Evaluating color and contrast

Understanding and judging color and contrast in your images are essential to producing good photography. Your lenses are where color and contrast begin and, to some extent, where they are controlled as they are processed as an exposure before being converted into a digital image.

Histograms are available for you to use as an evaluative tool on your camera’s LCD, as well as later in software image-editing applications. These graphs, which give you information about shadows, highlights, midtones, and colors in your images, provide a quick view as to your overall exposure quality and insight on how to adjust it, if necessary. There are two types of histograms: luminance and RGB.

LUMINANCE HISTOGRAMS

Every digital image has a tonal range that includes all the colors from the brightest to the darkest in red, green, and blue (RGB). Histograms more commonly come in a black-and-white mode where red, green, and blue are averaged; these are called luminance histograms. Your camera shows luminance histograms on the LCD, and your image-editing software often defaults to this mode.

The luminosity of an image refers to the overall distribution of brightness in a digital photo, and it takes into account your eye’s sensitivity to different colors (for example, red and blue appear less bright than green to the human eye). Luminance histograms calculate a weighted average of RGB colors in an image to create a single representation of the image that is simple to read and interpret.

RGB HISTOGRAMS

An RGB histogram shows shadows, midtones, and highlights for each of the three color channels (red, green, and blue) — essentially, it’s three histograms in one. Viewing a histogram in color, where each RGB color is shown in a range from 0 to 255 (assuming an 8-bit JPEG file), gives a far more detailed look into your image’s exposure. Yet while there’s more information in an RGB histogram, it can be more difficult to read because you’re trying to compare individual colors with a full-color image and there is triple the amount of data to interpret.

Using the typical sliders in your image-editing software, you can make on-the-fly adjustments to individual color channels, which can be useful for tweaking images where specific color attributes may be off. Figure 5-5 shows an image where the red histogram has been isolated and displayed in ACDSee, an image-editing software.

This colorful image of a traditional dancer in South Korea is represented by the histogram showing red shadows, highlights, and midtones. These tones can be selectively adjusted in software. Taken with a 1D Mark lln with a 70-200mm f/2.8L lens at ISO 640, 1/400 second, f/4.

While you can take time to fully read a histogram, frequently just a glance can help you quickly adjust an image. Generally speaking, a properly exposed image where all colors are included and the histogram is in black and white (such as what you see in your camera’s LCD screen) appears as a gentle bell curve where the midtones are higher (shown in the middle of the graph) and the shadows and highlights taper off on each side. Of course, when you begin to look at individual colors, such as in figure 5-5, there will be significant variations to the bell curve — even in a well-exposed image. However, strong spikes to the left or right may indicate under- or overexposure in an image — or, at the very least, significant portions of the image where this is occurring (which may or may not be what you want in a specific shot). And, while most typically a properly exposed image will have a histogram that is very reflective primarily of the main subject, there are photographs that lack enough midtones or a pronounced-enough subject that you won’t see a strong bell curve. This is where your subjective judgment as a photographer and artist is important; rarely would you rely solely on the histogram to decide if an image is correctly composed and exposed.

HIGH AND LOW KEY IMAGES

Images where lighting is shadowed and unevenly lit, such as in an artistic black-and-white nude, are called low key; these images emphasize the range of shadows and, partially, midtones of a histogram. Conversely, high key images are more evenly lit and feature stronger highlights and mid-tone ranges. They might be the result of bright sunlight as in figure 5-6 or they could be for a studio cover shot for a glamour magazine that features very few shadows and even bright lighting.

This high key image has few shadows and a large number of midtones and highlights (shown in ACDSee Pro histogram, but with the full RGB color view). Taken with an EOS-Mark IIn at ISO 160, 1/640 second, f/18.

For the most part, you can judge whether your image is high key or low key before you take the photo. In the studio, photographers specifically determine whether lighting should be high key, such as in the case of a family portrait, or low key, such as if they want to produce an artistic or more dramatic effect. Depending on your exposure settings, you can make an image darker or lighter and achieve, to some degree, effects of high key and low key; however, most often the amount and type of light in your photo determine your shadows, midtones, and highlights. Unless you have full control over the scene, such as in a studio, only your position and whether you use a flash determine what the image will be.

Looking at the histogram after you take a photo also tells you if your image is high key or low key. From there, you may be able to make adjustments. For example, if you take a photo in a program (automatic) setting and the histogram shows it to be lower key than you would like, you can shift to a manual mode and adjust the settings on your camera to shoot a brighter exposure.

Significant engineering in digital cameras has been focused on helping them to reproduce low key, low-light images where few areas of the image are overexposed. Low-light images possess more information that can be adjusted — especially if shot in RAW where there is a very detailed tonal range — than do images with overexposed areas. You can always take information out of an image, but you can’t add it in. Areas of overly highlighted images, where there are solid-white sections, are referred to as being blown out.

While underexposing images is generally safer than overexposing, you can of course take it too far; I generally don’t suggest underexposing more than one stop in situations where you’re uncertain about the light. Too much underexposing can result in a noisier image when it has been edited to a correct exposure, so you’ll want to be careful.
A histogram can help you by showing an area has too much data on the far right, which is the highlight section — meaning you may have at least parts of your image that are overexposed. You can also turn on the highlight alert feature in your Canon dSLR, which shows you over-exposed areas of your image in the LCD preview by flashing those parts of the image. Unless you actually want that part of the image to be overexposed, this tells you that you should manually adjust the image to be lower key by reducing the aperture size or increasing the shutter speed (or both); you could also lower studio lights if you are shooting in a controlled-lighting situation. While some blown-out areas are acceptable in certain images, such as in a reflective area on water or the center of a streetlight, determining the correct amount is a subjective artistic decision made by you.

CONTRAST

A logical aspect of a histogram, because it shows highs and lows of shadows, midtones, and highlights, is that it also displays contrast, which is the difference between dark and light areas in an image. A low-contrast image has a more narrow histogram curve and includes flattened areas, while a high-contrast image has a broader curve. Histograms are complex, of course, so you may have an overall low-contrast image with one part of it that is high contrast, or even have several areas with high contrast even though the rest of the image is low contrast.

Figures 5-7 and 5-8 show the same scene taken with slightly different exposures. After taking the shot in 5-7 and viewing the in-camera histogram, it seemed to not have enough contrast. So, I adjusted the exposure from 1/125 second to 1/250 second.

These dried fish, shot in a Kuwait marketplace, were a bit overexposed with too little contrast, as shown in 5-7. After adjusting the exposure to 1/250 second to add more shadowing, it had better contrast. Photos taken with a 1D Mark lln, 24-70mm f/2.8L lens at ISO 640, f/5.6. Figure 5-7 at 1/125 second and 5-8 at 1/250 second.

Reading a histogram is like following a GPS unit in your car: You can’t just blindly use it without giving some thought to the real world. You may have an image that shows low contrast or odd peaks in shadows and highlights, or that seems too flat, and yet artistically it may be just what you want. However, you should be generally familiar with the parts of a histogram that relate to a specific image, as shown in 5-9, which connects related areas of the histogram to specific parts of the image. Using the histogram and looking at your photo, you can decide how to adjust subsequent images and determine which areas of the image have more or less contrast, and whether the shadows, midtones, or highlights need to be emphasized or lessened.

The arrows connect the related points of the histogram to the image. You can see how shadows, midtones, and highlights are represented differently, and how you might interpret the histogram when taking a photo, and then make adjustments for the next shot. Photo taken with a 1D Mark II and a 70-200mm f/2.8L lens at ISO250, 1/250 second, f/7.1.

Category: Science of Lenses

Comments are closed.