I would like to make one more comment regarding up-sampling and sharpening. As a scientist I work with a lot of algorithms regarding sampling and convolution. Every algorithm I have seen has artifacts, and those artifacts get worse the harder you push. But more importantly, no one method consistently does better than another. Many are pretty good and "good enough" for many applications. When you really want to push the limits, try different settings for each algorithm you have, then choose the best parts of each test image and blend those together.

For example, if you want to improve the sharpness of the bird's eyes in an image, you will likely need one set of parameters for your sharpening algorithm, then when if you want to sharpen the bird's feathers that are slightly out of focus, you will need a different set. Optimize the settings for each section of image then dlend those sections together. That's what I did on the image I posted on Aug 30.

Roger