The jpeg image you get from your camera depends on in-camera sharpening
and noise reduction algorithms. If you look through camera comparisons,
f.i. at dpreview.com , you'll see that different manufactures have
different visions of this balance. Some of them prefer less aggressive
noise reduction (Panasonic), some more (Kodak). So an jpeg image from
Kodak camera looks smoothly with less noise, but with less details.
When you convert image from RAW with UFRaw, you get it closely to "as
is" from camera matrix. Afterwards you can apply sharpening or/and noise
reduction. Noise reduction plug-ins can be found at
or any other from the page
I'd recommend you selecting some "critical" parts (like faces) before
applying noise reduction with greycstoration, overwise it would take a
lot of time and memory.
Gimp-user mailing list