The Relationship of Light to Dark

For this lesson, we will manipulate only two of the variables that I mentioned at the beginning of Step 1: light-to-dark and in-focus-to-blur. Remember, it is the person creating the image who decides the journey that the viewer’s eye will take. And it is that journey that causes the viewer to see the story you wanted to tell.

How we control where the viewer’s eye will go is by manipulating variables such as focus, and light and dark. I contend that when we view anything at all, there is both an unconscious and a conscious element involved. First, our unconscious eye, or the anatomic structure that makes up the eye, scans in the predictable manner I described above. Then, the conscious eye, the mind’s eye, interprets the image seen by the unconscious eye. It is how we control the unconscious eye that determines how the viewer interprets the image. This is a theme to which I will frequently return.

In general, I like to begin manipulating light-to-dark, thereby exploiting the unconscious eye’s tendency to move from light areas to dark ones. For this specific image, I want the viewer’s eye to go first to the face, then to the torso, then to the rest of the image.

You must first decide what ratio or relationship of light-to-dark to create within this image. I wanted the face to be brightest, so I set it at 100%. As light’s circle of illumination increases, its intensity diminishes, so set the hair and torso at 50%. The background should be darker than the face, hair, and torso. The light on the background should go from light to dark as the eye moves from right to left. (I’ll explain why in a moment.) Set the right side at 25% and the left side at 0%.

After those values are drawn on the L2D IM layer, the image will look like this (Figure 1.3):

Placing Key and Fill Lights

If I had had all my lighting equipment at the photo shoot, I would have started by lighting the background, and I would have cross-lit it from left to right. Then I would have set key lights and fill lights. With a portrait, you usually want the viewer’s eye to go first to the subject’s eyes, so that’s where you should put the key light. Then you might put fill lights on the lips and to some degree, the torso.

Create a new layer and name it LIGHTING IM (for Lighting Image Map). Pick a color other than red from the Tools palette to draw your lighting choices. (I chose blue.)

Here is the LIGHTING image map (Figure 1.4) showing the notation of the percentages used: eyes 100%, face 50%, torso 25%, and background areas 50% and 25%.

The Illusion of Depth of Field

When shooting portraits, I find that a shallow depth of field, where the background is out of focus and only the subject is sharp, is visually pleasing. When you focus on the subject’s eye that is closest to the camera, the depth of field (the zone of acceptable sharpness) on the face will extend from the tip of the nose to a little past the ear. Generally that means shooting at f/5.6.

Create a new layer and name it D OF F (for Depth of Field), and pick a new color to use for your next set of image map notations. (I chose light green.)

In this case, the image was shot at f/6.3, with the model standing right against the wall, underneath the diffuser. The result? Too much depth of field, with everything in focus, including the background.

You now have a basic workflow to follow for manipulating the lighting and depth of field of the Challen Cates image in Photoshop (Figure 1.6). (Keep in mind that the values I chose to use are only approximations, and reflect relationships specific to this image.) You are now well on the way to a believable probability, the original vision of the image, from a completely flat-lit photograph.

Make the image maps invisible. They will be out of the way, but available when needed.

Step 2: Correcting Digital Sensor Color Cast

Now that you have an image roadmap, working from global to granular, let’s address the next biggest issue, correcting the image’s CCD/CMOS color cast. It is an important issue because it will affect all of the image editing choices you will make from this point on.

All RAW images, from any digital camera, exhibit a color cast as a result of the interpolation process that occurs when you bring that image into digital manipulation software such as Photoshop. The Challen Cates image has a magenta/yellow haze.

The most effective way that I have found to remove this type of color cast is to first define the black and white points of the image on which you are working. Finding the white point is a bit more problematic than finding the black point, but you are going to find them both by using a Threshold adjustment layer.

There are some rules that are important to know when removing the color cast caused from the interpolation of the data from a CCD/CMOS sensor. First, when looking for the black point, select your sample point from an area of “meaningful” black rather than using the first black pixel you see. If you select the very first black pixel that you see, it generally has RGB values that are R:0, B:0, G:0. If no information was recorded, no color contamination exists. What we are looking for is a black pixel that has RGB information in it.

Finding a white point is completely different. You do not want to select a white point from an area of “meaningful” white. Rather, you want to find the pixels that are closest to pure white, without actually being pure white. (A pure white pixel will have RGB values of R:255, G:255, B:255, which is of the same usefulness as a black pixel that has RGB values that are zero.) What makes finding a white point so problematic is that, much of the time, visible white and measurable white are two different things. (Measurable white, using the Threshold adjustment layer method, will always be biased to 60%G, 30%R, 10% B. Visible white generally consists of equal values of RGB and tends to be a lot bluer than measurable white.) In addition, there are instances when there is no “white point” and occasionally there may be aspects of the white point color cast correction you may not like, i.e. you may actually like aspects of the color cast. For these reasons, it’s a good idea to separate the black and white points into two curves adjustment layers; it gives you options.

How To Find Black and White Points Using a Threshold Adjustment Layer

You have successfully removed the CCD/CMOS color cast from the “white” aspects of the image, thereby eliminating the color contamination inherent in the process of converting RAW files to any usable file format.

Step 3: Merging Layers in “The Move”

You will now merge copies of the layers you’ve created so far into a single new layer, while preserving the individual work layers that you created earlier. Adobe Photoshop refers to this as Merge Stamp Visible. I prefer to refer to this as doing “The Move.” Note that you are merging the layers into one without flattening the image. This is an important distinction because if you just merge layers, you also flatten the resultant image thereby losing all the original layers. If you do this, you will have no exit strategy, no way to return to the image, and you will be unable to practice at practicing.

Make sure you are at the top of the layer heap by making the topmost visible layer active. For CS2 and above, press and hold Ctrl-Alt-Shift-E/Cmd-Option-Shift-E. For CS and below, press Ctrl-Alt-Shift/Cmd-Option-Shift, then type N, and then E.

You now have a base image layer on which you can start to work to make other aesthetically pleasing changes. Name it MASTER 1. “Save As” the file (Shft+Ctrl+S/Shft+Command+S) and name the file SHIBUMI 16BIT and save the image as a Photoshop document (.psd).

Step 4: Correcting for the Blue Color Cast of Sunlight

Wherever possible, I try to come up with the “Photoshop” way of doing things, and most of the time I succeed. But what makes Photoshop such a unique software package is that Adobe designed a way for us to use add-ons (called plug-ins) without messing with the core of the program. Some things can be done more easily or elegantly with third-party technology.

In my pursuit of a smoother workflow, I have become an advocate of the filters made by Nik Multimedia. For example, if you aren’t using their Skylight filter, I think you should.

A skylight filter can correct for the fact that shade and shadow light tends to be bluer than direct light, and direct light tends to be bluer than early morning and late afternoon light. The filter scans the image and determines how much, and where, red needs to be added to counteract any blue cast.

1.	Duplicate the MASTER 1 layer and rename the copy SKYLIGHT Note A quick way to duplicate a layer is to use the keyboard command Control+J for windows or Command+J for Apple. Although this is the keyboard command for duplicating a selection, you can also use it to duplicate a layer.
2.	Choose Filter > Nik Color Efex Pro 2.0: traditional filters > Skylight Filter. The Skylight Filter menu appears. The default setting (25%) is fine for our purposes, so click OK (Figure 1.8). Figure 1.8. The Nik Skylight Filter interface. [View full size image] First and foremost, you want to make sure that you are creating a believable probability. In the real world, shadows tend to be bluer or “cooler” than areas that are well lit. Areas that are “lit” tend to be “warmer” (red/yellow) than areas that are not. People look better “warmer” than “cooler.” Knowing this, what you want to do is create selective warmth in the image so that it follows the roadmap that you defined in your initial image maps (L 2 d L2D IM and LIGHTING IM layers). You will accomplish this is by using a layer mask (Figures 1.9 and 1.10). Figure 1.9. Before using the Skylight filter. Figure 1.10. After Skylight color correction.
3.	To selectively apply the Skylight filter effect, use a layer mask. To create a layer mask, go to the bottom of the layers palette. Holding down the Alt/Option key, click on the “Add Layer Mask” radio button. (This is the third radio button from the left.) This is the one-click way to create a layer mask and fill it with black.
4.	Make the L2D IM and LIGHTING IM layers visible so that you can analyze how to selectively warm this image (Figures 1.11a and 1.11b). Figure 1.11a. Image before painting in warm tones. Figure 1.11b. Image with image map before painting in the warm tones.
5.	Choose a very soft brush, roughly the size of the eye socket. Start with an opacity of 50% (press the number 5 key), and make sure the foreground color is white. Note The keyboard command to change a brush’s opacity is simple. Just type in the number value of the opacity percentage you want: 0=100%, 5=50%, 2=20%, 1=10% and so on. You now face a conceptually tricky task. Because you are working with a layer mask filled with black, but painting with white, you are dealing with a negative image as you would in a black-and-white darkroom. Therefore, you are working in reverse. You are going to be building up different levels of intensity of warmth (the “warmth” that you have just created using the Skylight filter), and since you want the eyes to be brightest, start by brushing that area at 50%. (Remember, all brushwork is cumulative.) When you’re done, this is the area you will want to be the warmest in the image. For the first part of the lesson, paint inside the lines on the image map. Don’t worry about being exact; just use the lines as an approximation (Figure 1.12a). Figure 1.12a. After the tones have been selectively painted in. The layer mask now has a whitish streak across it. This is the area you are revealing. Next, re-brush over the face at 50% twice so that the eyes will be at 100% opacity and the face at 75% (Figure 1.12b). Figure 1.12b. The layer mask.
6.	Select an opacity of 20%, increase the brush size one step by pressing the right bracket (]) key, and paint in the torso. Shrink the brush three sizes by pressing the left bracket ([) key three times, and brush in the hair. Now increase the opacity to 50%, increase the brush size four times, and brush the face and the area behind the subject.
7.	Lastly, reduce the brush four sizes and paint the area in which background light should fall off, as you originally mapped it on the image maps.

Unmasking Layer Masks and the 80/20 Rule Layer masks are a powerful Photoshop tool and an important way to control the aesthetics of an image. A layer mask works by allowing you to hide or reveal (either completely or partially) filter effects, adjustment layer corrections, or anything you want to selectively control. The best way to remember how to work on a layer mask is by remembering this simple mnemonic: “Black conceals and white reveals.” That means that black will block the visibility of the effect you are creating, whereas white will allow the effect to be visible. A good way to ensure an efficient work flow is to use what is referred to as the “80/20 rule.” If you want to reveal 80% and conceal 20% of the layer, create a white mask and paint with black to conceal. If you want to conceal 80% and reveal 20% of the layer, start with a black mask and paint the areas you want to reveal with white. Since you are working in grayscale when you are working on a layer mask, varying levels of black or white opacity determines how much or how little of the effect on that layer you allow to be seen. In other words, if you are working on a white layer mask, the darker the gray, the less that is revealed, and when you are working on a black layer mask, the lighter the gray, the more that is revealed. 0521 Rule I have found that there is a relationship between 100%, 50%, 25% and 10% opacity. The 0521 rule is based on this observation. The way the rule works is this: 50% of 50% is 25%. So if you’re working at 50% opacity, and you want to increase the effect of the thing on which you’re working, you will generally want to increase it about 25%. In the same way, 20% of 20% is 4%. So if you’re working at 20%, you will want to increase the effect about 4%. And if you want a little more, then 20% of 24% is approximately 5%, and so on. The keyboard command for changing the opacity of a brush is: 100% opacity is 0, 50% is 5, 20% is 2, and 10 is 1. (Hence the name 0521 rule). Being in Two Places at Once What follows is a way to have both the image on which you are working and the layer mask of that image visible at full size on the screen at the same time: Go to the Window Menu > Arrange > and select the New Window for (image name) from the bottom of the menu. This opens a new window of the same image that is “Live.” After each stroke, that window will be updated so you can see the effect of each change you make. Put the two images side by side (or if you use dual monitors, put one image on each monitor), and Alt/Option-click on the layer mask of whichever of the images has the layer mask you want to show. Now, depending on whether you want to work on the image or the layer mask, make that image active, and you will be able to see what happens as you do your brush work (Figure 1.13). Figure 1.13. [View full size image] Error-Free Layer Masks Now that you have a layer mask, wouldn’t be nice to have one that has no missed areas or unwanted overlaps? Here’s how. The important thing to remember, as I discussed in the 80/20 Rule, is that when you are working on a layer mask, you are working in grayscale. This means that when you change the opacity of white or black (depending on whether you are revealing or concealing), you are changing the density of the color with which you are working. That change in density manifests itself as shades of gray on the layer mask. As I discussed in the 0521 Rule, there is an inter- relationship between 100, 50, 20 and 10% when you are working with opacity on a layer mask. The problem is that when you brush over something, you might miss an area. If you brush the way most people do, brushing is cumulative. So it’s almost impossible to match up the areas you missed with the areas that you didn’t. Rather than try the conventional approach to brushing, treat the layer mask as a color, the color gray. Approach 1: Make the layer mask visible at full screen (See the Being in Two Places at Once section of this sidebar.) and make the layer mask active. Select the brush tool and set the brush opacity at 100%. Alt/Option-click on the area that you want to sample (the Alt/Option key brings up the sample tool eyedropper when you are in the brush tool), and brush over the area you missed or in which you have an unwanted overlap. (Remember, you are painting with gray.) Approach 2: Set the brush opacity to 100%. Alt/Option-click on the area that you want to sample. Select the Polygonal lasso tool. (The keyboard command is the letter L. If you want to scroll through any tool, hold down the Shift key and click on the tool’s letter until the desired tool appears.) Mouse click around the area and, when you have the desired area selected, Alt+Backspace or Option+Delete to fill with the foreground color. Approach 3: (The Peter Bauer Method) 1. In order to make the layer (rather than the mask) active, click on the layer thumbnail to the left in the Layers palette. 2. Layer > Layer Style > Stroke. (The default color is red, so stick with red as the color unless the image has a lot of red.) Increase the stroke width to 10 or 15 pixels. 3. Click OK in the Layer Style dialog box. 4. Click on the layer mask thumbnail to the right in the Layers palette. 5. Where you see blobs of red in the image, there are stray black or white areas, so (with the layer mask active) simply paint over the red spots with the appropriate color and watch them go away. 6. After cleaning up the mask, delete (or hide until next time) the Stroke layer style.

Step 5: Creating Selective Depth of Field

When I talked about creating the image map for depth of field, one of the issues I addressed was that I wanted the back wall to appear lit as if it was positioned well behind the subject. What I actually photographed was a subject standing up against the wall. If I could have cross-lit the background, I would have positioned the subject about 5 feet from the wall. To create a believable probability, I had to add the illusion of optical depth. You will do that by creating a layer of blur between two layers of sharpness.