[Photography] How to Read (and Use) Histograms for Beautiful Exposures

The post How to Read (and Use) Histograms for Beautiful Exposures appeared first on Digital Photography School. It was authored by Darlene Hildebrandt.

This article was updated in December 2023 with contributions from Darlene Hildebrandt, Anne McKinnell, Herb Paynter, Rick Berk, Simon Ringsmuth, and Darren Rowse.

Are you struggling to understand how histograms in photography work? Do you want to know how to read a histogram so you can capture consistently detailed exposures?

Before digital photography, you had to wait until you developed the film to know for sure whether you got a good exposure. Now, by using the histogram, this information is at your fingertips during most – or even all! – of the shooting process, as well as during post-production.

In this article, we’re going to look at everything you need to know to get started with histogram photography, including:

• What a histogram actually is
• How to understand the peaks of a histogram graph
• How to use a histogram to prevent overexposure and underexposure
• Histogram pitfalls and mistakes

So if you’re ready to become a histogram expert, then read on!

What is a histogram?

A histogram is a graph that represents the tones in an image: the highlights, the shadows, and everything in between.

Every image has a unique histogram, which is displayed on your camera and by most post-processing programs.

Why is a histogram useful?

In photography, a major goal is to capture a detailed exposure of a scene (i.e., a photo with well-rendered shadows, highlights, and midtones).

And while you can always check image exposure by looking at your camera’s LCD screen and/or electronic viewfinder, or by viewing your image on a computer, the histogram offers a more objective method of evaluating tones.

If an image has blown-out (detailless) highlights, this will be visible on the histogram; if an image has clipped (detailless) shadows, this will be visible on the histogram; if an image is just generally too dark or too light, the histogram will make this clear.

That’s why photographers love histograms so much, and why learning how to use a histogram is essential. If you can read a histogram, you can quickly and accurately check the exposure of your image while out in the field or when editing at home.

How to read a histogram: step by step

As I explained, a histogram is a graph – which represents the pixels in an image, like this:

Image by Anne McKinnell

The left side of the graph represents the blacks or shadows, the right side of the graph represents the highlights or bright areas, and the middle section represents the midtones of the photo. 

The graph peaks represent the number of pixels of a particular tone (with each peak corresponding to a different tonal value). So a peak at the right side of the histogram (such as in the example histogram above) indicates a large volume of bright pixels in the image. Whereas a peak at the left side of the histogram indicates a large volume of dark pixels in the image.

Here’s how I recommend reading a new histogram:

Step 1: Look at the overall curve of the graph

Is the histogram skewed to the right? Skewed to the left? Or just generally centered?

A left-skewed histogram often (but not always!) indicates underexposure, as the shot is full of dark pixels.

A right-skewed histogram often (but not always!) indicates overexposure, as the shot is full of light pixels.

And a balanced, generally centered histogram tends to indicate a beautifully detailed, well-exposed image, because the shot is full of midtones.

Step 2: Look at the ends of the histogram

A histogram with peaks pressed up against the graph “walls” indicates a loss of information, which is nearly always bad.

So check both the right and left ends of the histogram. Look for any clipping – highlight clipping along the right side, and shadow clipping along the left side.

What will a histogram tell you?

A careful analysis of a histogram will tell you two things:

1. Whether an image is broadly well-exposed
2. Whether an image has clipped tones

You can tell that an image is well-exposed if it’s balanced toward the center of the frame, with no obvious skew. Ideally, the graph is spread across the entire histogram, from edge to edge – but without edge peaks, which indicate clipping.

Here’s an example of a well-exposed histogram:

This is how an ideal histogram might look: evenly distributed and not up the sides, stretching across the entire graph.

If your histogram looks like the one displayed above, then your exposure is likely perfect and requires no adjustment.

However, if the graph is skewed to the right and/or includes peaking against the right end, it’s a sign you should reduce your exposure (try boosting the shutter speed) and retake the image:

And if the graph is skewed to the left and/or includes peaking against the left end, it’s a sign you should increase your exposure (try dropping the shutter speed or increasing the ISO) and retake the image:

Histogram pitfalls and mistakes

In the previous section, I talked all about ideal histograms and how you can use a histogram to determine the perfect exposure for a scene.

But while this is generally true, and the histogram guidelines I shared above are generally reliable, you may run into three issues:

1. Your scene may be naturally darker or lighter than middle gray

A well-balanced, unskewed histogram is ideal for images that include plenty of midtones and are generally centered around midtone detail.

But certain scenes just don’t look like this. For instance, if you photograph a black rock against a night sky, you might end up with a significantly skewed histogram, even if you’ve captured all the detail correctly:

This is a histogram for a dark subject. It is not wrong; it is just shifted to the left to represent the tones of the subject. This might be a dark rock at night, or a black cat on dark pavement.

And if you photograph a white tree against snow, you might get skew in the other direction because the scene is naturally lighter than middle gray:

This is a histogram for a light subject (e.g., a snow-covered valley) with mostly light tones in the scene and few dark areas (e.g., trees). See how it is shifted to the right compared to the dark subject? This is what you want, assuming your scene is mostly light-toned. If you change your exposure to keep the graph centered, you will end up with gray snow, not white snow.

So before you look at your image’s histogram, ask yourself:

Should my scene average out to a middle gray? Or should it have an obvious skew? Then use this information to guide your approach.

2. You may wish to overexpose or underexpose for creative reasons

Sometimes, even though an image is technically overexposed, underexposed, or clipped, it still looks great – so if you’re after a creative result, you don’t need to worry so much about an “ideal” histogram, assuming you know exactly what you want.

For instance, you might blow out the sky for a light and airy look, or deliberately underexpose for a moody shot; really, the possibilities are endless! Just remember to check your histogram no matter what and aim for a specific, deliberate result.

3. The dynamic range of the scene exceeds the dynamic range of your camera

While it’s good to avoid clipping, you’ll occasionally run into scenes where clipping is unavoidable, simply because the scene contains both ultra-light and ultra-dark pixels (e.g., a sunset with a dark foreground).

Here’s a histogram with this exact problem:

High contrast graph

In such situations, you’ll generally need to use a graduated neutral density filter to reduce the strength of the bright pixels, or capture several bracketed shots that you’ll later blend together in Photoshop. You can also embrace the clipped exposure (see the previous section on creative overexposure and underexposure) – though it’s often a good idea to bracket anyway, just to be safe.

Here’s an example of a scene that will likely go off the histogram at both ends, thanks to the bright star and the dark walls:

In the above shot, I’ve left the exposure as is, and I think the shot looks fine. But check out this image with bright windows and dark shadows:

Notice the skylight at the top of the roof is blown out, and the deep shadows have little detail.

Using advanced techniques like image merging and blending, HDR, or careful post-processing, you can compress the tonal range of a scene to fit within the histogram and get a result like this:

Notice how, in this image, the details have been retained in both the highlights and the shadows.

For the image above, I’ve used four bracketed images (taken two stops apart) and the HDR tone mapping process to prevent clipping.

Histogram examples

There is no such thing as a perfect histogram. It’s just a graphical representation of the tonal range in your image. It’s up to you, as the artist, to decide what to do with this information. Having solid blacks and bright tones (provided they are not blown out) is not necessarily a bad thing.

Let’s take a look at some examples of how histograms will look for different types of images:

High-key scene

When you have a scene that is high key, it has a lot of bright tones, and not so many mid-tones or blacks. When you are photographing a scene that you want to be high key, your histogram should be stacked up on the right side – but not going up the right edge. If you want your scene to be high key, but your histogram is showing a lot of mid-tones, your whites are probably going to come out looking more gray than you would like.

Image by Anne McKinnell

Histogram for the image above.

Low-key scene

A low-key scene is one that is dark, which you would expect when photographing at night. In this case, your histogram will be stacked up on the left side. You may have a spike on the left edge, which indicates solid blacks.

A low key or dark scene will have mostly tones on the left side of the histogram. (Image by Anne McKinnell)

Histogram for the image above.

High-contrast scene

A high-contrast scene is one where there are lots of very dark and very bright tones and perhaps not so many tones in between. In this case, your histogram will show data on the left and right, and not so much in the middle.

A high-contrast scene features extreme brights and extreme darks with little in the middle. (Image by Anne McKinnell)

Histogram of the high-contrast scene displayed above.

Low-contrast scene

A low-contrast scene has a lot of mid-tones and few bright or dark tones. Your histogram will have a bell shape.

Image by Anne McKinnell

Histogram for the low-contrast photo displayed above.

Again, it’s up to you as the artist to choose what to do with this information. You have to decide whether the information in the graph is what you want or not. It’s just another tool in your arsenal to help you transform your artistic vision into a photograph.

If you’re not happy with your histogram, use your exposure compensation to adjust the exposure by making the image darker or lighter. Or you may choose to affect the light on the scene instead by using a flash, a reflector, or a diffuser. The choice is yours.

Understanding the color histogram

You’ve probably noticed in the examples above that the histogram not only shows the tones in grayscale but also shows you colors. Yes, it’s possible to blow out a color! If there is one particular color that is very bright in the scene, sometimes that color will become so saturated that you lose detail. This commonly happens with red flowers, for example.

Image by Anne McKinnell

The histogram for the image above.

To combat this, you can slightly desaturate the color in post-processing to bring back some of the detail in the flower petals. The histogram above shows the increase in red tones toward the brighter end of the scale.

When to use the histogram

In the field, you can use the histogram in conjunction with Live View before you make an image (though you can generally turn on the histogram in the viewfinder if you have a mirrorless camera with an EVF). You can also see the histogram afterward when you review the photo on your LCD screen.

Either way, you must use the histogram to check your exposure while you’re in the field. That way, you have an opportunity to make another exposure while you are still out with your camera.

Another key point: Don’t rely on your LCD to give you feedback about exposure. It’s great for checking composition and focus, but not exposure. That’s because the brightness of your LCD doesn’t correspond perfectly with the brightness of your image since you can adjust the brightness of your LCD.

For example, you can brighten the LCD so you can see it more easily outside on a sunny day. But then if you don’t reduce the brightness and you look at the LCD at night, your images are going to appear super bright when they are not.

The histogram is also available to you while post-processing your image. Use it to see where adjustments need to be made and to ensure you don’t create areas that are too bright or too dark while processing your images.

Additional resources on histograms

This video does a great job of explaining how the histogram on your camera works and how to read it:

I do slightly disagree with his tip that having a mountain in the center is the best option, as it does depend on the subject and your artistic vision!

Histogram myths and misconceptions

There have been dozens of articles and videos published about the purpose and interpretation of the histogram in post-production. Unfortunately, many of them contain significant errors.

I’ve been adjusting images for decades – long before the histogram graph was publicly introduced! – and I’ve worked with it every day since it became popular, so I’d like to weigh in on rumors and clarify some facts. Once you understand the histogram’s primary function and limitations, you’ll find it to be a solid feedback resource.

Myth 1: The vertical lines reflect image contrast

FACT: The horizontal axis does reflect the image’s tonal range (from the darkest tones to the lightest), though the vertical lines reveal little about its contrast. Actually, the horizontal distribution is what reveals the overall contrast. Tones located mostly on the right reveal very light (or high-key) images, while tones favoring the left side are darker (low-key) images.

The extreme right side wall represents white and the extreme left wall of the graph represents solid black. The highest (vertical) peak of the graph merely indicates the highest ratio of pixels containing that particular color tone as it relates to the others. The lowest vertical level on the graph indicates the tone color with the least number of pixels in the image.

Myth 2: There is one best histogram shape

FACT: There are as many histogram shapes as there are images. There is no such thing as a good or bad histogram, and there is no such thing as an ideal histogram. Because these graphs reflect each image’s distribution of tones, you’ll be hard-pressed to find any two that are alike.

Only images containing mostly midtones will actually display a true bell (mountain) shape on the histogram as in the top left. (Image from Herb Paynter)

Myth 3: Histogram clipping is always bad

FACT: Depending on whether the image is high-key (medium contrast on a pure white background) or low-key (dramatic lighting with a black background), either side of the mountain may resemble a tonal cliff.

Real-life lighting dynamics make these wall-climbing graphs quite acceptable. Photos captured against white seamless backdrops are purposely exposed to produce a skewed histogram and a blown-out background.

The U.S. Air Force Thunderbirds screaming across a very bright mid-summer South Florida sky were captured against totally white clouds. The apparent clipping shown in the histogram accurately reports the totally white clouds. (Image by Herb Paynter)

Myth 4: A histogram should stretch from deep shadows to bright highlights

FACT: Real-life lighting doesn’t demand that every scene contain both deep shadows and bright highlights. Images are sometimes brightened or darkened unnecessarily; this is a typical rookie editing mistake.

Often, a lack of deep shadows or bright highlights establishes an emotional mood that would be lost if the images were over-corrected to produce a fully stretched histogram.

This low-key shaded scene contains very few highlights. (Image by Herb Paynter)

Adjusting the image destroys the rich colors of the low-key scene.

Myth 5: A histogram should have no gaps

FACT: The concern here is that any gaps or breaks in the histogram represent a banding effect, which looks bad. In truth, however, there are only 256 vertical bars presented in the histogram. Each horizontal bar represents less than one-half of one percent (0.4%) of the total tonal range (100% / 256 = 0.390625%). Even if a photo contains a very gradual change in tones across a wide area (like an unclouded sky), your eyes will only perceive “banding” if the JPEG image has been degraded by repeated “Save” functions.

JPEG images contain a maximum of 256 levels (8 bits) of tone between black (solid color) and white (no color). Once JPEG files have been opened and saved several times, the number of tone levels can become significantly reduced and tone-banding may occur.

These two histograms appear quite different, though the images are very similar. As you can see, “gappy” pictures rarely interrupt the flowing tones in a photo and can actually clarify detail. (Image by Herb Paynter)

Myth 6: A histogram shows every tone and color in an image

FACT: Each histogram does reveal the relative placement and distribution of all tones and colors, but due to the size of the graph, its accuracy is seriously limited. Since editing software histograms are based on a horizontal graph only 256 pixels wide, each representation is a basic overview at best. If the full range of possible colors were truly represented by a single graph, the chart would occupy the wall of a good-sized room!

Let me break down the numbers. This 256 pixel-wide graph portrays each image’s potential color range using an 8-bit (256 level) interpolation. This means that all 16.8 million possible colors are represented in a mere 256 horizontal point histogram. Tones change levels in 0.4% increments. The graph significantly exaggerates the difference between minor shifts in tonal value.

Image from Herb Paynter

Human eyes barely perceive a half-percent (0.5%) difference between tones, so 256 levels in a JPEG image provide the illusion of continuous tone. This means the histogram uses less than two vertical columns to represent a single percent change in value.

The two patches on the bottom (marked “57%” and “58%”) are separated by only 1% in tone, though your eyes may have trouble seeing the difference. The top band shows a gradual change between the two values.

What does all this mean? Quite simply, the histogram delivers a good estimation of overall tone distribution but cannot be relied on for accurate measurement. A few gaps in the graph will rarely be visible to the human eye.

How to read and use a histogram: final words

Well, there you have it:

A simple guide to reading and using histograms for beautiful exposures. No, histograms aren’t foolproof – but they certainly allow you to improve your exposures, and will significantly enhance your photos.

Now over to you:

What do you think about using the histogram in photography? Do you have any advice? How will you approach the histogram from now on? Share your thoughts in the comments below!

The post How to Read (and Use) Histograms for Beautiful Exposures appeared first on Digital Photography School. It was authored by Darlene Hildebrandt.

it's from Digital Photography School https://ift.tt/QsRWy7k
via Darlene Hildebrandt