In this tutorial, I will break down my thought process of designing a portrait image and being able to apply different lighting scenarios to the same image.

Set Up

 

This is my setup for clip studio paint. Main points of interest for this tutorial are the separate window, the color wheel, the color slider, and the sub view.

 

The separate window is like the navigator window. Minus the red box highlighting the active painting area. I have it open to constantly check the initial and overall read of the image while I zoom in and work on the image.

I prefer this to the navigator window because the red box is an additional color that can affect how you perceive the image in the navigator view. Also you can flip the image in the separated window in case you don’t feel like flipping the window while you’re working.

To bring up a separate window of the image you are working on to go Window> Canvas > New Window

 

The color wheel is set to HSV color space rather than HLS. You can toggle this through the menu or the icon in the bottom right.

 

HSV = Hue, Saturation, Value

HLS = Hue, Lightness, Saturation

 

The main difference is Value and Lightness. Lightness adds white to the color while Value is the black and white value of a color.

If the color is changed to red you can see the main differences in the numbers on the side as well as the gradient difference in the slider. At a “pure” red the Lightness % is at 50% while the Value % is at 100%.

 

I prefer using the HSV color space because it’s easier for me to keep track of the values and usually I don’t want to add white to a color when picking with the color slider because it will affect the values of my image. This will be explained further in the Rendering section.

 

Sub view is for loading in reference images. You can drag and drop images into that window to use as reference in there, switch between them, and color pick from them.

Composition

I use a rule of thirds template and a perspective grid to use as guides to arrange my composition. The perspective grid is for where I want the eye level to be and the rule of thirds is for placement of important parts I want the viewer to focus on. I find this more useful when working with more complicated compositions as a starting point to arrange the elements in my composition.

I also don’t make compositions with the rule of thirds in mind all the time. I usually have an idea or a gesture that I like and use composition templates like this to frame and refine the arrangement of my ideas in a more pleasing way. And not everything that you want focused on has to be on a third or a guide, it can be a little on the nose, so being approximate is also okay.

 

But since this is a simple portrait, I’m just looking for how the figure breaks up the canvas space and the shapes and negative shapes that are made from the silhouette.

 

If there is a story or personality I want to communicate I do that through how the character holds themselves or how they stand and how the audience is looking at them.

• If the camera is at a low angle, the character will seem taller than the viewer.

• Around eye level, the viewer is face to face with the character.

• If the camera is at a high angle, the character will seem shorter than the viewer.

• Closer to the camera, the viewer is closer with the character.

• Farther away from the camera, the viewer is more distant to the character.

 

Of course these are just describing the physical relation of the viewer to the character but it can also be used as a metaphor for the relation of the viewer to the character.

 

• If the camera is at a low angle, the character is looking down at the viewer, the character could be thinking that the viewer is beneath them.

• Around eye level, the viewer is face to face with the character. The viewer and the character are on equal standing in this moment.

• If the camera is at a high angle, the character will seem shorter than the viewer. The character looks up to the viewer. The character might be admiring the viewer in this moment.

• Closer to the camera, the viewer is closer with the character. The viewer and character might be intimate or are in a close relationship.

• Farther away from the camera, the viewer is more distant to the character. The viewer and the character aren’t close in relations at all.

 

The meaning trying to be communicated in these images can also be further changed or refined by changing the expression of the character as well.

 

 

When planning out my lighting for my images I usually keep to grayscale. I’m mostly interested in the shadow shapes and separating my foreground, midground, and background.

Again because this is a portrait this reduces the image to two main planes, background and foreground/midground, so as long as the character and the background don’t blend in together is the important thing.

 

If you were to see the image as a small thumbnail you want to be able to clearly read the main shapes of the image.

 

Drawing

Before doing a clean line art pass I make sure to correct the structure of my rough drawing.

I use a shorthand to recreate the skull. I do this because facial features do not float on the face; they have anchor points. Eyes are set into the skull and noses are attached to the nasal bone. I want the structure to be correct for when I start rendering.

I do a clean line art pass based on these corrections.

Base Painting

I select the negative space of the silhouette based on my lineart using the magic wand tool then invert my selection (CTRL + SHIFT + I) and fill in the silhouette with a 50% gray.

 

Base colors are applied on a second layer using a clipping mask and set to [Overlay]. The colors are filled in with the Lasso Fill Tool. This is essentially the same as if I were to just set the BaseColor layer to [Normal]

 

Remember the rough lighting sketches from earlier? By sampling the base gray that is used from the sketches and applying it to the MainShape layer the base colors are adjusted to match the values of the sketch. This is a non destructive way of manipulating the values of your colors and reincorporating previous work that has already been done.

Grays darker than 50% will darken the overall value of your colors whereas anything above 50% will add white to your colors and wash them out.

Since there are other ways to readjust the values and I want to get the most detail I can get in at the largest value range to start off I will be keeping the MainShape layer at a 50% gray for the base painting. This will serve as the midtones.

Next I add another layer and set it to [Brightness] and 50% opacity. I’m trying to slowly build my values down based on Ambient Occlusion.

Ambient Occlusion is for shading indirect objects so if there is no direct light in the scene this is how you create shadows to create a visual separation. The shadows get darker where surfaces are close or come into contact with each other. Since the first lighting that is going to be done is overcast lighting, there is a lot of ambient light and no direct light source. The shadows should be soft so I used an airbrush and resized it to get sharper edges. I’m not that precise about it since I will be painting over this later.

What the [Brightness] layer setting does is directly adjust the value of an image. If I set the opacity to 100% since I am using black it will move those values to black or 0% value. So if I change the color I’m working with to a 20% gray instead all the values will move towards 20%. Change it to 80% all the colors will move towards 80%.

While what I’m doing right now will essentially produce the same results, I prefer using this to [Multiply] because what [Multiply] does is add value and color to anything below it. Think of it as mixing traditional paint, the more colors you add the darker the color gets. Where as with [Brightness] anything under the [Brightness] layer will move towards that value.

 

I add an additional layer [Multiply] set to 50% to push the darker areas even further, because it gives the same results either way.

 

On top of this I start refining what I have by painting over it, still keeping to grayscale. I basically just painted enough so there was some definition and form in the hair to work with and started to paint over the line work.

This is an okay point to start refining even further so the BaseColor layer is turned back on. All the layers are contained within a folder.

 

 

Rendering

From here I will mostly just be painting directly. Create a duplicate of the folder and merge it. Using the magic wand tool select and create a copy of each different material. I usually separate it out to skin, hair, and clothes.

Skin is complex because the color will vary from various factors. There is a translucency to skin and this can be proven by putting a light directly under your hand, being able to see your veins under your skin, or when severe light hits the ear and it glows a little red. Examples of more obvious changes in color on the skin are the lips, when someone has dark circles around their eyes, or when someone blushes. This is just a roundabout way of saying that skin isn’t a flat color and that things under the skin as well as on top of the skin(i.e. makeup, sweat, oil, hairs) are going to affect the perceived surface color.

That said you can’t really just add red and blend it in. In order for it to look natural the red you’re painting in needs to be the same value as the area, you are painting it into. This is why I prefer using HSV color space in order to keep track of my values.

To check values add a new layer on top of your layers and set it to [Color]. Fill the layer with black, white or any gray.

----Do not use a Hue/Saturation/Luminosity Adjustment layer to check values ---- because it does not reflect how we would see the values of color. For example, yellow looks lighter than the gray in is on and blue looks darker than the gray it is on.

If I use a hue and saturation adjustment layer to check my values this contradicts the information that was initially perceived. The results below are the actual results of doing this, the boxes are there but barely visible. You can very easily test this out yourself as well.

Whereas with the [color] layer the black and white values are closer to what was initially perceived.

 

Values are important because though contrast control this is how you give images a sense of light, form, and depth.

I usually gradually add in color since it can be strong off the bat. When choosing color variants to add in I move the hue on the slider or the color wheel to the color I want and slightly increase or decrease the saturation depending if I want the color change to be obvious or not. Also because the brightness of the base color can affect the value of the color I’ll shift the Value % by a few points.

Paying attention to the V% works for when working within the same base color. If you remember the values of yellow and blue are both very different while they are both considered 100%V on the Color Slider. So it won't really work as well if you're trying to add in vastly different colors like for makeup. This is just a technique to help with adding color variation within a base color.

For the clothing and hair I’m a little more straight forward with the rendering and refining the form and getting the materials to read better.

 

Eventually, I duplicate the layers I am working on and merge them down and clean up the edges and paint over to edit the silhouette. Rise and repeat until satisfied.

 

Overcast Lighting

From here, I can start integrating the character with the background better. Since it’s just an overcast sky, there isn’t a whole lot of color editing involved.

 

The image itself as-is is also fine, but since the character is pretty close to the camera I want the overall image to feel that way as well. I also want to create a better distinction between the character and the background as reflected in my initial lighting sketch.

 

I use a levels layer to darken the overall values of the character in order to create a more distinct separation from the background. To create a new Levels Adjustment Layer: Layer > New Correction Layer > Level Correction...

Level Adjustments are for contrast control. For editing the overall value range of the character we will be looking at just the bottom bar of the layers window which is the output range. Shifting the left tab to the right will lighten the blacks and shifting the right tab to the left will darken the lighter colors.

 

From this point, the painting can be used to apply different lighting scenarios. The bulk of the work is done so it’ll mostly be adjustment layers, color editing, and paint overs.

Daylight

I created a duplicate of my overcast lighting painting. Turn off the Levels layer.

 

Using a reference I recreate the sky colors I want.

 

Using my lighting sketch, since I kept my shadows layers separate I duplicate that layer, clip it to my rendered character, set it to [brightness], adjust the opacity to how I like it, and clean it up.

 

I select the shapes in the shadows layer by CRTL + Left Click the layer thumbnail. Create a new layer and fill the selection with blue from the background, set the layer to [color], and adjust the opacity. The ambient light from the sky while not direct is going to affect the shadow colors.

Selecting the shapes in the shadow layer again, I invert the selection, create a new layer, fill it with red, set the layer to [soft light], and use gaussian blur on the layer. I select the shapes from the shadow layer and create a mask for the [soft light] layer. I clean up the layer to only affect the skin. This recreates the effect of subsurface scattering.

I use level layers to bump up the brightness so the character feels like they're standing in sunlight.

 

Quick run down for the tabs on the levels adjustment window and what they do.

 

Left tab - Shadows

move shadow tab to right > affects darkest pixels in image

 

Middle tab - Midtones

move midtone tab left > overall image lighter

move midtone tab right > overall image darker

 

Right tab - Highlights

move highlights tab to right > affects lightest pixels in image

 

Because the goal is that I want the character to feel like they're standing in sunlight and the lightest color like the shirt is actually darker than the sky behind her, I moved the highlight tabs to the left so that the lightest pixels in the image are lighter.

I add in highlights on a separate layer through paint overs and layers set [screen] or [brightness]. I usually create highlights in the skin by using a light greenish/yellow. I also do paint overs on separate layers to clean up the edges. I do this as many times as I feel like it's needed.

I added a slight glow by duplicating and merging the layers down for the character, filtering it with Gaussian blur (Filter > Blur > Gaussian Blur...), set the layer to [overlay], and adjusting the levels by shifting the middle gray tab on the histogram to the left, this affects the midtones. I create a mask on a layer, fill it with black, add in where I want with a soft airbrush set to white on the mask. Adjust the opacity as needed and do additional paintovers as needed.

Sunset

I created a duplicate of my overcast lighting painting. Turn off the Levels layer.

 

 

 

Using a reference I recreate the sky colors I want.

Using my lighting sketch, since I kept my shadows layers separate I duplicate that layer and clean it up, create a new group, select the shadows from my shadows layer by CTRL+ Left Click on the layer thumbnail, invert it, and apply it as a mask to the group. This group is used to only affect the areas that will be hit by light.

 

Inside that group I make a Color Balance adjustment layer

Layer > New Correction Layer > Color balance...

I select the highlight gradient balance and move the Yellow-Blue slider towards yellow to add yellow to the image.

 

I also add a levels layer to brighten up the areas that are being hit by light by moving the right/highlights tab to the left.

Create a new group, select the shadows from my shadows layer by CTRL+ Left Click on the layer thumbnail, and apply it as a mask to the group. This group is used to only affect the areas that will be in shadow.

 

I make another Color Balance adjustment layer and a Levels layer. Cyan - Red slider is shifted towards red and the Yellow- Blue slider is shifted towards blue. This gives the shadows it's purpleish color. For the levels layer the right tab on the output slider and shift it to the left in order to reduce the lighter colors in the shadows and make it darker overall.

 

This is just another way to handle adding color to the light and shadows in order to integrate the character so they feel like they belong in the scene that they are in. You can definitely recreate the same method that was used in the Daylight section, apply it to this scene, and get similar results.

I add two [Soft Light] layers, adjust the opacity, and using an airbrush paint in red in areas where the light will hit on the hair and the face to simulate subsurface scattering.

From here I did some paint overs

I added a slight glow by duplicating and merging the layers down for the character, filtering it with Gaussian blur (Filter > Blur > Gaussian Blur...), set the layer to [overlay], and adjusting the levels by shifting the middle gray tab on the histogram to the left, this affects the midtones. I create a mask on a layer, fill it with black, add in where I want with a soft airbrush set to white on the mask. Adjust the opacity as needed and do additional paintovers as needed.

Night

Night lighting isn't that strong unless there's artificial light somewhere off-screen and out of frame or it's been edited in post. So having a night scene reference to match lights and darks will really help you aim to try and get the image to feel like night time. But either way there will be some distinction between light and shadow it just won't be as strong.

 

I create a duplicate of my overcast lighting painting. I keep my levels layer on.

 

Using a reference I recreate the sky colors I want. And using my lighting sketch, since I kept my shadows layers separate I duplicate that layer, clip it to my rendered character, set it to [brightness], adjust the opacity to how I like it and clean it up. I then adjust the levels layer to adjust the character to fit the image better. I also add in color layers to help integrate the character more into the background.

 

Essentially I just reused previous techniques mentioned in the Daylight and Sunset section I was just less systematic about it.

General breakdown of my layer useage:

-[Brightness] - for adjust value by painting in

-Levels layer - for adjust value in more general areas

• [normal] - paint over

-[color]- add color only without affecting values

I didn't clip the adjustment layers to the character in this case because the group containing those layers is set to [normal]. What that means is those adjustment layers will not affect anything outside that group.

[Lighten] is an additional way to add lighter values to an image. What I did in that layer, in particular, was use the gradient tool to paint in areas that I wanted to be lighter. Then I created a mask for that layer and filled it with black then painted in areas in white that I wanted to be affected by that layer on the mask layer.

 

This is a quick way to introduce smooth transitions in order to push form or color in certain areas while being nondestructive.

I added an additional color layer to push the color a bit more. I also added two [Overlay] layers. One using the method mentioned in Daylight and Sunset to add a glow and an unedited merged copy of my character. What [Overlay] does is just make the darks darker and make the lights lighter. I did this so the darker colors read darker.

Backlight

I create a duplicate of my overcast lighting painting. I keep my levels layer on.

 

Using my lighting sketch, since I kept my shadows layers separate I duplicate that layer, clip it to my rendered character, set it to [brightness], adjust the opacity to how I like it and clean it up. Using the same background from my sketch since it’s not really meant to be natural lighting. I then just the levels layer to adjust the character to fit the image better.

 

I add a [brightness] layer and airbrush in some rim lights. [Soft light] to airbrush in some red to simulate subsurface scattering.

Then just paintovers to clean up the edges.

Cast Shadows from Other Objects

I create a duplicate of my overcast lighting painting. I keep my levels layer on.

 

Using the same background from my sketch since it’s not really meant to be natural lighting. I then just the levels layer to adjust the character to fit the image better. I wanted the lighting to be a little garish and over-exposed so I moved the right/highlights tab further to the left.

Using my lighting sketch, since I kept my shadows layers separate I duplicate that layer, clip it to my rendered character, set it to [brightness], adjust the opacity to how I like it and clean it up.

I then add in a [soft light] layer. Selecting the shapes in the shadow layer again, I invert the selection, create a new layer, fill it with red, and use gaussian blur on the layer. I select the shapes from the shadow layer and create a mask for the [soft light] layer. I clean up the layer to only affect the skin.

I paint over this to push the garish effect of the lighting, integrate the image better, and clean up the edges.

And that is pretty much it. I hope this helps in some way

 

-l_aciel