How To Read A Histogram
Photographic histograms are a graphic representation of an image’s exposure. More specifically, they plot tonal values onto a graph. What are histograms useful for? They quickly provide photographers with a more accurate understanding of exposure. But how is this useful, you might ask, since we can all see the image right there in front of our eyes? Actually, this is exactly why a histogram is useful. The human eye can be fooled, and it’s especially susceptible to being incorrect when the lighting in which we’re reviewing a digital image is especially bright or dark.
For instance, if you’re photographing at night, you might review an image you just captured by looking at it on the back of the camera LCD. It looks bright and glowing, so you assume the exposure is fine. But in the dark, almost any image would appear bright and glowing. Are those objects depicted in the frame actually exposed correctly? Are they perhaps too dark or too light? The histogram will tell you.
Another instance of histograms being crucial is the exact opposite situation when you’re working in very bright light. We’ve all had experiences on sunny days where it’s next to impossible to see the camera’s LCD accurately. Your shot could be overexposed, underexposed, or exactly right—but you can’t make that judgment by eye. So, instead of guessing, you rely on the histogram and know exactly where the tonal values in an image lie.
Even on the computer, you might be looking at an image and thinking, “Something’s not quite right.” A glance at the histogram can explain: the image is too bright, or too flat, or too contrasty. It’s all right there in the histogram.
So does this mean that there’s a “correct” way that a histogram should look, a standard to which we should always strive? No, there’s not. Because depending on the tones in an image—whether the subject is a white cat in a white room, a black cat in a black room or anything in between—the “correct” histogram will look dramatically different.
For instance, if an image is low key—the aforementioned black cat/black room combo—most of the tones will be dark so the corresponding peaks on the histogram will be toward the left of the frame. In a high-key image—white cat in a white room—those peaks should be toward the right. If you’re photographing a white cat in a white room and the histogram peaks are in the center of the frame, you know your image is underexposed and all that white stuff might actually appear gray.
The first step in understanding a histogram is to know how the tones are mapped. It’s not a pixel for pixel representation of the sensor, rather it’s a bar graph. Each horizontal step across the X-axis represents an individual tonal value, from 0 to 255 (black to white). The farthest left edge of the histogram represents pure black, and the farthest edge on the right of the histogram represents pure white. And each pixel in between, moving horizontally along the X-axis, represents one step in density from dark to light.
So knowing what the histogram’s X-axis represents is step one. What does the Y-axis—or the height—of the peaks on the graph represent? This is a way to quantify the number of pixels of a given value. The taller the peak, the more pixels in the frame are of that particular brightness. A tall spike means there are a lot of pixels of that tone, a shorter spike means fewer pixels of that tone. No spike, or zero information on the histogram in that area, means there are no pixels of that particular tonal value.
This last bit is important because it means we know that a histogram that touches either side of the graph is more likely to be throwing away information. If the graph peaks to the left of the frame, at the very edge, then the darkest tones in the image are pure black. And if it peaks at the right edge, the darkest tones of the image are pure white. And if that peak appears to be cutting off your bell curve of tones smack dab in the middle (or close to it) you’ll know that you’ve distinctly over- or under-exposed the image because you’ve thrown out some of the data, over the brink at either edge of the histogram’s frame.
If you think of that mountain peak of tonal data, its placement from left to right on the histogram shows whether the image is mostly made up of dark tones, mostly light tones or mostly middle tones. If it’s cut off at either edge, all of that data that might otherwise have been subtle distinctions in shadows or highlights, it’s all clipped off and becomes pure black or pure white. The moral of the story is, don’t let your histogram spike at the edge of the frame. Keep that bell curve contained within the frame of the histogram and you’ll avoid throwing away valuable image-forming information.
If a histogram has a gentle curving shape, without much “rise” on those mountain peaks, you’re probably looking at an image that’s flat and lacks contrast. If the histogram shows, on the other hand, peaks on both the left and right of the frame, that’s a high-contrast image. The spike to the left is dark tones, the spike to the right is highlights and the lack of much in the middle is the missing midtones. Most scenes have a good balance of shadows, midtones and highlights, with more in the middle than at the extremes.
Some cameras and imaging applications will divide histograms into multiple channels representing color information. So not only can you see the overall luminosity of an image represented graphically, you can visualize whether particular colors are spiking appropriately or not.
The ultimate takeaway to understanding histograms is to know that the graphic is simply plotting pixel values from left to right on the X-axis, depicting everything from shadows through midtones to highlights. If your subject is made up of low-contrast midtones and you want it to appear that way in the photograph, you should hope to see that reflected in the histogram shape. If your subject is predominantly light or dark, these too should match the values display on the histogram. If they don’t correspond appropriately, adjusting the exposure will make an instant change and you’ll see it in the histogram. (This is true, too, when working on the exposure of an image in Lightroom or Photoshop.) Opening up the aperture, lengthening the shutter speed or cranking up the ISO will provide more exposure and move the peaks on the histogram to the right. Stopping down the aperture, shortening the shutter speed or lowering the ISO will underexpose and move the histogram peaks to the left.
Copy and paste this URL into your WordPress site to embed
Copy and paste this code into your site to embed