.

.

..

How to make a Simulated Infrared

Taking Photographs

(6 - The Evolution - Histograms)or Back to Table of Contents

On most digital cameras you will come across an option for using a histogram. These appear complicated at first but they are not. First a pixel is just the small dots that make up your image. It was called picture element (PIXture ELement) in the old days. Histograms are a simple way to see how your exposure has come out. Each image you take can quickly be calculated into a histogram using graphics viewing software most of the time. Now days many point and shoot cameras will show you how the histogram looks right when you take the picture. Use this to your advantage and you will better predict how the image will turn out.

First of all the word histogram basically means statistical bar-chart. These charts show you how many dark and light elements are in your image. More importantly histograms tell you when either light or dark are in the extreme or information is being lost that you need for a quality image.

Histogram of Hawaii Sunset image

Hawaii Sunset image

How do I use a histogram? Try this! Remember that right is bright with regards to the tiny pixels that make up your image. On the right side of the histogram is where the bright values will be calculated. While in the left side of the histogram the dark pixels will be calculated. In the middle is where the medium range values occur. These middle range values or Mid-tones are where the best image occurs. It's important to note that many histograms are shown as mere black and white graphs. Those are the same graphs without the fancy shading of the one below. These will help you visualize better.

Histogram of the same Hawaiian Sunset image

The main goal is to have all your camera settings take in light so that mountains and peaks are in the center of your histogram. They don't have to be pretty mountains and peaks so long as they don't get clipped off on either end of your histogram display. In this way you have gotten enough light to see what is in your photograph and enough dark so as to not have the opposite brightness in the image burn white-hot spots in the scene with light. Notice the histogram in the Hawaii sunset above. It's a fairly balanced image with pixels that are not too dark and not too much light. It makes for a pleasing image.

Morobay California coast image

Histogram of Morobay California coast image

When elements in your image are too bright then detail is washed out. This generally means it is overexposed. Details can't ever be retrieved because you might see it flowing off the histogram's right-side. Notice in the histogram above this is true but acceptable in this case. In this image the ocean fog has diffused the light and obscured the detail but if it were not intentional we would not want the bright pixels flowing off the right side of the histogram. This one I took on purpose for the fog effect. The rise in information on the right side of the chart clearly shows saturated areas of the image. Those elements go off the right side of the chart if only in a small amount. This means detail within that clipped portion of the image isn't retrievable. In this particular instance it's not a bad thing. Seeing detail in fog was not expected.

Suppose you took a picture of a shiny golden bell in a room with lights. If the surface has metal that reflects like a mirror then you will probably get some spots in the image that overflow off the histogram even if it's only a small amount. That is because any mirror like surface will reflect most of the light going to it right into the camera. If controlled this can be an acceptable appearance. However, when you are out in the sun at a concert and you take a picture of the band members things are different. Let's say their skin is glowing from the sun this is not acceptable as you can lower the exposure, use a neutral density filter (which is like sun glasses for your lens), or simply wait for a better shot when they're in the shade.

California Sunset image

Histogram of a California Sunset image

On the sunset over the mountain image above, the histogram shows dark areas at the left edge of the chart. This is an indicator information is present but in the shadows. In fact being at the far left with no clipping off the sides means the color is close to pure black. This can be seen as good given the information isn't clipped off the left side. While this is perfectly acceptable, we can brighten the dark sunset as done below. Notice how the dark parts of the histogram are no longer on the edge and we see detail in the mountain side. If the dark areas had been pure black such as a football uniform then keeping pixels to the far left would have been preferable. It is up to you as the photographer what the final image looks like. When you are photographing you can change your exposure settings to move the historgram mountain back and forth till you have your information centered. If you were going for an artsy effect of a washed out background then you might want to intentionally brighten untill your histogram went off the edge of the display. Remember that basically pixels flowing off the right side of the histogram means there was so much light that the camera was 'blinded' or oversaturated with light energy. Make sure if this happens that you really intended it to.

Brightened version of California Sunset image

Histogram of a brightened California Sunset image

Now try this mental exercise. Look at any image and then try to picture its histogram based on the light and dark values. Are some areas washed out as too light or do they vanished in the dark shadows? Now retrieve its histogram using any number of free image viewing software. Was your prediction close?

Let's go in reverse. Take a look at the histogram below. This histogram is from an image of three deer sitting on the lawn of an office building in downtown. What are your predictions about how this image will look? There are certainly some dark areas that could be shadow. They are not clipped so this is great. There are also some light areas that show light in the scene. The fact that you can see that so much variety across the histogram means there will be good contrast between objects. If we had a singular narrow peak on the histogram then the contrast or variety would be low. For instance in the Hawaii sunset notice all the similar colors made the histogram narrow. Now, Imagine how this might all come together and then click on the histogram to reveal the image.

Click the histogram to see the associated image

In some software packages you are given the ability to filter parts of your image by narrowing which parts are displayed using your histogram. The software may have two pull-bars with one on each side. In this way you can bring the one on the left toward the center to avoid showing dark pixels. You can do the opposite with white pixels. Usually these type programs are giving you the ability to control the contrast and stretching of what is displayed on your computer monitor.

Another item people wonder about all the time is what happens if you have peaks that reach the top of the chart. Is it bad to have information touch the top? The answer is no problem, since having more of one tone is not the issues so much as what the particular tone is.

Technical Mumbo Jumbo: (Histogram Breakdown)

Now that you know how to read a histogram how is it created? Remember that a histogram is a statistical bar-chart showing the colors in your digital image. When the dots or pixels that make up the image on your digital display are "turned off" then the color from it is black. Black is given the value of zero. Likewise, when all the colors on your digital image are "on" for one pixel then it is white and the highest value is 255. When the histogram is made it plots everything from the blackest pixels to the most bright white pixels. Similar tonal ranges are all grouped into one of the hills you see on the historgram. It may now be easy to see why having a large part of your image against the edge of a histogram would not be desired. Think of histograms as another tool to assist you for a better photograph outcome.

Break down of a histogram

<< Back or On to the next Step >>

Written March08 and updated April09 by - Tracy Lee Rose