Hello Ken,
I recently got a new product from
Alienideas who are the publishers of FocusOn: a guide to digital imaging. Its a self training and reference DVD e-book with video clips but covers all of the core skills of digital photography - I found it very useful and to prove it I've copied and pasted a short section on sharpening for you
Sharpening
Image files produced by scanners and digital cameras tend to
be softened due to the actions of electronic filters which are
used to remove rogue pixels - an effect known as aliasing.
This means the detail in image files can be improved by
resharpening them.
The most common method for performing sharpening is
Unsharp Masking (USM). Although this sounds like a misnomer
for an image sharpening process, it is actually based on an
old printing technique which was used to enhance image
edge details. When an image is in Lab mode, we recommend
sharpening is performed on the L (Lightness) channel:
Filter > Sharpen > Unsharp Mask
When in RGB or CMYK use Unsharp Masking (USM) with
particular care: image sharpening is definitely better underdone
than overdone. In RGB or CMYK gamuts USM sharpening
adjusts the colour of the image - so if over saturation is
noticeable it may be best to convert the image to Lab mode
or try the High pass technique that follows.
With a layered image, flatten the file before applying unsharp
masking to ensure the sharpening is applied equally. As this
implies it is also possible to apply sharpening to layers, or to
selected parts of an image.
One alternative is to use the Photoshop Sharpen filters, but
these do not offer the fine controls of Unsharp Masking:
Filter > Sharpen / Filter > Sharpen More
Guideline Unsharp Masking for digital
camera files:
Amount - 130-200%
Radius - 1-2
Threshold – 2-20
We suggest the Amount is dropped to 150 and increased
to 200 to make a ‘bracketed’ comparison for fine tuning. A
setting of somewhere between 130-200 should be suitable
for most high resolution images for printed output.
Once you are comfortable with the effect, try small adjustments
to the Radius and Threshold as required. It is worth noting
the settings you use, especially if you have a single source,
a single output device and a preferred output size. It is also
common, for example, for social photographers to keep images
of people slightly softer than commercial images, as sharp
detail can be unflattering.
The higher the radius figure the more pixels are affected. USM
should only be used to sharpen the edge pixels so this figure
should generally be no more than 1-2 in the majority of cases
(this setting is in fact resolution and image dependent).
The threshold is used to determine which pixels are counted as
edge pixels. This measures the difference between pixels on
a scale of 0-255 (this number being the scale that produces
continuous tones in each colour channel). A figure of 5 will
alter all the pixels which differ by 5 shades or more........
It goes on and on but that should help you.
Best
Arthur Ruttler