box, shown below, is found in the Filter menu (choose Filter > Sharpen > Unsharp Mask). This filter doesn’t detect edge detail per se; instead it looks for pixels that have different tonal values than surrounding pixels. It then increases the contrast of those surrounding edge pixels, causing the lighter pixels to get lighter and the darker pixels to get darker, creating the sharpening halos I mentioned earlier. The sliders in the USM dialog box are used to control the size of these halos. Here’s a closer look at the controls:

Note: You may notice that there are two other sharpening filters in the Sharpen menu, named Sharpen and Sharpen More. One would think that these are the main sharpening tools, but they aren’t. They are blunt instruments that can’t be controlled. In the many years that I’ve been using Photoshop, I’ve never used either one.

• Amount. The Amount slider is used to control the amount of contrast between differing pixels, which affects edge contrast. Higher values equal more contrast and lower values equal less contrast.
• Radius. The Radius slider is used to determine the number of pixels that are changed when the filter sees tonal variation. Higher values increase the size of the halos, causing the sharpening to be obvious. Because of this, the Radius slider is the most important slider in this dialog box. Exercise caution when using it because too high of a value creates sharpening halos that are noticeable. To high of a Radius value combined with a high Amount value causes the image to look oversharpened. Keep in mind that this value is going to vary depending on the subject matter. A lower value works best with photos rich in edge detail, while a higher value can be used for photos that don’t have as much detail in them.
• Threshold. The Threshold slider is used to determine how different in tone the surrounding pixels need to be before they’re considered edge pixels, causing them to be sharpened. For example, a value of 5 affects only neighboring pixels that have a tonal difference of 5 units or more (on a scale of 0 to 255). The default value of 0 causes all pixels in the photo to be sharpened.

Tip: Apply creative sharpening to a duplicate image layer in the master layered file. This insures that you can undo any sharpening later by discarding the layer.

Using the Unsharp Mask requires a bit of a balancing act among these three sliders. The best way to understand how they’re used is to take the USM filter for a test drive. Follow these steps:

1. Download High_Desert_Flower.tif and zoom to 50 percent. This photo has lots of edge detail on the main subject, with little edge detail in the background. Remember that the USM filter, like the rest of Photoshop’s filters, only affects the currently selected layer.
2. Choose Filter > Sharpen > Unsharp Mask. The Unsharp Mask dialog box appears.
3. Type 200 for Amount, 2.0 for Radius, and 10 for Threshold, as shown below. Notice how much sharper all of the edge detail becomes.
4. Change the Amount value to 300. This oversharpens much of the edge detail on the flower and makes the image too contrasty.
   
   Tip: Though the USM dialog box has a preview window, you won’t be using it. It’s much better to use the actual image for evaluation purposes. So go ahead and drag the dialog box to the side so you can get a good look at the flower.
   
   

   

   
   
5. Now increase the Threshold setting to 25 and notice that the sharpening tapers off. This is especially noticeable in the middle of the crown of the flower where there are so many similar colors. That’s because only edge pixels with a tonal difference of 25 or more are sharpened.
6. Increase the Amount value to 300. The sharpening increases, but it’s still more subtle than the Step 2 setting.
7. Now increase the Radius value to 4. Notice how much this small adjustment affects the flower’s edge detail. It’s beginning to look too crunchy, which means it’s becoming over-sharpened.
8. Decrease the Amount value to 200. The crunchiness is reduced, bringing the edge contrast back into line. As you can see, there isn’t necessarily an exact set of sharpening values for this image because of the way these settings affect one another. If this photo were being printed, it would benefit from the settings in Step 3 or Step 8. Something else to notice here is that the out-offocus bee is still out of focus, even though its edges are sharpened.

I find that low Radius and Threshold settings and higher Amount settings are usually the best place to start when using the Unsharp Mask. In the previous exercise you began with 200 for Amount for this 6 × 9 photo. If the photo were larger, then a higher starting Amount value (such as 300 for an 11 × 14) would be appropriate.

Then adjust the Radius and Threshold values to match the subject content of the photo.

When the sharpening is complete, take one more look at the tonality and contrast of the image. That’s because large amounts of sharpening affect tonality and contrast. If you’re using adjustment layers for managing the tones, make any further adjustments to them before flattening the file and/or saving it.

Tip: Sometimes sharpening causes a color shift. When this happens, it’s easy to fix. If the sharpening is applied to a duplicate layer, change the layer’s blending mode to Luminosity. If the sharpening is applied to the main image layer, choose Edit > Fade and change the Mode to Luminosity in the Fade dialog box. (Just remember that Fade must be the very next step after sharpening is applied.)

Using the Smart Sharpen Filter

One of Photoshop’s newer filters is the Smart Sharpen filter introduced in version CS2. This filter is considered smart because it treats various regions of the image differently based on the content of those regions. The Smart Sharpen filter attempts to sharpen only the areas of the image that have detail without affecting areas that don’t. This is different from the USM filter that affects all areas of the image equally. The illustration below shows the Smart Sharpen dialog box. Notice that this dialog has Basic and Advanced modes. The difference is that the Shadow and Highlight tabs are added in the Advanced mode. This dialog box doesn’t have a Threshold slider because it isn’t needed. It also has some controls that aren’t in the USM dialog box.

Take a closer look:

• Amount. This slider is used to control the amount of contrast between differing pixels, which affects edge contrast. Higher values equal more contrast and lower values equal less contrast. It functions the same as the sliders in the Unsharp Mask.
• Radius. This slider is the same as the Radius slider in the Unsharp Mask dialog box. It’s used to determine the number of surrounding pixels that are changed when the filter sees tonal variation.
• Remove. This is a cool feature that adjusts the way the filter works, depending on the problem. The pop-up menu has three options: Gaussian Blur, which is the same algorithm used by the USM filter; Lens Blur, the best choice for most digital camera files; and Motion Blur, which attempts to compensate for blur caused by motion during the exposure. When Motion Blur is selected, the Angle setting is activated. This allows you to input the direction of the motion that caused the blur. For example, if the blur is from a sideways motion during the exposure, use a value of 0.
• Advanced. Selecting this radio button gives you more control by allowing you to work with the shadows and highlights independently of the rest of the image. It adds two new tabs to the dialog box, Shadow and Highlight. The figure below shows the Shadow tab, which is identical to the Highlight tab. You can use Fade Amount to adjust the amount of sharpening and Tonal Width to restrict your adjustments to the shadows with the Shadow tab, and to the highlights with the Highlight tab. This is quite useful when you have lots of noise in the shadows that you don’t want to sharpen.
• More Accurate. Selecting this option provides a more accurate sharpening effect, but the process takes longer.

Note: The only thing I don’t like about this filter is that the dialog box is huge, which can make it hard to see the image while using the filter — especially when working on a computer with a smaller display. Because I don’t use the dialog’s display, I scoot the dialog to the left until I can’t see the dialog’s display, freeing up more room for viewing the image.

In Step 1 of the previous exercise, I asked you to zoom to 50 percent before using the USM filter. There are two reasons I specified that zoom ratio. The first reason is that historically, Photoshop does the best job of drawing the image onto the screen with zoom ratios that are multiples of 25 percent. This has changed with the introduction of OpenGL image processing in Photoshop CS4. When OpenGL is in effect, the screen is drawn with equal accuracy at any zoom level. However, if you’re using Photoshop CS4, but your system doesn’t support OpenGL, then this won’t be the case for you.

The second reason I specified 50 percent is because it is the zoom ratio that usually comes closest to approximating the actual size of the printed image on your screen, while meeting the multiples of 25 percent rule I describe previously. One would think that zooming to 100 percent would display the image at its actual size. However, because a computer monitor isn’t capable of displaying an image at 300 ppi, it has to spread the pixels out to the resolution it can display — usually between 70 and 90 ppi. (That’s why 72 ppi is a common resolution for photos intended for the Web.) This causes a photo with a resolution of 300 ppi to look bigger on the computer monitor than the actual print. If you’re viewing at 100 percent while sharpening, you probably will be disappointed in the results because your preview doesn’t match reality. You’re seeing the photo at a much larger size than the final output size.

The optimal zoom ratio can vary from system to system. Here’s how to find out which zoom ratio is closest to the actual size for your particular viewing environment:

1. Choose View > Rulers to turn on the rulers. The illustration below shows a photo with the rulers displayed. If your rulers are already turned on, you don’t have to do this.
2. With the Photoshop rulers showing, hold a real ruler up to the screen while zooming the image using the keyboard shortcut presets (Command++/Ctrl++ to zoom in and Command+–/Ctrl+– to zoom out). When Photoshop’s ruler and the real ruler match, the file is displayed at its actual size. If OpenGL isn’t functioning on your system, you want a multiple of 25 percent. Try 25 percent and 50 percent to see which is closest to reality. On all of my monitors, 50 percent is a little bigger than reality and 25 percent is a little smaller. I use 50 percent because an image that displays a little bigger is easier to look at and evaluate. If OpenGL is functioning, you may find that 33 percent, the preset between 25 percent and 50 percent, comes closest.

Note: When sharpening for the Web, zoom to 100 percent to display the image at actual size because the output is intended for a computer monitor.

Following these guidelines helps insure that the sharpening you see on your screen more closely matches the final output size. However, be advised that there are a couple of other variables that come into play when adding output sharpening.

• Display versus Print. Even in a perfect world, the sharpness shown on a monitor won’t always translate to printed output. That’s because the way a computer displays an image is different than the way the image looks on paper. In fact, different monitors often look different from one another. When I first switched from a CRT monitor to an LCD flat panel monitor I noticed that everything looked sharper on the new monitor. That’s because of the resolving power of this newer style of LCD monitors. I had to make a mental adjustment to compensate for this.
• Printing Processes. Different kinds of printers and printing paper affect the way an image looks. Prints on glossy paper always look a little sharper than prints on matte. Some inkjet papers, such as fine-art papers, are very absorbent and really soak up the ink, which diminishes the effects of sharpening.
• Size Matters. Sharpening is dependent on the dimensions of the print. That’s why it’s important to do all final cropping and sizing before output sharpening is applied. The thing to take away from this section is to experiment with your intended output process until you’re comfortable predicting how your computer’s display translates to a printed image. Make some prints and compare them to the images on your monitor until you feel comfortable predicting any discrepancies between the way the monitor displays the image and the way it looks when printed.