Shadow/Highlight Detail in Photoshop -- Part IV

Article and Photography by Ron Bigelow

www.ronbigelow.com

Photoshop CS or Photoshop CS2 Used in this Tutorial

Through all the tools that have been presented, two problems have remained constant: the small number of tones in the shadows and the poor shadow SNR. None of the tools previously covered have been able to overcome these two issues. Consequently, the results from these tools have been limited in the shadow areas. The two tools presented in this article overcome both of these problems. Due to this, these new tools generate far superior detail enhancement in the shadow areas than any other tool available.

Multiple Exposures

The basic idea behind this approach is that the camera is set on a tripod and multiple shots are taken, each at a different exposure. A typical example would be where two shots are taken. One shot is exposed for the midtones and highlights. A second shot is exposed for the shadows. An even more common approach involves three shots. One shot is exposed for the midtones, a second for the highlights, and a third for the shadows. This approach is very easy with cameras that have auto-bracket capability. The amount that the exposures are to differ by is set up in one of the camera's menus prior to the shot. At the time of exposure, the shots are rapidly fired one-after-another. With many cameras, the photographer needs to press the shutter only once and all three shots will fire. Once the images are transferred to the computer, they are combined with the use of masks to produce one final image that contains good detail in all tonal regions.

In essence, this technique gives the shadows extra exposure. In fact, this technique works best when the shadows are overexposed and the brightness of the shadows is turned down in the raw converter or image processing program.

This technique has a major impact on both the number of tones in the shadows and the shadow SNR. To understand why this technique produces superior results, it is necessary to go back to some theory. Table 1 (from Part I of this series) shows the number of tones in a JPEG image (before application of any tonal curve) from a camera that has a dynamic range of five stops. This table shows the number of tones that each tonal range, from shadows to highlights, would get if a single exposure were taken. From this table, it can be seen that the shadows, which got one stop of exposure, contain only sixteen tones.

If multiple exposures are taken so that one has additional exposure added for the shadows, the situation changes dramatically. If the shadows are given two stops of light (rather than the one stop that would result if only one exposure were taken), the first stop of light generates sixteen tones, and the second stop of light generates another sixteen tones. The shadows now contain thirty-two tones, rather than just sixteen. The number of tones in the shadows has doubled -- this significantly increases the amount of detail in the shadows. If the shadows are given another additional stop of light (so that the shadows get three stops), the third stop of light would add an additional thirty-two tones. This would bring the number of shadow tones to sixty-four. This is a quadrupling of the tones over the original sixteen.

Table 1: Distribution of Tonal Values for a Five Stop Dynamic Range Image (JPEG File) Prior to Application of Tonal Curves
Light Level
Tonal Values
Notes
5 Stops
128
Highlights
4 Stops
64
Three quarter tones
3 Stops
32
Mid tones
2 Stops
16
Quarter tones
1 Stop
16
Shadows
Things get even better when the SNR is considered. Table 2 shows the SNR from the darkest shadows to the highlights (indicated in stops of light). From this table, it can be seen that the darkest shadows have a SNR of only 1.85. Increasing the exposure to the point where the darkest shadows get a full stop of light increases the SNR to 32.65. Adding yet another stop of light improves the SNR to 50.87. This considerably improves the quality of the shadow detail.
Table 2: Simulation of SNR
  Darkest Shadows 1 Stop (Shadows) 2 Stops 3 Stops (Midtones) 4 Stops 5 Stops (Highlights)
Signal
50 Units
2,188 Units
4,375 Units
8,750 Units
17,500 Units
35,000 Units
Constant Noise
20 Units
20 Units
20 Units
20 Units
20 Units
20 Units
Variable Noise
7 Units
47 Units
66 Units
94 Units
132 Units
187 Units
SNR
1.85
32.65
50.87
76.75
115.13
169.08

Notes

1. The signal is for a pixel with a full well capacity of 35,000 photons. The darkest shadows are assumed to be slightly above the noise floor.

2. The constant noise is assumed to be 20 photons.

3. The variable noise is calculated as the square root of the signal.

Figure 1: Image with Dark Shadows

Theory is great, but how well does it work in practice? Figure 1 shows an image of tidepools reflecting the sunset. Multiple exposures were taken of this scene. In particular, we will look at two of the exposures. The tidepools are the key element here. They are in shadow, and it is necessary to show some shadow detail with good image quality. Figure 1 is too small to see much detail in the tidepools. However, this detail will be visible when the image is printed at 13'"x 19" or larger. By examining crops of the tidepools, the issue becomes more obvious.

Figure 2 shows a crop of the tidepools. This crop is from the image that was exposed normally (i.e., the shadows did not get any extra exposure). Problems immediately become apparent. The image has a major noise trouble. This is particularly obvious in the water. Instead of a nice smooth reflection, the image is dominated by ugly noise. Moreover, the rock has major degradation due to posterization. As a result, the rock takes on an artificial appearance. Figure 3 shows the image with an additional two stops of exposure to improve the shadow detail. An immediate improvement is obvious. The noise is mostly gone, and the posterization of the rock has disappeared.

Figure 2: Shadows with Regular Exposure
Figure 3: Shadows with Two Additional Stops of Exposure

Generally, the multiple exposures will be combined into the same image through the use of masks, and the image will be further edited in Photoshop using the tools previously covered (e.g., Curves and Blend modes). The final quality of the shadow detail will depend on which tools are used for the editing.

This technique can not be used in all circumstances. In situations where fast moving objects are being photographed, the technique is not feasible. However, in those conditions where multiple exposures are possible, the technique will improve the shadow image quality far beyond that possible with any of the other shadow enhancement techniques.

Merge to HDR

Figure 4: First Merge to HDR Dialogue Box

Merge to HDR (High Dynamic Range) is a Photoshop Command that allows up to seven exposures of the same scene to be combined automatically into one image that contains detail in all tonal regions (the exposures should be about one or two stops apart). Figure 4 shows the first of two Merge to HDR command Dialogue boxes. Figure 5 shows the second Merge to HDR command dialogue box.

Figure 5: Second Merge to HDR Dialogue Box

The Merge to HDR command is simply an automated way of using multiple exposures. Consequently, the Merge to HDR command has the same benefits as those covered in the previous section, and there is no need to review this command in detail.

One thing to keep in mind is that anytime a photographer uses an automated tool, she loses some degree of control over the image. This is true of the Merge to HDR command. Using multiple exposures and masks, as mentioned in the previous section, allows for more control but also takes more time than the Merge to HDR command.

Summary

With digital images, photographers have many options that can be used to enhance shadow and highlight detail. These tools range from simple tools that entail no preplanning and only small amounts of work in Photoshop (e.g., Shadow/Highlight) to tools that require planning before the image is taken and call for significant image editing (e.g., multiple exposures). It is up to the photographer to develop an understanding of how the various tools impact shadow and highlight image quality and to utilize that knowledge to select the best tool for the job.

Articles

Shadow/Highlight Detail -- Part III