Could anyone to explain me or point me to some resource which explains how to interpret various curves in post-processing applications ? I'm using UFRaw, and RawStudio.
RawStudio tends to have a very low-contrast default curve. I find that pictures are quite muddy and it takes a bit of adjusting to make them punch.
Generally I start by creating a node near the very bottom (dark) end and pulling it down for contrast, and then creating a node near the midtones and bringing it up. Hard to explain exactly what shape I'm talking about; check out this screenshot. See the way it goes down a tiny bit and then curves up gracefully in the middle? (I find that I can't bring up the midtones too much, or else I blow out the highlights.)
I haven't ever used UFRaw, but it sounds like it has a similar 'base curve/correction curve' concept as my fave raw processor, DarkTable (great program, but only available in Linux). The base curve should already be set for you, and it imitates the sort of tone curve that your camera would automatically apply to a JPEG. In actual practice, the pre-set base curve should look a lot like the curve in the above screenshot.
You can think of the base curve as a 'sensible default' curve. It takes the rather low-contrast RAW image from the CCD and compresses the shadows/highlights to make it look like it did in the camera preview. (Rawstudio's rather muddy base curve is more true to what the camera sees, but nobody likes that look. As an interesting aside, most consumer-grade film is high-contrast, because people tend to prefer it. Terrible for taking really subtle, artistic photos with lots of tonal range, but it makes your average snapshot look exciting. Digital cameras do the same thing with JPEGs.)
Then you go and fine-tune in the 'correction' curve (or whatever it's called in UFRaw). It took me a while to understand this concept, that they both do essentially the same thing – but the base curve does it on the incoming picture and the correction curve does it on the post-processed picture.
In a nutshell, don't touch the base curve unless you're getting consistently bad results with each photo. Adjust the other curve instead, and you should be good!
For a more in-depth description of what's going on in UFRaw with the two curves, check this discussion out: In ufraw, how to use the two curves? The answerer has a different opinion from me; they think it's better to adjust the base curve rather than the correction curve.
Heres an article for the curves tool in Photoshop.
The ideas should be the same in all tools.
Here's another one which talks about curves for different colour channes: