Classic Luminosity Masking
Since writing the first luminosity mask tutorial I’ve found countless uses for them and there are always new options to explore. For new users, though, it can sometimes be a bit daunting to know where to start. Understanding the basics of using luminosity masks can be a good stepping stone to more complex applications. A recent image provided a straightforward example of why luminosity masks can be so valuable. I’ll use it to review some basic concepts both in words and pictures.
Here are three important things to remember about luminosity masks:
- They select tones in the image, not individual elements. It helps when using them to start thinking “tonally” about what will be selected with a luminosity masks instead of trying to use them to make a precise selection of a specific part of the image. Luminosity masks work best in situations where tonal differences are well-defined instead of in situations where there are obvious pixel edges.
- The edges of luminosity masks are perfectly feathered for blending. These masks are created from the brightness values of individual pixels. Just as a photograph is a continuous-tone image, luminosity masks provide continuous-tone blending. Sometimes this feathering can be a bit too perfect, especially in the initial masks, bleeding an adjustment into even weakly selected tones. But it’s easy to narrow the tonal range selected using techniques in the original tutorial or using calculations for 16-bit luminosity masks. Some feathering is necessary and highly desirable for insuring perfect blending of any adjustment or other Photoshop maneuver into the rest of the image.
- They are incredibly precise when properly used. Tonal selection and perfect feathering make it possible to use luminosity masks to make extremely targeted adjustments. Painting through an active luminosity selection is perhaps the best way to take advantage of this since multiple brushstrokes can be applied to the same area, slowly building up the desired effect, while also insuring it blends flawlessly into the image.
The image below of a cloud from a clearing storm against a mountain background is the example I’ll use to illustrate these principles. This is the nearly finished version of the image. The main problem left to fix is that the cloud isn’t as well defined as it could be. It’s an accurate depiction of what was captured, but it lacks good textural quality because the tones, especially in the brightest areas on the left, are too close together to provide meaningful definition for the viewer. The cloud still needs some work to bring out the tonal texture that is present, but hidden in the brightest tones.
But how to isolate the cloud for additional development? It doesn’t have any good edges. The wind-blown wisps along its outer portions would be a challenge to select with any standard Photoshop tool. And the hard edges created by these tools would be equally difficult to feather into the image. However, thinking tonally, the cloud is distinctly separate from its background, so the tonal selection provided by a luminosity mask would be ideal.
While there are an infinite number of luminosity masks, it’s usually easy to spot the right one. It’s the mask that is whitest in the areas of the desired adjustment and very dark in areas where no adjustment is required. In this case, it was the Lights-3 mask (shown below). With this mask, the cloud clearly stands out from the background.
The red-overlay “view” mode (below) shows even more clearly how perfect this mask is. Not only is it more selected in the whitest areas of the clouds (darker red), but it also feathers very nicely to the edges of the cloud (lighter red). And, to top it off, there is no red in areas immediately adjacent to the cloud. So an adjustment using this mask will affect the whitest cloud areas most and feather perfectly to the wispy edges. There will also NOT be any haloing around the cloud caused by a poorly feathered selection. The luminosity mask will confine the adjustment to only those pixels where it’s needed. NOTE: The snow-covered mountain tops and upper cloud are also showing a small degree of selection (pale red) but it is a simple matter to paint these areas black in the final mask to exclude them from the adjustment made to the cloud.
Once the mask is decided on, it can be added to an adjustment layer. For this image a Levels adjustment works well. It takes the targeted tones and easily adds contrast to create more texture in the cloud, especially the blob-like, white areas. The Properties panel for this adjustment is shown below.
The image below shows the results. Rolling the mouse over this image shows what it looked like before this adjustment. (The rollover might not be visible in the email feed, but is visible on the blog website.) It’s easy to see how the cloud (and only the cloud) has had its texture significantly improved. This result displays one of the ideal qualities of luminosity masks, namely that they can separate tonal differences present at the pixel level which is nearly invisible to the eye. This is exactly what I was looking to do here, and the luminosity mask made it very easy.
This is a good start. The cloud shows improved tonal separation and is more congruent with the textural qualities present in the rest of the scene. However, this adjustment also had the unintended effect of graying-down the cloud. The tonal separation has been significantly improved, but some of the brilliance has been lost.
This now is a good example of a situation where painting through a luminosity selection (Luminosity Painting) makes a huge difference. This technique lets me selectively restore the crisp whites to the cloud by painting white onto a “Dodge” layer exactly where I want to add brightness. The luminosity selection controls which pixels receive paint and how much they receive.
LUMINOSITY MASK PEARL: The “7½” and “2½” zone selections are my initial go-to masks for removing this type of midtone grayness from an image. Painting white through a Zone-7½ selection lightens the lighter grays, but, because the very lightest tones in the image are subtracted off, it prevents blowing out the whites. The Zone-2½ selection does the same for the midtone darker grays. Painting black through the Zone-2½ selection darkens these grays, but, because the darkest blacks have been subtracted off, maintains enough texture to keep the details from going black. A 30% opacity brush is a good starting point, and plan on using multiple strokes to slowly build up the desired effect.
Painting white through a Zone-7½ selection in this case nicely targets the cloud without leading to a loss of texture in the brightest whites. It also prevents spilling paint onto the darker tones in the mountains behind the clouds. Some care is taken to use a brush size that allows paint to mostly fall on the gray clouds that need to be brightened and to allow selectively painting some areas multiple times.
The “Dodge” layer for this luminosity painting is a blank pixel layer set to Overlay blend mode. This painted layer, placed against a gray background, is shown below. It demonstrates how the Zone-7½ selection very nicely confines the white paint to the cloud and how I was able to add more paint to some areas to increase the effect exactly where I wanted it.
The final image is shown below. The rollover is the image before luminosity painting.
In summary, this image, with its poorly separated cloud texture, is a classic situation for using luminosity masks. Three important luminosity principles were applied:
- Think tonally−Choose luminosity masks in situations where there is adequate tonal separation which can be exploited to create a useful mask or selection.
- Find the right mask and feathering−Look for a mask that is clearly lighter in the areas that need adjustment compared to areas that should not be adjusted.
- Paint for precision−Use luminosity painting to precisely burn and dodge the image to create the proper tonal balance.
These concepts can be applied to other situations where luminosity masks are being considered as a tool for image development.
If you’d like to experiment with these techniques, a downsized-size jpg of this image without the adjustments described in this tutorial can be downloaded here.