n the previous article I introduced the concept of sharpening an image and why it is necessary. The problem with using the basic approach of the Unsharp Mask (USM) is that it applies the sharpening to the entire image. This means that both areas of detail as well as uniform expanses of colour are affected by the USM. This makes itself noticeable in the form of sharpening artefacts (broad expanses of colour start to look a little noisy).
The easiest way to sharpen only what you want sharpened while leaving the broad areas alone is to create a second layer - on which the sharpening will take place - and mask off the areas that will not be sharpened. In some cases this is in fact sufficient. However if you have interpolated the image to a large degree or intend on enlarging the image in a print, the transition between the sharpened area and the unsharpened lower layer can become evident, particularly in the shadows which are already those most at risk to digital noise. This article therefore looks at a way in which areas of detail can be sharpened while leaving alone areas that lack detail.
For the purpose of illustration I’ve chosen an image that has both detail and large areas that are out of focus. What I’m wanting to do is create an accurate mask that only selects areas of detail while masking off the out of focus portion. The image of two Julia butterflies mating is tricky for conventional masking as areas that require sharpening are very small and intricate.
The Sharpening workflow
Step 1 – Assuming that all other necessary processing has been completed, a new sharpening layer is created above all other layers. The easiest way to do this is to hit SHIFT+Ctrl+Alt+’E’ which merges all the visible layers into a new layer at the top of the layer stack. After this new layer is created rename it ‘sharpening’ and create a new layer copy.
|