Voss wrote:
I've been using Auto Sharpen in Elements a lot.
It's hard to say what anything "auto" related to sharpening might do! I'd avoid it like the Black Plague!
Sharpening is a really difficult concept to understand. Unlike exposure, where there are 400 conflicting opinions for every 200 photographers, with sharpening there are 6 conflicting methods for every 1000 photographers. Lots of silence... but no agreement either.
A few brief points:
1) Every image originally made with a camera that uses a Bayer Color Filter Array to encode color needs to be sharpened in some way to some amount.
2) Sharpening is the last step in a production workflow. That is true because everything else, and in particular how an image is scaled to a specific size for display, changes how sharpening should be done. It is very
very different for a 600 pixel wide thumbnail for the WEB than it is for a 300 PPI image printed at 24x30 inches for framing. It's different if that print is on canvas than if it is on Glossy Lustre paper too!
3) There are several different kinds of sharpening. None of them, with one possible exception, will result in higher resolution. They all increase acutance, or the appearance of sharpness in whatever level of resolution already exists. (Richardson-Lucy deconvolution is an algorithm that can, if there is enough information about whatever is causing image blur, actually increase resolution. With typical Street Photography we just don't have that kind of information to see any significant increase in image resolution.)
4) The two most basic types of sharpening come in many forms, and each editing program is different. The values used for one program to not apply to another, and values for one type do not match those for the other. Unsharp Mask (USM) is one method, and Sharpen (actually a high pass sharpening algorithm) is the other. Often at least a touch of each should be applied to every image, though sometimes one other is not needed. Generally, but not always, USM will have more effect on an image that has been down sized from the original and Sharpen will seem to be more effect on an up sized image.
5) USM operates on single edge transitions. Sharpen operates on multiple consecutive tonal transitions. Hence a picture of a white house with a large black door that has a picket fence around the front yard will show an effect on the door to wall transition when USM is used, and the picket fence will be affected when Sharpen is used.
6) Consider that sharpening, no matter how it is done, operates on the contrast of high frequency spatial detail, and then realize that down sizing an image is a very effective Low Pass filter! So if an image is first sharpened at the original pixel dimensions and then down sized to thumbnail size for WEB viewing... the effect is to virtually remove the sharpening.