- Pro Video
- Pro Audio
- TVs & Entertainment
- Optics & Outdoor
- Shop Categories
- Used Dept
One of the magical things that digital photography gives us is the ability to review an image instantly on the back of our cameras, or inside of an electronic viewfinder.
Frequently, one possible mistake that digital photographers make is when reviewing an image on their camera—they evaluate exposure based on the reproduced image. Why might this be a mistake? Well, both your camera’s LCD and EVF likely have adjustable brightness. Also, you might be viewing your images in bright sunlight or in the pitch black of night. Just as viewing a computer screen at work, or a television in your home, ambient light, screen brightness, and other factors affect the image that you view.
You should evaluate your LCD for composition and look to see if your depth of field, sharpness, and motion blur or freezing of action is what you were aiming for. Of course, if you have completely missed on your exposure, you may see that in an extremely bright or dark image.
So, how can we fine-tune our exposure if we cannot trust what our eyes see on the reproduced image? The answer is: the histogram. Luckily for us, the manufacturers of digital cameras have given us the histogram to use as a tool to evaluate exposure on a digital image more precisely.
The Camera’s Histogram(s)
|Luminosity Histogram||Color Histograms|
Most modern digital cameras have four histograms. The primary one is the luminosity histogram that shows overall brightness of a scene. This histogram usually has a monochromatic display—either white data on a black chart, or vice versa. The other three histograms are the color histograms, representing the red-, green-, and blue-sensitive pixels on the sensor. These histograms are generally displayed in their respective color.
(It may help to have your digital camera with you while you read this article, so that you can figure out how to find your camera’s histogram and interpret it. Digital cameras from different manufacturers have different menus, interfaces, and settings that govern when and where your histogram or histograms will appear. Consult your owner’s manual or an online source to utilize your camera’s histogram display.)
Now that you have located your histogram, how do you read it?
This example shows a nice mid-tone image. The dark area at the top appears as the peak on the histogram’s left side, but it does not extend all the way to the left edge of the graph.
How to Read the Histogram
First of all, there is an enormous amount of math behind the histogram. Luckily for the math-challenged photographers like me, you do not need to know any of it. For those of you who like numbers, I will attempt to sprinkle a few into the article, but know that the numbers behind the chart are inconsequential to reading it on your camera. If the numbers were critical to translating the data, the graphs would have numbers labeling them. However, there are probably more than a few websites and tutorials out there that dive into the mathematical abyss of bit depths, dynamic ranges, and some other sources of the math behind the histogram. Personally, I like to stay focused on the practical applications of the histogram. If you want to know more, hit me up in the Comments section at the end of the article, and we can nerd-out a bit! OK, back to the histogram…
This histogram’s horizontal (X) axis shows the luminance of the image from pure black on the left edge of the graph to pure white on the right edge. Growth on the vertical (Y) axis indicates the relative quantity of light for the given luminance. To illustrate the functionality using an extreme example, take a photo with your lens cap on and you will produce a histogram that has one spike, from bottom to top, on the left edge of the histogram. Opposite of this, take a long exposure on a sunny day and you will achieve a spike on the right. An image with a balanced exposure will show a “hump” in the middle region of the chart that tapers off as you move left toward black or right toward white.
This middle region of the histogram is for midtone luminance—the gray area(s) between black and white. You may have heard of “50 Shades of Gray.” Your camera, if it does 8-bit sampling, has 255 shades of gray. If you must visualize numbers, the X-axis of the histogram goes from 0 (black) to 255 (white) as you move from left to right.
To effectively use the histogram, you need to know three things:
Let us expand on these three things a bit.
How to read it: The histogram basically shows you the brightness of an image. If you take an image and see the majority of the body of the graph toward the right, this means you have captured a “high-key” image that may appear overexposed. Opposite, a histogram with the data showing mostly on the left is a “low-key” image that might appear underexposed. If you are making an image of a high-contrast scene (very dark and bright areas), you might see a U-shaped histogram. There are almost infinite combinations of light and dark that will register on the histogram.
Here is a low-key image and its accompanying histogram. You can see how the “weight” of the histogram is on the left side of the graph and it tapers rapidly as you move toward the center. The hump in the middle corresponds with the illuminated areas under the benches. Note that the data barely touches the left edge of the histogram. That would indicate clipping of the shadow detail.
The scene: I feel this is important, and something I do not see discussed very much on the topic of exposure and histograms. If you take a night photograph of a building and half of your image is a sky as black as a raven’s wing at midnight, you should expect to see a large spike on the left edge of your histogram. Conversely, if the sun is in your image, get ready for a right-side spike. If you are taking photos of a scene with dark shadows or bright sunlight, you need to expect to see this on the histogram and you should not be surprised when the spikes appear on the right or left edge.
Your goal: For those of you who read my article about “Understanding Aperture”, you are familiar with my thoughts about the term “proper” when it comes to exposure. Many guides will say that a certain histogram shows a proper exposure. “Proper” is subjective and photography is art. Art is subjective. If your artistic vision is a photo that is overexposed or one that is underexposed, and you intentionally cause that effect for your image, then “proper” is what you have achieved.
Here is an example of the U-shaped histogram caused by a high-contrast image. The band of LED lights in the middle of the frame and the bright foreground give the histogram a short peak at the right while the majority of the pixels are dark, and this is reflected by the mass on the left side of the histogram.
So, you take those three elements—knowledge of what the histogram is showing you, knowledge of the scene you are capturing, and knowledge of the final image you wish to produce—and then you look at the histogram and evaluate how and if you want to adjust your exposure for the next image by tweaking your aperture, shutter speed, ISO, or by recomposing the scene to reduce the amount of dark or light area in the image.
Beside personal artistic goals, the biggest reason to adjust the exposure, based on histogram data, is if your histogram tells you that you have experienced the horrible, yucky, and unfortunate phenomenon known as “clipping.”
Your camera’s digital sensor is much more limited than the human eye in its ability to gather information from a scene that contains very bright and very dark areas—a scene with a broad “dynamic range.” In photography, dynamic range is defined as the ratio between the maximum and minimum areas of luminance in a given scene. The camera will, unless you are manually controlling exposure, try its hardest to create an image that is exposed for the widest possible range of lights and darks in a scene. Because of the limited dynamic range of the sensor, this solution might leave the image with pitch-black shadows or pure white highlights.
A spike touching the left edge of the histogram means that there is shadow clipping. The dark areas of the image are outside of the camera’s dynamic range to the point that the camera cannot discern any information from those regions. The camera says, “By exposing for the major portion of the image, I have created an area of the photo so dark that I cannot see anything there, so I am going to call it pure black.” Spikes touching the right edge are representative of the camera saying the opposite, “When I expose for the major portion of the image, this one region is so bright that I cannot tell if there is an object in that region, so I will call it pure white.”
Here is a simulated (no, I did not blow the exposure!) high-key image. All of the data in the histogram is off to the right and—oops—lights of highlight clipping too!
Here is the same shot, as it was taken, which shows a low-key exposure. Note the spike on the right, corresponding to the illumination from the light bulbs, that shows up on the histogram and shows clipping. This is a good example of the sneakiness of highlight clipping, as there is virtually no buildup of brighter sections leading to the spike on the right edge of the histogram. We also have some shadow clipping as well. In scenarios like this there is really no way to avoid clipping on one or both edges. You just need to adjust your exposure to get the effect you want. Your own eyes will have the ability to see into the shadows while not getting blinded by the lights.
Clipping represents, unfortunately, the loss of data from that region of the image. Digital cameras are known for their ability to extract detail from dark shadow regions of an image, but once the histogram touches the left edge, that data is all but lost to a black abyss, and no amount of post-processing will pull detail from those shadows. Areas of pure white will also be just that—pure white—on your screen or on a print.
Shadow clipping is usually pretty noticeable on the histogram, since there is usually a buildup of data on the left side of the scale that reaches the left edge. Highlight clipping can be a bit more subtle, especially when taking night photographs, as only a small number of pixels in the image might be blown-out highlights. Keep an eye out for the narrow spike on the right edge of the histogram. Many of today’s digital cameras will have a flashing highlight indicator, or “blinkies” function that causes the blown-out light and/or dark regions of a previewed photo to blink so that the photographer can see, without the histogram, that areas are lost to black or white. A good practice is to use both the blinkies and histogram so that you can evaluate what regions of the photos have been clipped and whether you want to make adjustments to the exposure to prevent the occurrence in the next shot. Of course, you can always leave well enough alone and move to your next great image and accept the shadows or highlights. Sometimes, depending on the scene and your camera’s dynamic range, there will not be a practical remedy to the clipping; however, if you can adjust your exposure to avoid clipping, by all means, do so.
Remember that RGB histogram that I mentioned? Well, basically, it shows distribution and clipping in the red, green, or blue channels. Pay attention to the RGB histograms because you might see color clipping in one or more channels while the luminance histogram shows no clipping. This “hidden” clipping might be a big deal for your image, depending on your photographic vision for the picture.
Another thing to mention: the histogram and blinkies are usually based on a JPEG rendition of your image. If you are shooting a RAW file, your actual image will have a slightly greater dynamic range and, the clipping, if there is some, should be reduced.
In summary, the histogram is a remarkable tool to have in your tool box and one of many gifts that digital photography has bestowed on photographers of all skill levels. When reviewing your images, be sure to base your exposure adjustments moving forward on the histogram data, look for clipping and blinkies, and do not judge an exposure based on the preview image.