Hi!
It behooves Boris to define his workflow to some degree.
Shel, I've no idea what "behooves" is, but I suspect I know what you're asking of me anyway. So here goes. I'll try to be as detailed as I can so prepare for rather long message.
1. I shoot RAW only. Color balance is almost always set to automatic. My camera is set to: Saturation: 0, Contrast: -1, Sharpening: +1. I've been told once that Contrast -1 slightly improves the sensitivity range. I use Sharpening +1 because it seems the images coming from camera are more on the soft side than I would accept.
2. I process with Adobe PhotoShop Elements 3. So, for instance, perspective correction through crop seems to be out of my reach.
3. When I process I adopt some of the technique suggested by Rob Studdert:
a. I would start with "As Shot" color balance and work towards better color rendition if I disagree with "As Shot" values. Very often for outdoors shot I would end up with about 5000 K and -5 - -15 Tint.
b. Then I would change exposure in such a way that no areas of the shot are marked as blown out highlights.
c. Then I would play with shadows so that I wouldn't have any areas marked as blocked shadows. Here I would also see that my general contrast is high enough. So I usually tend to make shadows darker.
d. Then I would set brightness as to compensate for potential global darkening which was caused in b. above.
e. Finally I would usually increase contrast a little via contrast control.
f. If I want to go b/w I might set saturation to -100. Usually I leave it at 100. That's because now I have Channel Mixer enabled via set of Earthbound light Effects package.
g. Finally I would set luminance smoothing to 0 almost always. Could be I am mistaken here. I simply found that if luminance smoothing is more than 0, picture seems to loose its sharpness...
h. Then I would set sharpening to 75-100. And finally color noise reduction I usually set to 0. That's because I rarely shoot in low-key/underexposure situations. So I don't need this, I think. Again, it could be I am mistaken with sharpness. As per Rob's suggestions, I would always enlarge a portion of image to examine it. If I see clearly that sharpening introduced artifacts, I would lower the sharpening level...
All together, from reviewing the above process it seems that I might be overdoing sharpening a little...
Also I usually process in 8 bit. Which could be wrong, but in Elements I have to be 8 bit, and also for printing I have to provide a JPEG file - so 8 bit is kind of forced on me.
Now, to post-processing itself.
1. If I have to crop, I would usually start with cropping to standard size, such as 45x30 or 40x30 or 30x30 (all centimeters of course). I would set the resolution to 200 or 300 dpi so that when I crop I would also res up for final size. If my crop size is non-standard, I would not res up.
2. I would then either apply Curves or more often perform Levels correction. Very rarely I would need to use Selective Color or Hue/Saturation to fix up the colors.
3. Finally I would use High Pass filter to sharpen it even more. Recently I learned that I don't have to always Overlay the resulting layer. I may use "Vivid Light" or "Soft Light" which gives more interesting results. Also I may push the high pass filter to higher values of threshold, such as 5 or even 10. Then I would Gaussian Blur it before merging it with the original layer. Could be this is a wrong idea all together.
4. As for USM, even technique suggested by M Reichmann from Luminous Landscape, which is quite mild, is still too harsh for me.
That would be it. I would always save my files for print at highest possible quality level of JPEG. Often though I use progressive optimization option. It seems to give slightly smaller sizes without sacrificing quality.
Finally, the "For Web" part. Here it could be I am doing more mistakes.
1. I would resize my image by one step. So I would set resolution to 72 dpi and either width to 700 pixels or height to 640 pixels. That is the most comfortable size for me to keep editing at 100% view ratio.
2. Usually the resized image is softer than the original. So I would High Pass sharpen it at 0.25 or 0.3 (at most 0.5) threshold.
3. Finally I would apply my standard frame (10 pixels white/black, 1 pixel black/white, and 20 pixels white/black).
4. When I am saving my image I wouldn't use "Save for web" (may be big mistake), I'd simply use Save As and set image quality to 7-10 out of 12. My ultimate goal would be to obtain an image which is not larger than 100 Kb. Sometimes, I cannot do it. Then I would save it at maximum quality. Then I would re-open it in IrfanView and save from there at 80% of original quality. Very often, that would result in smaller image.
That would be it.
Now, if you see that I am doing something wrongly, please indicate how exactly I can correct myself. Then I could try this and learn.
Finally, to make it crystal clear: I am not abandoning Pentax. Not after spending few years as a member of this club :).
Boris
