Contents

Understanding Image Histograms and Curves

Histograms

Image Histograms

Color Channels

Histograms and Pixel Structure

Local Histograms

Using Histograms as a Scanner Tool

Prescan Histograms

Postscan Histograms

Evaluating Histogram Area

Image Exposure and Tone Curves

The Scanning Process

Changing Brightness and Contrast

Color Corrections

Examples

Summary

Interactive Demos

Setting Exposure

Setting Image Curves

Appendices

The Photoshop Levels Function and Curves

Why Is There No Luminosity Tool?

Average Skin Tone

Using Histograms to Track Scanner Performance

Further Information

Order Hardcopy Version

Home page

Color Corrections Using Curves

The proper means of performing image-wide color modifications is through curves -- not by using the sliders.

Modifying colors with curves is similar to modifying tones. When we speak of modifying tones it is in terms of lightening or darkening. For colors, on the other hand, we speak in terms of addition or subtraction. In addition, a modification of a color component -- red, green, or blue -- can also be expressed as an inverse adjustment of its color complement:

bullet

Adding/subtracting red is equivalent to subtracting/adding cyan

bullet

Adding/subtracting green is equivalent to subtracting/adding magenta

bullet

Adding/subtracting blue is equivalent to subtracting/adding yellow

For example, if your image has a cyan cast you correct it by applying a curve that adds red, its complement. Or to say that an image lacks yellow is equivalent to saying that blue must be subtracted.

The common practice in finding the correct color setting is to evaluate each correction quickly because the human visual system rapidly accustoms itself to erroneous settings.

Example of application of color correction curve function

Post Color Correction

After a color change is applied, a compensating adjustment of brightness may be required if the image's tonality is to be preserved, something you may be familiar with if you've developed color prints in the darkroom.  If the color correction curve was shifted upwards, which has the side effect of increasing brightness, the compensating RGB curve would be downwards, and vice versa.  Image tonality, or luminosity, is especially sensitive to changes in the green channel, which has a far greater weighting than blue or red in calculating luminosity (see Color Channels and Luminosity). 

see hardcopy for more on post color correction

Use Analog Gain to Control Exposure; Use Curves to Correct Tone and Color Balance

Analog gain is for adjusting scanner exposure, not for controlling tone or color balance.  Although analog gain affects color balance, it only does so as a side effect to setting exposure, its sole function.  The proper order of scanner operation is to set for optimal exposure, if necessary, through analog gain (a pre-scan operation).  Optimal exposure only coincidentally results in correct balance.  Then the image's tone and color balance are achieved through curves (a post-scan operation).  In addition, because it operates linearly like a slider, analog gain is a generally inadequate and cumbersome tool for controlling tone and color balance.  If you value your sanity, don't try to accomplish both with analog gain.  Keep the tasks conceptually separate and address each with the appropriate tool.

see hardcopy for more on where the BW points and curves should be applied

Modifying Brightness and Contrast  Scanning Examples