Follow Ron on Twitter

Another common way to use Curves for managing contrast is through the use of the Eyedroppers. The Eyedroppers are in the lower right hand corner of the Curves dialog box (see Figure 1). The left Eyedropper is the Black Eyedropper and is used to set the black point or shadow values. The right Eyedropper is the White Eyedropper and is used to set the white point or highlight values. The middle Eyedropper is the Gray Eyedropper and is used to set neutral gray in the image. In the Eyedropper approach, the Black and White Eyedroppers will be used to set the shadow and highlight values respectively.

One of the advantages of the Eyedropper approach is that the shadow and highlight values can be made neutral. The drawback is that forcing these values to be neutral can sometimes cause color shifts in the rest of the image. It is therefore necessary to track the color values at three tonal locations: the shadows, the highlights, and the midtones. This is easily done with the Color Sampler and the Info palette. These are used to put three color samplers in place. The first two color samplers are set down in the shadows and highlights as was done in the Info Palette Approach. The third color sampler is set down in an area of midtones. This area should be as neutral as possible.
Figure 2 shows an image with the three color samplers. One point that needs to be noted is that this image has a slight, blue color cast because it was shot in the shadows. This will be addressed during the editing. It should also be noted that there are no perfectly neutral highlights in this image. Figure 3 shows the histogram of the unedited image, and Figure 4 shows the info palette with the color values of the three color samplers.












The Info Palette (see figure 14) shows the result of this work. Color sampler #2 originally had values of 225, 229, and 236. The values have been changed to 254, 254, and 253 (a neutral highlight).
Now, the Info palette shows a couple of very interesting pieces of information. For color sampler #2, as just mentioned above, the color values before editing were 225, 229, and 236. Before editing, this highlight was not neutral. There is more blue than the other two colors. This is not surprising as this image was shot in the shadows (the light in shadows, generally, has a blue tint). Thus, the image, as shot, had a blue tint. The values for color sampler #2 after the editing are 254, 254, and 253. Thus, after editing, the highlight values are neutral. This has caused a color shift in the highlights. In particular, the amount of blue has been reduced compared to the other color values. Color sampler #3 shows that this color shift has not been restricted to the highlights, but it has also occurred in the midtones. Before editing, the color sampler #3 (midtone) color values were 159, 157, and 158. This was a fairly neutral point. After editing, color sampler #3 has color values of 179, 173, and 169. This point is no longer neutral. As with the highlight values, the amount of blue has been reduced.
This color shift is clearly seen in Figures 15 and 16, which show the original image and the image after the shadow and highlight values were set.
One of the advantages of the Eyedropper approach is that it gives neutral shadows and highlights (if the Color Picker is set up properly). However, Figures 15 and 16 show that forcing the shadows and highlights to be neutral can cause color shifts. When comparing the two images, it is clear that a color shift has occurred in the second image. Now, in this case, the color shift worked to the photographer's advantage. The color shift got rid of the blue tint in the original image and allowed the natural, warm tones of the old wood to come forth. This saved the photographer some work as an additional color correction step will no longer be needed.
However, in other images, color shifts could result that will be to the detriment of the photographer. Undesired color shifts could be introduced that would need to be removed by a color correction step.



Some additional contrast will further improve the image. The contrast adjustments are made in the same manner as with the Info palette approach used in Part II of this series (see Shadow Values, Highlight Values, and Contrast: Info Palette Approach).
Figure 17 shows the final Curves adjustments and Figure 18 shows the final histogram.


Figure 19 shows the Info palette after all of the adjustments have been made. It can be seen that the Curves adjustment slightly changed the values for the highlight values. This could be readjusted with the upper end point on the curve. However, for this image, the changes are slight, so no further adjustments will be made. Clicking the OK button closes Curves.
Figures 20 and 21 show the original image and the image after all the adjustments were finalized.



The previous techniques all used the RGB channel. Thus, all three colors were adjusted at the same time. This works well for most images. However, sometimes it is desirable to adjust each color channel separately. This is most often the case when the color histograms have significantly different shapes. In this situation, the individual color channel technique is an option. Figure 22 shows an unedited image that falls into this category. Now, the first thing that should be noticed is that this image is underexposed (this was due to wind that forced a higher shutter speed than the light level would normally require). Figure 23 shows a histogram of this image. The histogram clearly shows the underexposure. It also shows that the color channels have different shaped histograms that start and stop at different points. Normally, I would not attempt to recover an image that was underexposed this much as the image quality is usually not up to standard. However, I like this image so much, let's see what can be done with the Individual Color Channel approach.



To start the editing, Curves is launched.

Next, the Channel pop-up is accessed (see Figure 26), which reveals the three color channels. For this example, the red channel is selected first. The red channel is now displayed (see Figure 27). All edits performed will affect only the red channel.

The shadow and highlight values can now be set similar to the way they were set in the eyeball technique. The only difference is that the edits only affect one color channel at a time.
To set the red shadow values, the lower left end of the curve is dragged to the right while the histogram and image are checked to see the effect of the edit. For most images, the left end of the histogram should just start to touch the left side of the histogram display area (this is correct for images that have dark shadows; for images without dark shadows, the histogram should not be moved so far to the left).
To set the red highlight values, the upper right end of the curve is dragged to the left while the histogram and image are checked to see the effect of the edit. For most images, the right end of the histogram should just start to touch the right side of the histogram display area (this is correct for images that have highlights; for images without highlights, the right side of the histogram should not be moved so far to the right).
Figure 28 shows the Curves Dialogue Box for the red channel with the adjustments for the shadow and highlight values. The upper end of the curve has been moved in rather drastically. This is required because of the underexposure condition. Properly exposed images will not require so extreme of an edit. Figure 29 shows the histogram for the red channel after the adjustments.









