Compensation; Plus or minus, when to use it and how much? A lot of articles, I've read of late, suggest that a lot of photographers, never use more than +1 to-1 for compensation. I suppose a lot of them use 0.0 then do their work in post processing. What are your thoughts on compensation?
Just do what you see you need to do.
If the image is darker than you want, then use + EV exposure compensation (or + EV flash compensation), to take the second try.
If it is too bright, then turn it down (- EV compensation).
This problem is shown at
How Camera Light Meters Work
On Nikon cameras (menu exceptions can be selected on the most recent 3 or 4 models), Exposure compensation affects both ambient light and TTL flash. Flash compensation only affects TTL flash.
With just a little experience (and thinking), we quickly learn to just simply see this in the scene when we first walk up, and know about how much compensation to dial in, in advance of the first shot.
The only one single reasonable use of the histogram is to judge brightness by judging how close the data comes to the right end side of the histogram. This does make the assumption that there are some bright areas or colors of the scene that ought to be that bright (it is not always the best answer, because scenes vary in content).
But for sure (no exceptions), if data is stacked up as a tall thin spike right at the right end, that is clipping (over exposure), and must be corrected with compensation.
To see this clipping, it is mandatory to watch the three individual RGB histograms.
The one single gray histogram is less than useless, it won't see the clipping.
Two types of Histograms
The one single gray histogram is Not real data. It is a mathematical manipulation simulating gray scale brightness (luminosity). But our color pictures are not grayscale, and the one single histogram is worthless for our purpose.