Local Light: Sketch-o-graphs

Fishhook barrel cactuses live up to their name. The spines look like fishhooks and are just as sharp. These, along with the hair-like structures, create an almost abstract covering for this common cactus. Looking closely at the edges, you can see the fine black lines that the sketch actions add.

An earlier post on local light discussed how photographing nearby subjects provides an opportunity to see them in new ways using different techniques, either during the capture phase with the camera or in post-processing. This past summer I decide to more fully explore this approach. My daily walk on the neighborhood streets offers plenty of interesting subjects in the form of Sonoran desert cactuses and other succulents. But I also wanted to do something different photographically with them and had my eye on the sketch technique I learned from Steve Dell. His photos using this technique along with those of Bob Hills made me want to explore this technique to see what works for me. These “sketch-o-graphs” can still look photographic, but they usually have enhanced edges and color. I was also planning to include the sketch action in the TK8 plugin and wanted to gain additional experience using it to insure the plugin included the best practices for getting good results. NOTE: The sketch actions in the TK8 plugin are accessed via the Workflow Extras button on the main interface of the Multi-Mask module (green outline, left image), and this opens a new interface where the Color Sketch and B&W Sketch actions are located.

Choosing the subject

Based on what I saw in Steve’s and Bob’s photos, images with lots of detail and texture seem to be the best candidates for sketches. Things like clouds, moving water, and out-of-focus elements are less affected by this technique. The sketch action finds edges in the image, and the stronger the edge contrast, the more prominent it will appear in the sketched version. My local subjects, the succulents along my street, were ideal for this as they had excellent texture and well-defined edges. However, I actually took it a step further and focus-blended most of the the images. Since many of the images were shot with a macro lens, focus-blending eliminated most out-of-focus areas and helped insure the entire image would convert well as a sketch.

The converted RAW file is on the left and the finished sketch is on the right. The sketch technique enhances both color and edges. Color sketches often soften fine textures in the image, which makes the resulting image look more sketch-like. The strong edges, though, become more prominent and make this an ideal image for sketching. Agaves create some of the most beautiful patterns in the Sonoran desert. There is a welcoming softness in the flower-like shape of their leaves, but the sharp tips are a reminder to not get too close.

RAW file conversion and initial sketch

The sketch actions can have quite an impact on the image, though it’s not easy to predict what the image will look like once it’s been converted to a sketch based on the initial RAW fie conversion. As such, I’m not necessarily looking to make a perfect image in Lightroom or Camera Raw. Instead, what I want from RAW conversion is an image that converts well to a sketch, and that usually means a little experimentation. In Camera Raw, I apply a linear profile, click Auto, and then make a few adjustments to balance brightness, color, and saturation. Images that have slightly under-saturated colors tend to work better as sketches since the sketch action usually bumps up saturation. The image is opened as a smart object in Photoshop, and a sketch action is immediately run on it. If it looks good, I’ll work with that sketch from then on. However, if I don’t like it, I’ll delete the group containing the sketch layers and tweak the RAW file by double-clicking on the smart object’s thumbnail to do additional processing in Camera Raw. I go back and forth this way, creating a new sketch after each RAW file adjustment, until I get a sketch I like. In this way, RAW file conversion and making the initial sketch are essentially one step. This initial sketch, created immediately after RAW file conversion, becomes the base image for additional development in Photoshop. Running sketch actions on different RAW conversions lets me zero in on the best starting point for developing the sketch in Photoshop.

These images demonstrate the enhanced color and edges that come from running the Color Sketch action. The image can look quite photographic still, but often much improved compared to the RAW conversion. These are cereus cactus flowers that bloom for just one night each year.

Color Sketch or B&W Sketch?

There are two sketch actions: Color Sketch and B&W Sketch. Which to use is often a matter of personal preference, and I’ll sometimes try both to see if I prefer one over the other. The actions make it easy to try both, and there’s no easy way to predict which will look best. One thing to keep in mind is that the Color Sketch sometimes pulls unexpected colors out of the deep-shadow areas or adds excessive color saturation to some areas, like blue skies, and if these don’t look right with the image, then the B&W Sketch might be the preferred alternative. However, even with the Color Sketch, overly colorful shadow areas revert back to being quite dark and less colorful after the initial sketch adjustments. Occasionally, over-saturated colors require a little extra effort to control later on, and it’s sometimes better to pull back their saturation in the RAW file before running the Color Sketch action (and is another reason to coordinate RAW file conversion with making the initial sketch).

A black and white sketch-o-graph of an agave clearly shows how the edges of the leaves have been enhanced with a black line, which looks quite natural in this monochrome rendition.

Initial settings

There are two settings needed for making the initial sketch. They’re both easy to set and it’s best to use the same setting on each image, at least to start. The layers produced by the sketch actions mean the initial settings can always be readjusted later.

  • Minimum–The Minimum dialog window opens as the action runs. The default is “4” and it’s best to leave it set to that as it generally produces output that indeed looks like a sketch, at least in the details and textured parts of the image. The Minimum setting creates the black sketch lines that define the sketch look, and the greater the edge contrast, the more prominent the sketch lines. After seeing the initial sketch it’s fine to go back and adjust the Minimum value, but using “4” as the starting point gives a good impression as to how well the image will work as a sketch.
  • Levels adjustment–The final step in the sketch action is the creation of a Levels adjustment layer on top of the other layers in the Sketch group. The Properties panel is opened (if it’s not already open) and the user can then make a Levels adjustment using this layer, and this adjustment is critical in creating the desired look for the sketch. The initial Levels adjustment is also (nearly) the same for every image: Pull the midtones slider to the far right edge of the histogram that’s displayed in the Properties panel. I’m not entirely sure why this works, but it does. This maneuver restores proper global contrast to the image while maintaining its sketch-like quality. Small deviations of where the midtones slider is placed can lead to slightly different (but noticeable) effects, but placing the midtones slider very near the right edge of the histogram is almost always a good starting point.

Adjusting the sketch

Once you’re satisfied with the initial sketch, there are still ways to fine-tune it. The image below shows the Layers panel after running the Sketch action and the main controls for adjusting the sketch after it’s been made.

  1. Minimum–Because a smart object is used to make the sketch, it’s easy to go back and adjust it. The main thing to experiment with is the Minimum setting. Simply double-click the word “Minimum” to open the Minimum dialog widow. Try settings in increments of 2 in order to see an effect. The default is “4” so try “2,” “6,” or even “8.” Higher numbers make the image darker and the black sketch lines more prominent. Lower numbers make the image lighter with less prominent sketch lines.
  2. Sketch group opacity–Lowering the opacity of the sketch group allows more of the original image to show through. This can make the result look more photographic and less sketch-like.
  3. Add a layer mask to the sketch group and paint black to conceal parts of the image–Some elements in an image, like sky and clouds, might not look as good after the sketch action runs while other parts with more texture and detail might look great. Adding a layer mask to the sketch group provides an easy way to conceal elements that don’t sketch well while revealing those that do.
  4. Change the sketch group blend mode–This is particularly useful when making a B&W Sketch. In this case, changing the sketch group layer’s blend mode to Luminosity changes the monochrome sketch into a color sketch, but with added detail compared to what the Color Sketch action provides. The video below by Dave Kelly has a nice demonstration of this.

Finishing the image

While the sketch action provides a significant transformation from the RAW conversion, it doesn’t finish the image. I see the sketch as providing a new base for further development. Burning, dodging, vignettes, adjustment through masks and even creative techniques, like glows, can still be used once a satisfactory sketch is in place. The sketch itself can look quite good sometimes, but that’s partly because it’s such a dramatic shift compared to how the image looked before running the sketch action. It’s fun to see what the sketch can do, but don’t hesitate to push further. Think about and experiment with additional possibilities. The sketch can certainly create a new creative path, but it will be up to you to follow it and see where it leads.

The image on the left is the focus-blended RAW file conversion. The finished image on the right shows the enhanced detail and color that were added with the Color Sketch action plus additional post-processing. This is a Santa Rita cactus that develops this purple/blue color as it dries out. Despite the extremely dry winter, it still managed to produce a large number of buds this spring.

Sketch-o-graph video tutorial

Dave Kelly does a nice job reviewing several of these steps in working with the sketch actions in the video below.

Additional examples

I’ve included some images from Steve Dell and Bob Hills below to show the variety of images possible from using this sketch technique.

The first gallery shows some Steve Dell’s sketch images. They again demonstrate how images with good detail work nicely with the sketch technique. In the image with the clouds, the clouds still pretty much look like un-sketched clouds. This is typical of image elements that lack distinct edges.

Bob Hill’s images below are from his “Sonoran Garden” series. Again, the texture and detail in these subjects create strong sketch images. Worth noting are how the distinct lines between shadow and sunlit areas on the leaves create even stronger lines in the sketched versions. In addition, the shadows are less dark and filled with more color.

I hope you’ll give the sketch actions a try. Sketch-o-graphs are easy to make and can produce some surprisingly unique and attractive images.

TK Quick Tip: Better vignettes via masks (Think H, S, and L.)

One way to get the most out of Photoshop masks that are based on pixel level data, like luminosity masks (where pixel luminance is the data source), is to always keep these masks in mind when solving problems or making adjustments in Photoshop. And the easy way to do that is to simply focus on the three basic characteristics that make up each pixel: Hue, Saturation, and Lightness. If the adjustment you have in mind or the problem you’re trying to solve can be isolated based on one of these three characteristics, then a mask from the TK8 Multi-Mask module might be of use. TK8 makes masks that target each of these characteristics as explained below.

Hue (or color)–If specific colors in the image need adjustment, an Infinity Color mask or Black & White Adjustment Layer mask can be used. Both make excellent masks that target specific colors in the image.

SaturationSaturation masks are based on pixel-level saturation values. The more saturated the color, the lighter the pixels are in the mask. Inverting the base Saturation mask provides access to Vibrance masks, which are brightest in the unsaturated colors in the image.

Lightness–To access pixel-level Lightness data, use Luminosity masks and Zone masks.

So, if you examine your image in terms of its Hue, Saturation, and Lightness, then it’s easier to pick the right type of mask for what you want to accomplish. As an example, I recently had an email discussion with a TK8 user on alternate ways to make image vignettes. TK8 Combo and Cx modules use an adjustment layer set to Multiply blend mode and a layer mask to provide darkening at the edges of the image. The amount of vignetting can be controlled by changing the layer’s opacity. The user’s concern was that this method sometimes made dark colors too saturated or too dark. Yes, this is a potential problem with any vignette, especially if the image’s edges already contain saturated or dark colors. My usual solution had been to paint black at varying opacities on the layer mask of the vignette adjustment layer to conceal the vignette in areas where it affects saturation or darkens the image too much.

In the Quick Tip video below, Sean Bagshaw demonstrates an even better method and it’s a good illustration of the how to think about adding pixel-based masks to the workflow. The user’s main concern was that dark colors were sometimes being affected too much by the vignette. So the question then becomes: Is there a way to select dark colors in the image using a mask? Well, yes, of course, there is. That would be a Darks-series luminosity mask. So the next question becomes how to use a luminosity mask to subtract the dark colors from a vignette so the dark colors are less affected by the vignette. The video shows how easy this is while still creating a really nice vignette.

One of the nice things about using these pixel-based masks is how perfectly the effect is revealed in the image with no visible edges or halos. In the example of the Freehand Vignette in the video, the vignetting still looked perfect even after Sean increased the opacity of the vignette layer. Seamless blending is one of the defining characteristics of pixel-based masks, so using them provides considerable latitude in the degree to which pixels selected by the mask can be adjusted before creating unwanted artifacts.

Summary: Think about Hue, Saturation, and Lightness when deciding if a pixel-based mask or selection would be useful in the post-processing workflow. If one of these characteristics helps isolate the areas that need adjustment, then the TK8 Multi-Mask module can be used to generate and refine a mask to better target the adjustment to where it’s needed most.

Extra Credit Challenge: Sean used the TK8 Mask Calculator to create an ideal vignetting mask. Can you think of a way to use TK8 to paint the same effect directly onto the layer mask WITHOUT using the Mask Calculator?

TK8: Color grading with masks

Color grading is the process of adding color to an image beyond the normal or natural color balance that existed when the photograph was taken. It has traditionally been used in films to impart a specific mood, like the color green that signaled the characters were inside the matrix in the “The Matrix” movie franchise. Photographs, however, can also benefit from color grading, and have actually been doing so for a long time. Things like warming up the colors for images taken around sunset and adding extra blue to images of ice and snow are common examples.

“Morpheus” from The Matrix movies. Notice how the green color grading lets viewers know he is inside the matrix. Image copyrighted by Warner Brothers.

Lightroom and Camera Raw added color-wheel-based color grading in 2020. Color grading an image could now be accomplished by simply dragging and dropping an icon on the color wheel to add that color to either the shadows, midtones, or highlights.

Camera Raw color grading interface.

Photoshop, however, still lacks a similar feature. Curves, Levels, or Color Balance adjustments layers can be used for color grading, but doing so requires using multiple points on the Curves adjustment layer or adjusting multiple sliders on Levels and Color Balance. There is no native, drag-and-drop color grading tool inside Photoshop.

The TK8 plugin now adds color-wheel-based color grading to Photoshop. Shadows, midtones, and highlights can all be adjusted by simply dragging and dropping a matching icon (black, gray, or white square) on the color wheel. The further the square is placed from the center of the color wheel, the more intense the color. In addition to changing colors, the brightness of the different tonal ranges can also be adjusted with the slider at the bottom of the interface.

TK8 color grading interface.

Using a color wheel greatly simplifies the color-grading process. Multiple colors and intensities can quickly be tested to find the right one. And while it’s certainly possible to color grade the entire image using the TK8 color grading interface, the real power of this feature lies in combining it with the different masks that the TK8 Multi-Mask module generates. Color grading has traditionally been associated with more global color adjustments to the image, but the TK8 masks allows it to be more specific and localized, while still providing the easy control that comes with the color-wheel interface. There’s even a specific color grading button in the module’s output section to a) automatically create the necessary color grading layer with the mask preview as the layer mask, and b) open the color grading interface to allow quick access to this functionality.

Red outline shows the color grading output button.

Color grading through masks also simplifies the entire color grading process. When color grading the entire image, it might be necessary to color grade all three tonal ranges—highlights, midtones, and shadows—to get the desired look. However, when color grading through a mask, it’s often NOT necessary to use all three tonal ranges. Simply color grading the midtones, for example, might be all that’s necessary to achieve the effect you’re looking for.

In the video below, Dave Kelly demonstrates how to combine color grading with different pixel-based masks generated with the TK8 Multi-Mask plugin. He first demonstrates color grading the entire image, but then starts narrowing it down, color grading the background separately. He then goes even further to work with luminosity masks, zone masks, and color masks as ways to constrain color grading to specific parts of the image. He even throws in using the mask calculator in the Multi-Mask module as a way to focus color grading right where it’s needed.

Color-wheel-based color grading provides a new option for controlling color inside Photoshop. Combining it with masks can make color adjustments even easier.

Be sure to subscribe to Dave Kelly’s YouTube channel for more TK8 videos.

Updates: Paint Contrast video and TK8-beta with Photoshop 22.5

Paint Contrast

Dave Kelly posted a good video last week demonstrating how to use the “Paint Contrast” action in the TK7 Combo and Cx modules. I’ve linked to it below.

My friend B. pioneered contrast painting. Dave’s video bumps it up a notch. I’ll explain. The Paint Contrast action allows adding contrast to an image simply by painting on a specially constructed pixel layer on the Layers panel. To make it easy, there are just three “colors” to paint with:

  • 50% Gray–Which darkens shadow values and lightens highlight values, i.e. increases contrast in both shadows and highlights.
  • Black–Which primarily darkens shadow values (increased contrast in the shadows) and affects highlight values less.
  • White–Which mostly lightens highlight values (increased contrast in the highlights) and affects shadow values less.

The neat thing about this technique is that you don’t need a luminosity selection to paint through in order to target specific tones. Painting with black paint is almost like painting through a Darks-1 or Darks-2 selection and painting with white is like painting through a Lights-1 or Lights-2 selection. Black paint automatically selects dark tones to darken and white paint automatically selects light tones to lighten. And, just like with luminosity masks, the painting blends seamless, especially when painting with a feathered brush. So, while painting with black affects shadow values most, it also feathers the effect perfectly to blend the change into the image’s midtones and highlights. Likewise, painting with white affects highlights most, but also tapers perfectly into midtones and shadows. And, of course, painting with 50% gray feathers the effect into both the shadows and the highlights.

Dave demonstrates painting with all three colors, but the new thing I learned from the video is to use a different layer for each color. I had been using this technique on a single layer, mostly painting with 50% gray and occasionally lighter or darker shades of gray. However, by doing a separate “black paint” layer and “white paint” layer, as shown in the video, you can affect shadows and highlights independently and better tailor the effect to the image. This offers more control in both tonal ranges even when working on the same area of the image that has both shadow and highlight values. This control also extends to subsequently fine-tuning the adjustments, like changing the layer’s Fill opacity. The painting contrast technique is super easy and often yields pleasing results. I hope you’ll give it a try.

TK8-beta with Photoshop 22.5.0

Adobe released Photoshop 22.5.0 almost two weeks ago, and I’m happy that TK8-beta still works with it. It’s always a bit nerve-wracking whenever there is a new Photoshop drop. Something is always broken that wasn’t broken previously, and it’s almost impossible to predict if the TK panels will be affected. This isn’t entirely unexpected. The new UXP architecture on which TK8-beta is built is still evolving with new features being added and talk of some elements eventually being deprecated. I really like what UXP can do, so overall I think it will be positive direction long-term. I’m committed to trying to keep up with Adobe and fixing things as soon as possible if they do break. If you do have problems with TK8-beta, be sure to contact me so I can investigate and get things working. Also, it’s a good idea to save your download link. It can always be used to download the latest version from the download server so that when bugs do get fixed, you can get updates installed soon after I post information on this blog.

There are two things worth noting after the recent update to Photoshop 22.5.0. The first is that some of the formatting for the TK8-beta user interface shifted a little. The spacing around some elements in the sub-menus might look a bit off. This doesn’t affect the functioning of the plugin, but if it bothers you, use your download link to get the newer version, which has been added to the download server, or contact me if you need your link reactivated. These formatting issues will also be corrected in the final version of TK8 to be released later this year.

Another problem related to Photoshop 22.5.0 only happens on Windows computers and concerns input boxes, those boxes where the user can type in a value, like the Opacity value on the Layers panel. If you click in an input box to create a cursor there and then click the “Backspace” key, it doesn’t delete characters in the input box like you’d expect. Instead, it deletes layer masks or even layers on the Layers panel. Definitely not a good thing. You could accidentally delete a layer mask or even a layer without realizing it. This is new to Photoshop 22.5.0 and can affect the TK8-beta modules as well. (Interestingly, TK7 modules appear to NOT be affected.) For example, if you click in an input box on one of the TK8-beta modules, like to change the Height or Width for web-sharpening, and then tap the backspace key, the active layer(s) get deleted. Adobe has acknowledged the problem in this post, and it’s bad enough that they’ll probably try to get it corrected in the next dot update. Again, this is only a Windows issue. The “delete” key or “fn + delete” on Mac doesn’t have this problem. The short-term workaround is to use the cursor to highlight the letters and numbers that need to be changed and then immediately type their replacements without clicking the backspace key. This works, but if you’re like me, the backspace key is used almost as often as the space bar. So it’s hard to remember not to use it when working inside an input box. I know Adobe is dedicated to fixing bugs and am sure this one will get fixed in the near future.