D-Lighting in Capture

I agree, the highlight slider does not to do anything for me either. It all means to me that when it is lost it is lost. Lesson for me is: don't overexpose!

 

I have always used PS since capture is slow darn slow. Hopefully, my new computer will be returned to me (fixed) this week, and I'll be able to try it out.

 

I am starting to use the D lighting but often am not happy with the "artificial" appearance and the added noise that I see. I also have tried "prophoto" as a profile offerred in PSCS which does essentially the same thing. I do see more response to the slider in prophoto (thanks Gale)
Dave

 

YW Dave,

that can be quite effective in some shots for sure.

I use many techniques, depending on the shot.

 

about the rule to expose "to the right"...

I believe, and I may be mistaken, that the rule to expose "to the right", e.g. slightly overexpose, rather than slightly underexpose is the rule for negative based shooting. When you overexpose, there is still info in the negative that can be coaxed out when you are in the dark room, however if you underexpose and don't convert any silver nitrate crystals in the negative you can't really help yourself in the darkroom. When you're shooting slides, the emulsion is your result so there is not help in the darkroom and you'd best not blow any highlights. Digital photography acts more like slide shooting then negative shooting and so you don't want to blow the highlights too much. A lot of this has to do with the perceptability of level gradation in the highlights as opposed to the shadows as well.

I believe that is the source of that rule. Of course now that I've spewed this out someone whe actually knows what they're doing can chime in and correct me

Dave

 

I use D-Lighting some of the times, but usually at a different setting than the standard, which is very aggressive. At times D-Lighting works miracles and hits the spot right away, but often I prefer using the curve tools.

Unfortunately D-Lighting is also very slow, and each time I open the RAW file it needs to recalculate.

I use PS only to work on specific parts of the picture, and that is really seldom. I'm also pretty new to PS.

 

Re: about the rule to expose "to the right"...

Well, that would not be me, but ...

Expose to the right (e^r) applies to digital photography when you are trying to eliminate noise in the shadow regions. Noise appears in shadows because of the way that numbers represent brightness.

The lower half of exposure in a twelve bit sensor is only 0 to 63 (6 bits, so small differences in brightness (say 7 and 15 - both nearly black to the eye) will render as very different bit counts and look different in a digital photo. This is what shadow noise is.

The brighter half of digital capture (bits 7 through 12) consists of the numbers from 64 to 3967. Here a few digita different do not make as much difference in the way the brightness looks in the image. So there is less noise in parts of the image that get at least half or more of the maximum amount of the exposure.

So, images exposed in the upper half of the 12 bits possible will be much less prone to noise than will those exposed in the lower half. Thus, to minimize digital noise, put the important stuff in the upper (right) half of the histogram.

Check out the 'white paper' by Bruce Fraser, or read his book on Camera Raw to hear this from someone who really knows what they are talking about.

 

good point about shadow noise

That is a good point about shadow noise. It is interesting to note that this feeds into a discussion of dynamic range. After all, a 12-bit pixel and a 10-bit pixel don't necessarily have different dynamic ranges. They may simply describe the same range with more (12-bit) or less (10-bit) level accuracy (resolution -- not image resolution, level resolution). When you couple to this, a logarithmic mapping of level to bit values you can get banding or more noticeable noise down in the shadows. Of course the 10-bit pixel could cover less dynamic range with more or less equal resolution thus providing less "noise", but a narrower operating range.

I am, by no means, an expert in how camera manufacturers record their raw data, but how they design the A/D converters that read out the voltages from their sensors can further cloud any a to b comparisons between different camera capabilities. After all, they are not required to follow either a linear or a logarithmic mapping of light intensity to quantization level. It really is interesting stuff, but as I've spent a whole day thinking about interesting stuff at work, I think I'll give my tired brain a rest and stop rambling. Good stuff in this thread though.

Dave