TK Quick Tip: Layer Mask mode

One of the best features in the TK7 panel is the ability to view luminosity masks and other pixel-based masks as fast as they’re created. Seeing the actual mask up front allows you to make an initial assessment of what will be revealed and what will be concealed before the mask is actually put into use. It also allows you to modify the mask to make sure it selects the parts of the image you want selected.

When you actually deploy the mask, though, sometimes it’s not quite doing what it was expected to do. Maybe a different mask would have worked better. Or maybe the current mask is pretty good, but still needs additional modification. It’s occasionally hard to know how a particular mask is going to perform until you actually see how it affects the image.

If the mask you created using the TK7 panel was applied as a layer mask, then there’s no need to start all over. The TK7 panel has “Layer Mask mode” that lets you modify layers masks or even change to a totally different mask without going through the process of generating and applying a new mask.

In the video below, Sean Bagshaw covers three situations where Layer Mask mode comes in handy.

  • Changing to a different mask entirely.
  • Modifying the current layer mask.
  • Exposure-blending to control dynamic range.

The key feature in all these examples is that the image itself drives the decision-making process. In Layer Mask mode, you no longer see the mask since it automatically gets applied as a layer mask to the active layer. What you see instead is the effect the mask has on the image. So in layer Mask Mode you’re choosing the mask based on how the image looks and not on how the mask looks. You can still look at the mask if you want to, but you’ll also be able to instantly see how the mask affects the image. As always, Sean does an excellent job walking you through the process. I hope you’ll give it a try.

Add YOUR Photoshop actions to the TK7 panel (new method)

In addition to generating luminosity masks, the TK7 panel also creates a more efficient Photoshop workflow with the buttons and menus in the Combo and Cx modules. Many photographers keep one of these modules open in their workspace because it provides quick access to many commonly used Photoshop features. Another way to improve workflow efficiency is to use Photoshop actions to perform repetitive steps. However, using Photoshop actions still involves opening Photoshop’s regular Actions panel. The Combo or Cx modules provide an alternative to this also. Both modules have a dedicated User Actions menu for running Photoshop actions. You just need to take a few minutes to add the actions you are already using. Once it’s set up, you can access your Photoshop actions directly from the User Actions menu in these modules, and this sub-menu automatically closes once the action completes. It’s a fast and efficient way to run Photoshop actions. To help you get started, the video below shows a new method for adding your previously recorded actions to the User Actions menu of the Combo or Cx module.

TK Quick Tip: Frequency separation

Luminosity masks are only one feature of the TK7 panel. There are also lots of buttons and actions to speed the creative workflow. Sean Bagshaw recently posted a TK Quick Tip video on one of the actions: Frequency Separation. While this technique has its origins in fashion and portrait photography, Sean shows how it can also be applied to nature and other photographic situations.

Frequency separation literally separates the image into two layers. One layer contains the color information (low-frequency) and the other the texture information (high-frequency). Once separated, these two component can be dealt with independently. Image cleanup is the main application for this technique and is especially useful when standard methods−like the clone stamp, patch, and healing brush tools−aren’t doing a great job.

Sun flare, human or animal tracks, and certain out-of-place elements in the image are situations where frequency separation is especially useful. The Frequency Separation action is found in the “TK►” button menu on the Combo and Cx modules of the TK7 panel.

blue-example

There are two important points to keep in mind when using the action.

1) The action stops to let the user enter a blur radius. A good choice for touching up color is the default 10-pixels value. For fixing texture issues, increase the blur radius so that it blurs out the unwanted texture or feature.

blue-example

2) Once the new layers are generated, the Clone Stamp tool is commonly used in the cleanup/repair process. It’s very important to set the tool’s “Sample” option to “Current Layer” in order to confine the needed cloning to just one of the frequency separation layers.

blue-example

Sean shows you how to do all this in the video below. I’m sure you’ll find it useful.

Be sure to subscribe to Sean’s YouTube channel for more great tips on photography and post-processing including those listed below.
Frequency separation
Mask-the-Rapid-Mask modification
“My Channels” masks
Infinity color masks
Linked vs. unlinked smart objects
Three ways to use Levels and Curves
Reusing saved luminosity masks
Developing a quality night sky
Split toning
Cloud sculpting
Exposure blending

Infinity Color Masks: A closer look

First off, a big THANK YOU for the positive feedback on the new infinity color masks. They seem to have resonated with a large number of photographers. Also a big thanks for the questions that were sent. I can see there is some interest in better understanding these masks and how they work. In this blog, I’ll look at three sample images using well defined colors to explain what’s going on when you make an infinity color mask. NOTE: This gets a bit technical, so I put the main points in bold-face type if you want to just skim through.

Color is defined by three values: Hue, Saturation, and Brightness, which are the three values in the HSB color model, and all three can be extracted using Photoshop’s Color Picker. All three are also part of an infinity color mask. The initial Color Picker selection extracts the Hue from the sampled area and this Hue is the foundation around which the mask is built. The gray values in the mask are then determined using the Saturation and the Brightness of pixels that match the chosen Hue. That’s the basic idea of how an infinity color mask is made.

Of course, the actual infinity color mask provides additional options. For example, it’s not limited to a single Hue value but instead creates a mask that encompasses a range of similar hues around the sampled Hue, and there is also a hue-based transition zone from the defined Hue range into nearby hues. Additionally, I’ve cranked up the mask’s lightness since most “normal” photographs produce relatively dark infinity color masks that would be hard to evaluate on-screen. But the exaggerated lightness adjustment does not affect the relative lightness within the Hue range being selected by the mask, except for unusually saturated colors, in which case the mask’s overall lightness can easily be turned down to properly restore the relative pixel lightness. In all cases, though, the infinity color mask is constantly and accurately adjusting with each image to take Hue, Saturation, and Brightness into account for what was sampled from the image.

In the first example below there are three different blue colors−a light (but unsaturated) blue on the left, a darker (and more saturated) blue in the middle, and a lighter blue on the right (with the same Saturation as the middle square). These colors differ only in Saturation and Brightness. They are all the same Hue: Hue = 206 degrees in the Adobe RGB color space. That means that no matter which color square is sampled to make an infinity color mask, the initial mask will be the same. Hue is the foundation for infinity color masks, and since the Hue is identical in all three color samples here, they all produce exactly the same mask when sampled, and this mask is shown directly below the image.

blue-example

The gray values in the mask are determined by the Saturation and Brightness of the colors that share the sampled Hue. The lightness of the grays in the mask varies depending on the Saturation and Brightness of the underlying color.

I’ve listed the pixel-based Hue, Saturation, and Brightness values for each color. Without going into the math, a color needs to have both Saturation and Brightness values greater than zero to show up as a gray tone in the mask. When two colors share the same Hue and Saturation value (second and third squares in the image), the different Brightness values determines the level of gray in the mask. When two colors share the same Hue and Brightness values (first and third squares in the image), Saturation determines the level of gray in the mask.

The lesson from this example is this: Hue is the foundation for all infinity color masks. These masks can’t exclude colors that match the chosen hue and sampling different colors that have the same Hue (but differ in Saturation and Brightness) results in identical masks. I specifically chose blue for this example since some photographers have been asking about using infinity color masks to select one shade of blue from a sky that is all blue. That’s probably not going to work too well. Remember that all three blue colors in this example have the same Hue and the Hue value is the foundation around which the infinity color mask is constructed. So you’ll get the same infinity color mask regardless of which color is sampled in this image to create the mask. In addition, because these colors share the same Hue value, you won’t be able to use the infinity color mask control window on the RapidMask module (shown below for this example) to alter the mask in a meaningful way. Neither the color range nor feathering sliders will make any difference in this example since all the colors are the same Hue. Bottom line here is that an infinity color mask is great at selecting a blue sky and displaying all the Brightness and Saturation nuances in that sky, but not good for selecting a specific shade of blue from that sky. NOTE: For slightly different hues you can try narrowing the hue range and feathering using the controls on the module (see next example), but for substantially similar Hue values, like in many skies, it might not be possible to successfully select a specific shade of blue exclusively.

blue-example

This next example was sent to me by a photographer who was trying to create a mask that selected a somewhat unsaturated orange element in his image that was sandwiched between highly saturated red and yellow elements. The image below isn’t the actual image but represents the situation. Below the image is the mask that is created when sampling the middle tan color, which is actually a desaturated orange color with a Hue value of 30 degrees.

blue-example

The mask generated shows a gray value for the middle square in line with the Saturation and Brightness of the sampled tan color with a Hue of 30 degrees. However, the transition zone colors for orange, red (Hue = 0 degrees) and yellow (Hue = 60 degrees), are much more saturated and brighter than the sampled color. Since red and yellow are in the initial transition zone for this tan color when it’s sampled and, because they have such higher Saturation and Brightness than the sampled color itself, they end up brighter than the sampled color in the mask.

Unlike the blues in the previous example, this is a situation that infinity color masks can handle. Because there are different Hue values for the three colors, the color range and/or the hue-based feathering can be contracted using the infinity color mask control window to exclude the adjacent hues, even though they are much more highly saturated. In this case I collapsed the color range to its smallest width and nudged the feathering slider just a bit to the left as shown in the image below.

blue-example

After I did this, the infinity color mask preview looks like the image below.

blue-example

The gray value of the unsaturated orange sample has not changed, but the red and yellow transition colors are now properly excluded from the mask. And even though the sampled color isn’t white in the mask, it’s still light enough to be useful for revealing an adjustment in this image. NOTE: The gray in the mask could be made even lighter using the MODIFY section of the TK7 RapidMask module.

There are two lessons from this example:

1) To make a good infinity color mask, sample a good color. And by “good color,” I mean one with a decent amount of saturation. The tan color in this example is a desaturated orange. It will never be pure white in an infinity color mask, but can still make a useful mask. As colors increase in saturation they get lighter in the mask (like the red and yellow here), and it’s generally easy to look at an image and identify which colors are saturated and which are unsaturated. A good rule to follow if you plan to use an infinity color mask is to sample colorful pixels. Black, white, and gray are “colors” with 0% Saturation. They always render as pure black in an infinity color mask, and colors approaching black, white, and gray will be very dark gray in the mask. So if you want to make good infinity color mask with plenty of light gray values indicating selected pixels, be sure to sample colorful colors.

2) The color you sample in your image will NOT necessarily be white in an infinity color mask. It might be white if it has enough underlying Saturation and Brightness, but the sampled color might also be a shade of gray in the mask, like it is here, when Saturation and Brightness are lower. And that’s OK in many cases. The mask is still properly displaying the pixel-based values for the selected color in your image, and the gray values in masks insure seamless blending when the mask controls how an adjustment gets applied to the image.

The final example image below has two unsaturated colors, dark gray and light gray, and a saturated red. The mask underneath the image shows what I got when I clicked on either of the gray squares to sample for an infinity color mask. Can you explain this?

blue-example

The two gray samples, lacking any saturation, should be black in the infinity color mask, and they are. But why is the red color square white in the mask if I sampled a gray color to make the mask?

The answer lies in the Color Picker when I sampled the gray color from my image. It’s shown below with the arrow pointing to the spot on the Color Picker chosen when I sampled the light gray square.

blue-example

Even though the sampled color has 0% Saturation (it’s a light gray and S = 0% in the Color Picker), Photoshop still requires every sampled color to have Hue, Saturation, and Brightness values in the Color Picker window. When the Color Picker opened, the Hue was set at its default value of 0 degrees. IMPORTANT NOTE: A Hue value of zero does NOT mean there is no hue. It means that the Hue is set to a value of “0 degrees,” which corresponds to the color red. Remember, infinity color masks always use the Hue of the sampled color as the foundation for building the mask, and, in this case, that Hue happened to be red (0 degrees). So for my image of two gray squares and a red one, the red square matches the sampled Hue, and because it is also bright and saturated, it ended up white in the infinity color mask of this image.

There are a couple of lessons here also:

1) The first is one I’ve already mentioned. Colors with zero Saturation (white, gray, or black) are always black in an infinity color mask. You cannot select colorless colors and expect them to be white or even gray with this type of color mask. Both Saturation and Brightness need to be greater than 0% to show any degree of selection in an infinity color mask.

2) The second one is that if a neutral color (white, gray, or black) is sampled to make an infinity color mask, the Hue that shows in the Color Picker will be the foundation around which the mask is built. If there are any colors in the image that are within range of that Hue, they will be selected in the resulting mask. This often comes as a surprise the first time occurs, but hopefully it now makes sense why this might happen.

Conclusion: Infinity color masks are a great way to generate masks using pixel-based Hue as the foundation. Colors with similar Hue value will all be selected by the mask. The other two components of color, Brightness and Saturation, determine the final mask. Bright and saturated colors will be lighter shades of gray or even white in the mask. Dark and unsaturated colors will be dark gray or black in the mask. Remember, infinity “color” masks are selecting COLOR. Objects without much color (unsaturated colors) will always be dark in an infinity color mask. So bright, saturated colors are what can be best selected with these masks. Or, to put it another way, a “color” mask is a poor choice for trying to select parts of an image without color. If you have a dark element or dark area in your image, or an unsaturated color, and need to select it, use a luminosity mask or a saturation mask. These types of masks are better at selecting pixels based on Brightness and Saturation.

OK, I know that’s a lot to absorb, but there’s quite a bit happening in the background when the TK7 panel makes an infinity color mask. Please feel free to leave a comment or contact me if you have any questions.

TK Quick Tip: Mask-the-Rapid-Mask

In the video below, Sean Bagshaw reviews the new blur features in the Mask-the-Rapid-Mask option in the MODIFY section of the TK7 RapidMask module. This addition is a request I received from a users who wanted more control over this process. Mask-the-Rapid-Mask lets you localize the effect a luminosity mask has on an image while the mask is being created.

The Lasso tool is a common starting point for choosing the specific parts of the image where you want the luminosity mask adjustment applied. In order to avoid a hard edge to the selection (that might be visible in the image), the Mask-the-Rapid-Mask action now opens the Feather Selection dialog. The panel calculates a generous feather radius based on the size of the image, but users are given the chance to adjust this. Unfortunately, Photoshop doesn’t preview feathering, so you might want to experiment when you start using this feature to get a sense as to whether you prefer more or less feathering than is suggested. Clicking “OK” in the Feather Selection window completes the action and applies the selection as a mask to the Rapid Mask.

At other times, users already have a dedicated selection they want to use and would prefer no feathering at all. The action also accommodates this. Simply click the “Cancel” button in the Feather Selection dialog and the selection mask is applied with no additional feathering.

Sean covers both possibilities in his video.

Be sure to subscribe to Sean’s YouTube channel for more great tips on photography and post-processing including those listed below.
Mask-the-Rapid-Mask modification
“My Channels” masks
Infinity color masks
Linked vs. unlinked smart objects
Three ways to use Levels and Curves
Reusing saved luminosity masks
Developing a quality night sky
Split toning
Cloud sculpting
Exposure blending

Make-It-Glow: A favorite TK7 action

The TK7 panel isn’t just about luminosity masks. There are also several actions to enhance images during processing. One of my favorites is Make-It-Glow.

The Make-It-Glow action is like the Orton effect in some ways except that it produces no noticeable blur. The color is blurred a little to help create the glow effect, but the detail and texture in the image remains unchanged. In addition to the more diffuse color, saturation is also enhanced, like with the Orton effect, so it’s important to use Make-It-Glow on images that can handle a little extra color saturation. I often use it near the start of the Photoshop workflow. This is often a time when the image can use a little boost in color saturation. The effect generally looks quite good, so it can help set the tone for further development. Below is an image before Make-It-Glow was applied. Roll the mouse back and forth over the edge to see the difference running the Make-It-Glow action can have. NOTE: The image is not visible on the email feed, so please visit the blog to see this. Also, the rollover image might take a few seconds to load.

Make-It-Glow actions

It’s not an overly dramatic change, but it works extremely well on this image to saturate the existing colors in way that looks natural. Photoshop’s Vibrance adjustment layer can have a similar affect on color saturation, but it tends to add saturation to cool colors more than warm colors, and this is the opposite of where glow usually looks best. Make-It-Glow has no color preference; it just works with whatever colors are present in the image. So compared to Vibrance, warm colors get a better saturation boost with Make-It-Glow. In addition, Make-It-Glow blurs the existing color a bit, and this provides a better sense of glow that a Vibrance adjustment layer can’t replicate.

Still, it’s important use Make-It-Glow in the right situations. Here are some that I’ve found useful:

  • Low-saturation images.
  • Intimate landscapes−those with land but no sky.
  • Autumn foliage.
  • Rocks, like sandstone and slot canyons.
  • Grand landscapes with lots of clouds, like stormy or overcast skies. NOTE: Clear skies sometimes get over-saturated with this technique.

The Make-It-Glow action is found in the “Color” section of the menu opened with the “TK►” button on the TK7 Combo and Cx modules. It has only one user input, and that’s a Gaussian Blur pixel radius. This blur radius determines the color diffusion for the effect. A radius that equals the megapixel count of the image is a good starting point, and this is what is calculated and suggested by the panel.

Gaussian Blur window

In addition to the blur radius, there are ways to customize the effect after the actions completes.

  • Run the action a second time to enhance the effect.
  • Adjust the layer opacity if it feels too strong. This is potentially useful if running multiple Make-It-Glow actions on the same image (which sometimes works quite nicely).
  • Add a layer mask to the Make-It-Glow layer to fine-tune the effect. The video below shows several methods to do this.

Of the various techniques that help improve the color in my images, Make-It-Glow is at the top of the list. It plays a role−sometimes a very large role−in determining where the image goes color-wise. It’s also a bit of a safety valve when I’m creatively stuck. Once I see what Make-It-Glow does, I can often find a way forward with the image.

Watch the video below for additional tips on using the Make-It-Glow action. View it full-screen in order to see the subtle changes from the action and the masks.

Stunning shadow and highlight details with the TK7 Triple Play

You’ll never outgrow luminosity masks or the TK7 panel. There will be lots of new opportunities once you become comfortable with pixel-based masks. Even I don’t know it all. Occasionally someone writes me with a method for using the masks and the actions in the panel that would never have occurred to me. Something that was not intended at all when I coded the panel, but once I try it I’m like: Wow, how did I miss that!?

Actions menu

The technique that follows (and also explained in the video at the bottom) was one of these and is provided by my friend Daniel Anderson. He was an early adopter of luminosity masks. When I released the Triple Play actions sometime around 2008 or 2009, he wrote back almost immediately excited about what they did for his images and said that he used them on almost every single one. The Triple Play eventually found its way into the panel and I recently connected with Dan and ask him to explain in more detail how he uses this particular action. I decided to try what he suggested am now using it regularly using it too. When properly applied, it indeed makes every image look better. Tonal separation is enhanced in both the darkest darks and lightest lights and details in these tones stand out much better also. The effect allows images, even jpegs viewed on the web, to show wonderful texture in areas where detail is frequently lacking.

The Triple Play is accessed by clicking the “TK►” button in the TK7 Combo and Cx modules to open the Actions menu. The “Lights TP” affects the light tones and the “Darks TP” affects the dark tones. I generally use the Darks TP first since my images tend to have a lot of darker midtones in the final image. And sometimes I only use the “Darks TP” since the amount of lighter tones in my image may be extremely small, so the effect would be minimal. However, if there are obvious light tones in the image, you’ll definitely want to give “Lights TP” action a try.

The Triple Play uses luminosity masks (some of them blurred) and blend modes to achieve its effect. The first step is to choose a pixel radius for a Gaussian blur. A dialog window opens to ask for this.

pixel radius dialog

I use 20 pixels for the Darks TP and 15 pixels for the Lights TP. Dan uses 20 pixels for both. You can experiment, but these numbers would be a good starting point.

Darks Triple Play

The Triple Play action then creates a set of Curves adjustment layers set to Screen or Multiply blend mode that are masked with the Dark-series luminosity masks (some of which are blurred using the user-entered pixel radius). The adjustment layers aren’t meant for adjusting, though. The desired Triple Play effect is achieved simply by turning ON the visibility of different layers and letting the blend modes either lighten or darken the image. Initially, the visibility of all layers is turned off, so there is no effect on the image until the visibility of different layers is turned on.

For the Darks Triple Play, I turn visibility ON for:

  • the “(S) Darks-4, blurred” layer, which lightens the image because it is in Screen blend mode, and
  • the “(M) Darks-4, no blur” layer, which darkens the image because it is in Multiply blend mode.

This combination is known as an S-4/M-4 adjustment. Dan says he starts off with S-3/M-3, but that effect usually feels a bit strong for me in the image’s dark tones. NOTE: The lower the number, the stronger the effect. S-3/M-3 has a greater effect than S-4/M-4 because the Darks-3 masks reveal more pixels than the Darks-4 masks. Many times the S-4/M-4 combination of layer visibility works perfectly and I call it quits. Other times, I might try different layer combinations or adjust the group’s opacity. But I don’t do any actual Curves adjustments. I only turn different layers on and off to find what works. What I’m looking for is better detail separation and better contrast in the darkest tones in the image without too much clipping (though a little clipping might occur and generally looks OK for dark tones).

Lights Triple Play

Whether or not I run a Lights TP depends on the image. For many of my images, the Darks TP is all I do, but if I have some strong white tones in elements that would look better with enhanced detail, I try a Lights TP too. Running a Lights TP is just like running a Darks TP. Choose a pixel radius (15-pixels is a good starting point), let the actions create all the layers, and then turn on the visibility of the ones that produce the best effect, i.e. increased tonal contrast in the very lightest tones without blowing out the brightest values. I usually start with S-3/M-3 since fewer light tones in my images usually means I need more revealing masks to see an effect. However, I’m also more likely to pull back the effect since clipped highlights tend to bother me more than clipped shadows. I’ve found a midtones mask on the “LIGHTS Triple Play” group a good way to insure my whites remain within range. Watch the video below for details.

Here are some important things to know and keep in mind about the Triple Play.

  • It should be LAST step in the workflow. The Triple Play is most strongly affecting pixels at the extreme ends of the tonal spectrum. When you stop and think about it, these are areas that are somewhat harder to adjust. Much of our workflow is focused on the midtones. Midtone adjustments can have a significant and positive effect on the image, so this is where we normally concentrate our efforts. The Triple Play is a chance to get into the darkest darks and lightest lights and make them look equally good. However, once you run these actions, it’s time to quit with the adjustments. Additional processing could run the risk of more noticeable clipping unless you’re really careful.
  • But there is a little flexibility. Dan applies his Triple Play actions after sharpening for print. I apply mine before. By including the Triple Play as the last step on my master file I get the benefit of seeing its effect in both my down-sized jpeg images for web presentation and in the final upsized sharpened prints. I find it scales pretty well, although I may pull back a bit on the amount of web-sharpening if the images look a little too sharp. However, the increased punch in both the light and dark tones is usually apparent when you compare jpegs with and without the Triple Play.
  • Zoom in to see what’s happening. Depending on how many light and dark tones are in the image, it may be difficult to see the effect the Triple Play has on a monitor-sized image. I generally zoom into at least 50% and maybe 66.7%. It’s much easier to evaluate what’s happening and decide which layers to turn on when you can see the details in the image better. I work to get the effect right at this magnification rather than relying on getting it right in a monitor-sized image. Also, be sure you’re viewing a dark part of the image if adjusting a Darks TP and a light part of the image when adjusting a Lights TP.

lightening bolt button

  • Choose your subjects. Some elements, like soft clouds, won’t necessarily look better with the Triple Play. Too much detail in the highlights of clouds can look unnatural. Harder subjects, like the European cathedrals in Dan’s images, would be the ideal subject to experiment with the Lights Triple Play, I think.
  • There is a lot going on in the background when the Triple Play actions run. For a large image or a slow computer, it might take a few seconds for the TP actions to complete. There are 11 layers, 10 masks (all 16-bit), and one group needed for each Triple Play action.
  • You don’t need to keep all the Triple Player layers. Once you find the layers you need to be visible to create the proper effect, click the lightning bolt button on the Combo/Cx module. It quickly deletes all the hidden (not visible) layers and makes your Layers panel feel normal again. You can also close the TP layer groups to make the final result more compact.
  • If you want to read about the Triple Play in detail, open the Settings window on either the Combo or Cx module and click the “TK” button. This will take you to a website where you can get a free download that includes a somewhat lengthy “Luminosity Mask Triple Play and Advanced Masking” PDF on how the layers and blurs are created.

Like Dan, I’m using the Triple Play as the last step on every image now. Even when I think I’ve done a good job processing, the Triple Play (especially the Darks Triple Play) makes the image noticeably better, for both the print and the jpeg, every time. There’s better separation in the dark areas of the image, and when you look at the print there is a surprising amount of detail in what would normally be very dark parts of image that lack it. I will definitely keep trying it on all my images in the future.

The video below is a quick review of applying Triple Play to some images. It’s best to view it full-screen since some of the effects are subtle. I really do like this technique and hope you’ll give it a try.