Sharpening works by increasing the contrast around edges. Edges in images always involve darker tonal values adjacent to lighter ones. We can emphasize the edges by making the dark tonal values darker and the light tonal values lighter.
In the analog darkroom, this was accomplished using an unsharp masksee the sidebar, "Why Is It Called Unsharp Mask When It's Used to Sharpen?" for details on just how this process worked. In the digital domain, we sharpen by identifying the dark and light pixels that represent edges, and lowering the value of the dark pixels to make them darker while raising the value of the light pixels to make them lighter.
When we do this, we create a "halo" that makes the edges, and hence the entire image, seem sharper. The concept is simple, but as with many things in digital imaging, the devil is in the details that we discuss throughout this book!
Figure 1-1 shows the same image before and after sharpening. (The image also illustrates the pitfalls of driving in rural Scotlanda good metaphor for the myriad things that can go wrong when we use sharpening inappropriately!)
Figure 1-1. Before and after sharpening
The only difference between the two versions is the sharpening. Figure 1-2 shows a zoomed-in comparison with an accompanying graph of the values of a single row of pixels, before and after sharpening.
Figure 1-2. Sharpening up close
Notice that the tonal range of the sharpened versionthe distance between the lightest and darkest tonesis wider than that of the unsharpened version. Notice too that the biggest differences occur at the edge transitions of the capital "O" while smaller differences emphasize the texture of the sign's background.
Sharpening is closely related to contrast, but simply increasing the contrast over an entire image just produces an over-contrasty image. Successful sharpening demands that we localize the contrast boost to those parts of the image that actually represent edges.