Selecting Camera Exposure


I have gone back to take another picture of the wall in my living room with my Nikon D810, this time, following your suggestion of "increasing the exposure by 3EV from the brightest area where texture should be preserved". Please see page 5 in the following PDF :!AkD78CVR1NBqlssSY7nCCsXBF13Iuw?e=e3Aoo9

In spot meter mode, I took a reading on the wall between the top of the tv set and the first picture frame.
Unfortunately, I did not realized that, yesterday, I had my camera set to ISO 125? And today's natural light was "lower" than yesterday?
So I opened up the diaphragm as much as I could, to "add 3 EV stops" to the exposure, going down to F/1.8.
At that point, I lowered the shutter speed to 1/20s.
In the viewfinder, the little "Exposure scale" went "past" the right bar.
I knew I was "overexposing" but maybe not by a full 3 EVs? But I think I had the "essential" of the technique right.
I did not want to shoot at slower speed to avoid moving during the exposure.

The result?
Judging by the new Histogram (DSC_1565.NEF), the exposure *was* increased as the whole histogram shifted right.

Question 1:
I am curious, at this point, why the scale extends from 0 to 64,000 (16bit) instead of only 16,000 (14bit)?
My camera bit depth is still 14bit.

Question 2:
I pasted the Statistics for each picture above the Histograms.
One thing I'm curious is why the Red channel maximum is reported as 15,783 yet when I look at the Histogram, the rightmost red pixels only extends to about 8,000? And why all four channels report the same max values? 15,873?

Question 3:
Please correct me if I am wrong but I get the impression, looking at the Green, Blue and Second Green histogram (DSC_1565) that I went as far as I could on the exposure without clipping? I mean, if the maximum 14bit is indeed 16,000 then the 1/20s f/1.8 exposure was pretty much ALL I could do?

Dear Roger,

Dear Roger,

A1: It's about dead space to the right, compare 


A2.1: Please switch on log mode for Y on the histogam ("Y-Axis Range" section, "Scale" drop-down), you will see the right side of the histogram more clearly.

A2.2: you have tiny amount of pixels clipped, please look at OvExp/UnExp statistics section in the main window; when clipping occurs, it is customary that all channels are clipped to the same value.

A3: yes, and the outlayers that are clipped may be specular highlights. You can put a checkmark into "OvExp" box in "Display" section of the main screen to see what are the offenders.

A3: I am delighted to

A3: I am delighted to discover the existing function for visualizing "clipped pixels" in the main interface!
I was telling myself, yesterday, "If only RawDigger had a way to visualize clipped pixels"....

Add new comment