Most DSLRs have great metering systems, but even a great metering system can fool you when you’re shooting in low light conditions. Many photos that seem well-exposed – at first – turn out to be underexposed or overexposed. That’s why your camera allows you to use a histogram alongside the image on your LCD screen.
This article will teach you all you need to know about the histogram, how to read it and how to correct the exposure based on the histogram data.
What is a histogram
A histogram is a graph that shows the luminance range in a photo (shadows, highlights and midtones). It basically “reads” the brightness distribution of an image with pure black on the left, pure white on the right and grey in the middle.
The next image shows an ideal (equal) distribution of the gray tones, where the medium gray falls right in the middle:
A histogram can help you determine if an image is properly exposed, which means it has shadows and highlights and most of the shot contains midtones.
How to read the histogram
The shape of the graph depends on the type of image and how you exposed it. The histogram contains three main areas: shadows (left), midtones (center) and highlights (right). Roughly speaking, brighter shots push the chart to the right, darker ones push it to the left.
A properly exposed image has a histogram spread all over, but there are also exceptions that only use one end or the other of the scale (e.g. a black cat in a dark place or a snowy scene). Here is how the brightness values are mapped out throughout the tonal range:
The height of the graph says how many pixels fall into this tonal range. It doesn’t really matter how high and low these peaks and valleys go in the histogram, what is most important is that it goes all the way across from left to right.
When pixel values are pushed to the top left / right indicate clipping in the highlights / shadows:
As you look at the histogram on your LCD, you will notice a blinking in overexposed areas, which means details are lost in those areas and you can no longer get them back. So it’s always best to underexpose and retrieve data later in post-processing. Clipping areas are marked by a tall peak on the right side of the histogram.
The most common uses of the histogram:
Most images tend to be slightly over or underexposed. It’s easy to say when a photo is severely overexposed because most of the image is blown out. In landscape photos, the sky in the background is often overexposed.
When the graph is pushed to the right side (pure white), it tells you there is a loss of detail in the image. Usually, overexposed images are pretty useless because anything that is blown out to pure white is lost forever.
When the chart is pushed to the left (pure black), it tells you that parts of the image are underexposed and that details are lost in dark/shadow areas. You can also tell that a photo is underexposed if the chart is on the lower right side of the histogram.
You also know your shot is underexposed if there is no line of the graph all the way to the right. Darker photos are easy to correct in post-processing by dragging the right slider in Photoshop’s Levels function to the extreme left edge of the graph and bumping up the midtones.
A balanced histogram has a mountain ridge-like shape where the pixels are spread throughout the chart. It’s interesting to notice the connection between the spikes in the graph and the areas of tone on in a shot. The higher the spikes, the more tones there are.
Choose your battles
Camera sensors will never match the dynamic range of what we see. The human eye can identify difference in tone over about a 10-11 f/stop range of brightness but a digital sensor can only recognize a 7-stop range with visible details.
The main difference between how we and the camera perceive a scene is how the eye adapts to light as it evaluates the scene. The pupils dilate when focusing on darker areas and constrict when looking at brighter ones.
Our brain arranges the snippets into a panoramic view that is seen as well exposed. However, the camera lens captures the entire scene at one aperture per shot. If you expose for the highlights, then the shadows are gone, compared to what is recorded by the eye.
So, in high contrast lighting like the midday sun, you must choose which part you’re going to expose for. If you pick the highlights, you’ll lose the shadows and if you expose for the shadows, then the highlights are lost. If you go for the midtones, you could still lose the highlights.
What is clipping?
Clipping is a sign that you’ve lost details in the shot as a result of washed out highlights or excessively dark shadows. You can recognize clipping by observing the left and right edges of the histogram – the high spikes that climb all the way up to the top.
The best thing to do is lower the exposure, but if your shot looks too dark when you do this, then the scene’s dynamic range is too great. The best way around it is to add fill light to the shadows and/or use scrims to lower the highlights.
You can get by with a little clipping, especially on the highlights of the sun reflected on water, a sunset or deep shadows. In all these cases, you won’t be able to prevent some loss either in your blacks, or whites, maybe even both.
Should you push the graph left and keep highlight details, or move it right and save shadow detail? There is no right or wrong here, it’s up to how you perceive the scene before you. If in doubt, take multiple shots and decide later.
However, blowing out whole areas like a person’s forehead looks terrible and often changes the colors. With practice you’ll get better and better at it and you’ll have to trust your visual response. The secret is to identify it and decide whether a loss of detail in that area is acceptable or not.
The RGB histogram
Apart from revealing the amount of tones recorded in a photo, newer camera models also display RGB or color channel histograms. These graphs display the same information as standard luminance histograms, but split the data to the three primary colors (red, green, blue).
The three colors are not recorded in equal amount each time you take a shot. Other colors are mixed by the image sensor and are processed to create different colors in your shot. The RGB histogram reveals the tonal values for each color, allowing you to see if one of the channels has lost data.
Usually, the red channel is clipped the most, while the information in the green channel is better preserved. If you only analyze the standard histogram, you may evaluate a shot as being correctly exposed, while some blue shadows or red highlights are actually lost.
This is especially important when your subject contains some reds that make the key object of your composition, like a parrot, for instance. As you can see in the image below, there is a good amount of details in the red feathers:
If the details would have been lost in the red channel, the bird’s feathering would have been a flat patch of red or magenta in your final shot.
When capturing scenes with vivid colors, always use the RGB histogram, because elements like flowers, birds and fabrics usually have a lot of details. If you clip the color channels, your shots will be oversaturated and flat.
Turn on highlight warnings
It’s not always easy to find the blown out areas in the histogram, but fortunately there is something that makes your job a lot easier: highlight warnings. When enabled, the blinkies will indicate the washed out areas and where you’re overexposing the image. It’s also useful in knowing how far you should go with the brightening.
To enable the flashing highlight warnings on a Nikon, preview the shot and hit the Up or Down buttons, near the “OK” dial, until you notice the highlights blinking. If you select the “Highlight” mode, the camera remembers your setting for your next shot. To enable blinkies on a Canon, hit the “Display” button until they appear on the screen when you preview the shot in full-screen.
Histograms vary depending on the scene you’re photographing and how well you expose it. You can’t have a good or bad histogram – it all comes down to what you want to accomplish. Sometimes you may want to trust your natural reaction to a scene more than the numeric values provided by the histogram.
You can correct the exposure to a degree in your image processor, but you should try to get it right in the camera or reshoot the image. If you check the histogram and see that your shot is overexposed or underexposed, you can use the in-camera Exposure Compensation function to solve the problem.
Increasing exposure moves pixels to the right end of the graph. Lowering exposure moves them all the way to the left. When the histogram is moved too far in one direction or the other and the lines don’t even touch the edges, then you can safely adjust your exposure to cover more of the tonal range and create a better final shot.
The key benefit from using the histogram is to see if any highlights have been lost. If you have blown out areas in your shot, you’ll see it as a tall spike on the far right edge of the graph. If you see clipping, lower the exposure to get the vertical line as close to the right but without touching the edge.
The histogram also comes in handy when it’s difficult to preview the photos on the LCD screen because of bright sunlight. Remember, the histogram values don’t have to be perfect, but when you see clippings on either edge of the chart, re-check your photos to see if there’s a loss of detail in important parts of the frame.
Working with the histogram is not as hard as it might seem at first and once you get a better hang of it, you’ll want to use in all your work.