Hi Ganesh / Nilanjan as discussed
I think of it in three stages and I try to stay true to film based principles which are very much applicable in Digital sensitometery as well.
[
b]How my equipment will respond [/b]. The typical problem is that the Luminance Range in nature far exceeds the capacity of the mediums we have
Eye Perhaps >10,000 : 1
Slide film 250 : 1
LCD projector 200 : 1
Glossy print 60 : 1
The CMOS sensor is somewhat similar to the Slide film. So calibrating and understanding how the sensor behaves and having that in control is the first aspect.
The sensor behavior is based on a Characteristic curve. This determines how it treats shadows and highlights . Not at all unlike a film. The famed Tri x had a specific Characteristic curve that users over the years learn to use and exploit to get a consistent look for their pics .
Pre visualizing the scene: Now fitting the scene into the Sensor recordable range is both a technical and an aesthetic decision. Technical to record the maximum information possible for later use. Including max amount of pixels. Keeping the Histogram to the right. Here we have to remember that built in Matrix metering will always give an average reading and make its own compromises. At times we might want to take that decision ourselves.
Aesthetic because we need to decide what to leave out as information based on what we want to show.
Interestingly here the challenge is same as film low exposure and extracting the image creates grain and poor image.
Processing the Negative to create the visualization This is where the greatest benefits of Digital world kicks in. In the analog world we were stuck with printing the negative on to a paper with a single type of tone response- That is a fixed characteristic curve.
In the Digital negative Photoshop allows us with curves to adjust the different tonal areas with different curves based on how we feel aesthetically pleased.
If we look at creating the visualization in terms of curves that would be the right way. Do I want to show a gentle gradation in the shadows with a greater leg portion of the curve ?
Do I want to show a greater Highlight gradation with the shoulder ?
Do I want to have a gentler slope in the mid tones to bring out a smoother range ?
I see clipping as a quick way of achieving this. I am not sure but I suspect by clipping and redistributing we may be loosing pixel information therefore loose very fine tonal gradation.
Again like in film the best tones adjustment is the minimum tone adjustment . That means a lot of things were thought out in advance.
Lastly better the bit depth better the tonal gradation.
Deft Definition :
Slide: Nothing to do with stock market or children play equipment
also known as transparency. Film based image where we directly get a positive image that can be projected. This used to have the highest dynamic range in Analog