Tutorial 7

How to sharpen an image

In this tutorial, we will show you how to sharpen an image.

Before:

After:

Typically after processing, the last step is to sharpen the image. This step is often important after performing image size reductions. You can use one of the following menus to sharpen an image:

Adjust/Enhance > Smart Sharpen
Filters > Sharp > Sharpen
Filters > Sharp > Unsharp Mask

In this tutorial, we will use the Unsharp Mask tool to sharpen an image. We took a screenshot and reduced its dimensions by 50% (see Figure 10.07.1). The words in the image look kind of fuzzy so we would like to have it sharpened a bit. You may right click on Figure 10.07.1 and save this image to your disk, then open it in Image Broadway if you want to practice it yourself while following the tutorial steps below. You may also right click on this image and copy it to the Clipboard and use the Image Broadway File > Clipboard menu to paste it in as a new image. Optionally, you may click the Navigator tab to display the Navigator window and use its slider bar to zoom in and out; or you may click the Info tab to display the exact pixel color and cursor information in the Info window.

   
   Figure 10.07.1

1. Use the Filters > Sharpen > Unsharp Mask menu to open the Unsharp Mask dialog box as shown in Figure 10.07.2.

   
   Figure 10.07.2

2. In this tutorial, we use 5 as the pixel radius to blur the image. Usually we use a number between 2 and 10 as the blur radius. We will apply 300 percent of the difference between the original and the blurred image to sharpen the image. The number in Amount (%) is typically between 50 and 500. The "0" threshold means to apply the difference no matter what. The result is shown in Figure 10.07.3.

* What the Unsharp Mask does is applying a Blur Mask to the image to remove noise. It uses the specified radius to blur the image. Then, it calculates the difference between the original and the blurred image. If the difference is larger than the specified threshold, it will apply the difference multiplied by the Amount (100 is 100 percent) to the pixel. It works on each channel separately, then combines the result.

   
   Figure 10.07.3