The image above depicts three similar images, along with a graph : Histogram. Before plunging into the details of what a histogram is and what it does, I would like to bring it to your notice that the histograms are something which we do encounter frequently while editing pictures, thus making it necessary to understand the concept of histograms. I would try to explain the technical details in least boring way possible for me, but still pardon if the effort goes hay-wire. The following is the image of Adobe Photoshop Lightroom 5.0 depicting the histogram as an essential past of the software:
The histogram is a simple graphical representation of intensity of pixels versus the number of pixels of a certain intensity in any given image. This determines if the given image is properly exposed or not. Any photograph may be under-exposed, over-exposed or properly exposed depending on how close the peaks in the histogram are to the mid-tones of the image. First image of this post depicts the same. We would go to details of usage of histogram, but before that, what exactly is a histogram??
A diagram consisting of rectangles whose area is proportional to the frequency of a variable and whose width is equal to the class interval.
Ehh!! Am I the only one confused? Why would I need the dictionary explanation? What do we do using a histogram? Have a look!!
A histogram is the pictorial representation of each and every pixel in an image with respect to their exposure values. The left side represents the darks or the shadows, while the right side depicts the whites or highlights and the middle section is for mid-tones, the properly exposed zone. The horizontal axis determines the exposure and the vertical height determines the number of pixels in the image of that certain exposure value. The area under the graph would hence determine the total number of pixels in the given image.
Note: A picture might be under-exposed and still look good, that depends totally on the photographer’s artistic instinct. Here exposure means closeness of the image pixels to the mid-tone levels.
We may learn many things about an image just by analyzing the histogram. We can tell if the image is properly exposed or not by checking if the histogram reaches the edges properly without any sudden peaks near any of the edges. If there is a gap on any of the edges, ie, the histogram doesn’t reach the edge itself, we may adjust the histogram such as to touch the edge without losing any significant detail.
The histogram shows that the given image was properly exposed. The peak is concentrated mostly near the midtones area of the histogram (the central part) and decreases smoothly to about 0 as it approaches both the ends.
The peaks of the histogram are concentrated near the left edge of the image, showing majority of the pixels are less exposed than the mid-tone region. The image, of which this histogram is, is as a result underexposed. Also, because towards the right end of the histogram, there is no pixel intensity shown, it would be safe enough to clip that part to average the histogram towards the mid-tones, without losing any significant details.
The peaks of the histogram are concentrated near the right edge of the image, showing majority of the pixels are much more exposed than the mid-tone region. The image, whose histogram this is, as a result is overexposed. Also, because towards the left end of the histogram, there is no pixel intensity shown, it would be safe enough to clip that part to average the histogram towards the mid-tones, without losing any significant details. This will eventually give a picture that is better exposed.
Then comes a very different sort of histogram, one which you cannot say is under-exposed or over-exposed, just by looking at it. It has peaks on both the ends, and most commonly it denotes any image which has high contrast. Such image have both set of features, bright highlights as well as dark shadows. Thus they yield to a pattern similar to as shown above in the histogram.
There are a few ways to recover this type of loss in detailing due to uneven distribution of pixel exposure in various parts of image, including image blending, HDR photography or editing in software like Lightroom or Photoshop. HDR and image blending usually average outs the range of pixel into a much uniform distribution thus yielding an image with much more details in both shadows and highlights.
Also, most of the cameras usually have the feature of displaying the histogram of image right after processing it. Although not accurate, because it is the histogram of the jpeg format of the image, but it gives a fair idea if the image is properly exposed or not.
I would like to add to this that having an image under-exposed or over-exposed can sometimes be the choice of the photographer himself/herself. I talked about a general term : Histogram. It is totally upto the choice of the photographer if he deliberately chooses the image to be under/over-exposed. Creativity Hold No Bars!!
“The best style is the style you don’t notice.”
– Somerset Maugham