This page shows the evolution of an image as it makes its way through my 4-step color correction process. Color correction is more about fine tuning the RGB contrast than it is about manipulating the color balance of the image. This image, however, was shot on amateur Kodak Gold 100 film that had a cyan cast. The color balance was adjusted with the scanner software. My scanner has a tool for recovering detail from shadows and highlights and this was used too.

Except for a reduction in size, this is the image just as it was scanned from a 35 mm negative with a Nikon Super CoolScan 5000 ED film scanner. It’s already a pleasant composition, but the image quality is going to get a whole lot better.

The first step is to use the Levels tool to set the black point and the white point of the image. Before this adjustment is made not all images span the full range from darkest black to brightest white. As you can see in the illustration above, Levels includes a graph called a histogram that shows the relative number of pixels at each brightness value. According to the histogram, this image already spans the full tonal range from black to white, hence no adjustment need be made. When I use Levels, I leave the gamma (gray level) set at 1.00.

The second step is to use the Curves tool, which does things that Levels can’t do. It takes a lot of skill and practice to get the best results from this tool. Above is an illustration of how I set Curves for this particular image. Every single image requires its own unique curve. A straight diagonal line represents equal input and output brightness, while deviations from the diagonal represent changes in brightness values.

Notice that in the case of this particular image I’ve flattened the contrast in the bottom eighth of the tonal range, while sharpening it in the mid tones and highlights. Your curves may vary radically from this one.

The third step is to adjust the color saturation of the image. Just the right amount of saturation will make an image come to life, but too much saturation will cause the pastel tones to block up and become solid. In the case of this image I cut saturation by 4%. When choosing slide film, there is a tradeoff between good shadow detail and bright colors. I get the best of both worlds by shooting mostly with Fujichrome Astia, a low contrast, low saturation film. Then I optimize contrast and saturation in Photoshop.

Some images benefit from corrections that can’t be made in RGB mode. Intensifying color with the saturation tool can decrease detail in an image. Fortunately, in Lab mode you can intensify color and increase detail at the same time. In Lab mode brightness and color are encoded separately. The L channel controls brightness, the a channel controls color balance between red and green, and the b channel controls color balance between yellow and blue. You can intensify color by using the Curves tool to increase the amount of contrast in the a and b channels.

The final step of quality enhancement is to sharpen the image. This is done with a filter that’s paradoxically called Unsharp Mask. This filter has three controls: Amount, Radius, and Threshold. The Amount setting will depend on how sharp or blurry the image is. The Radius will depend on the resolution of the scan, the higher the resolution of the scan the higher the number for this setting should be. For Kodak Master Photo CD scans at 2200 dpi, the optimum setting is 2.4 pixels. For 35mm film scanned on the Nikon 5000 ED at 4000 dpi, the optimum setting is 4.4 pixels. I always leave the Threshold set at 0. Photoshop CS2 introduces the even better Smart Sharpen filter, which works well with the same settings as Unsharp Mask. The image as sharpened with the Smart Sharpen filter is shown below.

Many experts say that sharpening an image should be saved for the final step of adusting it. If you’re scaling up an image’s pixel count, sharpen after scaling. However, when scaling down the pixel count I sharpen before shrinking the file.

It’s common knowledge that shooting with smaller apertures results in a greater depth of field. What’s not so well known is that as the aperture gets smaller diffraction increases, resulting in less sharpness where the image is in focus. Thus an image shot in dim light at ƒ8 or ƒ11 will often require a smaller Amount setting than one shot in bright light at ƒ16.

Unsharp Mask and Smart Sharpen work by increasing the contrast between adjacent pixels, in effect increasing local contrast rather than global contrast. Some say they only produce the illusion of increased sharpness, but when carefully applied I’d say they do the real thing. Too high an Amount setting will take away subtle tonal gradations and make an image less detailed rather than more detailed. Be sure to view the image at 100% magnification when applying these filters.

Once more, for the sake of comparison, here’s the uncorrected image just as it was scanned, except for the reduction in size. As you can see, color correction makes an image come to life in a way that a raw scan simply cannot match.

If you have Photoshop on your computer, it’s definitely worth your while to learn the process I describe here. The Levels, Curves, and Hue/Saturation tools can be found via the Image menu and the Adjust submenu. Accessing these tools via the Image menu will change the pixels permanently. If you want Levels, Curves, and Hue/Saturation to be reversible and forever editable, use the Layers menu and the New Adjustment Layer submenu instead.

Unsharp Mask and Smart Sharpen can be found under the Filter menu and the Sharpen submenu. These filters change the pixels permanently, so you should duplicate your background layer for sharpening and leave the original untouched.

This concludes the lesson on color correction with Photoshop. For further information on how to use these tools, see Photoshop Help.