The unsharp mask is found in a number of programs, as in PSP 7. It seems rather misnamed, in that unsharpening is a method of paradoxically acheiving just the opposite - a sharpening of certain edges in the original. The advantage of using unsharp, rather than the typical sharpening filters, is that unsharp can be very carefully controlled. It provides a different sort of sharpening. But the effect is very similar, and need not introduce highlights, or jaggies, or other unwanted artifacts produced by sharpening filters, alone.

One can start with any image. This one, above, just seemed sort of colorful. It's not particularly blurry, or soft focus. It's pretty sharp to begin with. But it can still show some improvement. 'Unsharpening' is performed by blurring the image, then recombining with the original to produce a sharper result.

    

The image on the left is just the original from the top of this page. The image on the right (or bottom) is the slightly sharper version just created, here.

Now, for 'Unsharpen'. The Unsharp Mask filter, more precisely, produces the same thing, in one easy step.

• from the menu bar: Effects, Sharpen, Unsharp Mask
• Set as follows:
  • Radius: .9, same as blur factor, above
  • Strength: 70
  • Clipping: 0

And run it through a second time, for an even sharper look. It might be just what the photo needs to get it to stand out on the page. Clipping tends to remove highlights. And it would be used particularly if you increase the radius much beyond 1 or 2. The filter works well at a tiny value, like this. So, best to maybe apply a couple of times, with a small radius, rather than all at once.
 
The radius in the unsharp filter is used to set the blur on the duplicate image in just the way suggested for the manual example. As you increase the radius, you tend to increase the artifacts from this method, as if you had increased the blur in the manual method. The blurred image is more like the softer areas in the original. So it tends to actually remove less sharp areas from the original when subtracted out to produce a mask. And so the mask, of course, contains the sharper areas. Then, by adding back that mask, the sharpness is emphasized, but without doing anything to the softer parts of the image. The intensity setting or, strength, is the amount by which the blurred image is subtracted. This filter effectively highlights both light and dark high contrast areas; specifically right at the boundries between light and dark. The effect is that the highlights and shadows can sort of grow, or spread, as the radius and/or intensity are increased. Clipping can help reduce this. Clipping is a threshold, beyond which the filter is applied, and below which the filter is bypassed or ignored. But the effects from a much larger radius would have to be something you are comfortable with. Again, generally, smaller values, maybe applied repeatedly, would be more typical for the use for which the filter is intended.

Now, at this point, some might be curious about the photographic method which inspired the digital filter - unsharp. Apparently it's even a bit more direct, and different, again. It involves, not surprizingly, creating a second plate based on the one you have, already. So you have a picture, a negative, to start with. Who knows where you got it? Now you create another 'negative', by shooting entirely through the first. It's just that this second plate is separated from the original by a certain thickness of glass - whether the glass, itself, or the slight distance away creates the blur, I don't know. But so you get film that picks up a blur, to some extent, of the original. It's a mask. And to create the final print, the two are overlayed. The mask goes 'on top', and the original faces the photo paper. The blurred areas soak up a bit of the light coming through, which tends to favor the sharper areas in the original. It's exactly different than the manual method, described here, which rather than work with an actual blurred or unsharp mask, uses a sharp mask which is then combined with the original. It's just the opposite. And, also, I don't know the origin of this lab method; who invented it, who took credit, and so on. But it certainly seems, to me, like something someone discovered by accident, by messing up a shot or two.

[My thanks to K. Zaklika, of JASC Software, for reviewing this tutorial, and explaining to me both a) the manual method used to simulate the unsharp mask and b) the photographic method which lent its name, at least, to the digital filter.]