Digital Coloring 231 Options

[Chaomancer Omega]

Digital Coloring 231

Lesson One: Basic coloring

Good morning, class. Or afternoon, as the case may be. Or night. Well, to heck with it. "Hi." Welcome to "Digital Coloring the Chaomancer Way". In this course, I'll be teaching you the very same techniques that I use to digitally color comics images. It should be noted that this is my way, not the only way, and not even the dominant way. It's just what works best for me. You may well find yourself making adjustments for your own personal preferences.

The image editing software that I use is Jasc's Paint Shop Pro version 8. However, any sufficiently advanced graphics suite should be sufficient; as long as it has layers and other basic image-editing tools and options, you should be able to follow along. I will be assuming you know at least the basics of using your graphics program; how to use the paint bucket, the brush tool, menus, etc. If not, just raise your digital hand, I'll explain what I can.

This first lesson will cover the basics of digital coloring: flat coloring. It and the lesson on shading will probably be the longest lessons of the course.

We start with this image that you, the students, selected in the polls earlier this month, Poison Ivy as drawn by Quinton Hoover:

I'll be honest, this is really a much more detailed image than we should be starting with, but I love Quinton Hoover's work, and the basic techniques are the same (except for that light ball, which we'll deal with specially.) If you want to follow along with a simpler image, the tutorial will of course be applicable to that as well.

Fortunately, this image is nicely inked and very clean, so we don't have to do any prep work on it. (Actually, it's not so much good fortune as by design; all the images in the poll had minimal prepping. Clean-up tutorials will follow in later weeks.) All we need to do is have this in a format where it'll be nice and easy to add colors to it.

Some people like to add layers on top and use the multiply blend mode; that's one way of doing it, but I prefer to avoid messing with the blend mode for the general coloring. It's a good technique for lighting effects, but it's not needed for general coloring, and my technique adds a bit of flexibility on the background -- and some forgiveness if color areas inadvertently overlap.

What I first do, is create an image the same size as the existing image, and then fill it completely with black with the paint bucket tool. (I trust I don't have to show that.) Next I add a mask layer, basing it off the original image. In Paint Shop Pro, this is accessed by going to the Layers menu, selecting New Mask Layer --> From Image. Select the original image file, source luminance, and invert mask data.

What this does is it creates a "mask" above the black layer. A mask is a layer that blocks out what's below, making it invisible. Black in the mask layer does the blocking, white lets things through. In this case, we inverted the mask data from the original image, giving us the reverse effect: white from the original image makes the layer below appear transparent, black leaves the black below intact, and shades of grey are partially transparent. We then select the layer group, and merge them together using the Merge Group command (right click or find under the Layers menu). This gives us a single layer which looks like this:

The checkerboard pattern is PSP's way of telling us it's transparent. You can close the original image now if you want; we won't be needing it again. Now, if our lines here weren't dark enough for our purposes, we could duplicate the layer we have by selecting Layer-->Duplicate, and this would result in a darkening of the lines. But that's unnecessary for this image. Our next step is to create another raster layer. Fill it with white, and then move it so that it is below the lines layer in the layer heirarchy. You can do this either on the layer pallette on the right-hand side or by selecting Layers-->Arrange-->Send to Bottom. Now we have an image that looks like the original image, but the lines and the white background are now in two separate layers -- and we can add layers of other colors in between. Make sure the image is in full color by going to Image-->Increase Color Depth-->16 million colors (24 bit). Now would be a good time to save your file if you haven't already done so.

Next it comes time to start adding the colors in. You add layers in between the background and the line layer. You can add in layers one at a time, adding the flat colors as you go, or you can create all the layers you will need at once. In either case, though, you need to have a degree of awareness of what needs to be in a separate layer. For this reason, especially on a detailed image like this, I usually create my layers one at a time. It gives me a greater ability to change my mind without having to rework as much, and I don't have to keep track of as many things without seeing them.

What do you need to have a different layer for? Two things. First, if something is in a different color, make it a different layer. Ivy's red hair and her green suit should not be in the same layer. Secondly, if two areas are distinct, but would interfere with each other's shading, it's often a good idea to have them in separate layers. For example, Ivy's arm, and her torso. They're the same color, but shading them will be easier if they're separate, as one casts shadows on the other. Now, if something is the same color, and doesn't interfere, it's okay to have it on the same layer. The ivy fringes on her tunic can be on the same layer with no trouble. It is helpful, incidentally, if you rename your layers when you create them to something that will help you remember which is which.

So now that we have some idea of how to choose our layers, it's time to get to coloring. You pick the colors you're going to use for the flats by the medium tone. This will give you more flexibility when it comes time to shade. I will start, in this case, with Ivy's arm. (Incidentally, the flesh tone I'm using is #FFE2CD; Ivy, after all, is fairly pale skinned, at least when portrayed as having flesh-colored skin. I am using #FF6815 as the base tone for her hair.) Essentially, you want to color in between the lines -- actually overlapping a bit with the lines, so that the entire area is colored. It's okay to overlap, because of course, the lines are on top so it isn't going to be seen anyway. Any tool you use to color it is fine; personally, I use the brush tool at 100% hardness to outline the area, then use the paint bucket on RGB match mode and 0 tolerance to fill in the area.

Now, I know what you're thinking. "Hey, you colored over the leaves too! They're not supposed to be flesh colored!" This is where this layering technique gains us an advantage over the multiply technique. Here, I can just create the "leaf" layer above this layer, and the green will completely block out what's below. If I multiplied, that wouldn't work; I'd wind up with a green/peach blend on those areas, and it'd look pretty ugly. So if I were to multiply, I wouldn't be able to be a bit lazy and just ignore those lines on the lower layer. As it is, I only have to do close-up detail work on the leaf layer. Thus, as we add layers, some of the layers will be placed at different levels in the heirarchy so we can control what's "on top" and visible. In general, if something appears "above" another object in perspective, it should be a a layer above that object. So we should have Torso below Left-arm-bracelet below Right Arm below Right-arm-bracelet.

Now we can proceed with coloring the rest of the layers. I recommend saving after each layer. Be sure to color the eyes white, rather than just leaving them transparent to the background. If you ever want to change the background color later, you don't want her to suddenly have the whites of her eyes turn purple. Leave the ball for last; that's a bit of a special case. I'd recommend the hand that's underneath it to be part of the same layer as the torso to make shading easiest later on. You may find it helpful as you go along, to temporarily set some of the layers to "invisible" while you pick out some of the finer details (such as the leaves in her hair.) This is easily done by clicking on the "Eye" icon in the layer palette.

When you're done, you'll have something like this, depending on what colors you chose:

That'll be it for this lesson. We have the flat-colors in, next time we'll shade the lady. Until then, the floor is open to questions.

[the superfly]

Excellent work Professor!

[Chaomancer Omega]

Just wanted to say I haven't forgotten this, and won't let it lie fallow for too long, but I've got some other things on my metaphorical plate right now that are moderately more time-sensitive. In the meantime, if anybody has questions, please ask.

[Chaomancer Omega]

Digital Coloring 231

Lesson Two: Basic shading

:: YAWN ::

Oh? What? There's still lessons to be held? Whoops.

Right. Guess I'll get to it then. I can't promise any regular release schedule, but I'll be putting things in here from time to time again now.

Okay, class, listen up. We flat-colored most of Poison Ivy in our previous lesson, now it's time to add some shades and highlights to her so she looks a bit more like a modern comic drawing and a bit less like a Hanna-Barbera cartoon shading. Not that there's anything wrong with that, mind you. Just not what we're going for.

We'll start by showing the basic technique, in this case, on her arm.

The first thing we do is select the layer her arm is in, and duplicate it. We then hide this new layer. We'll be using it later. Re-select the original arm layer.

Next, we use the magic wand, with its mode set to RGB match and tolerance 0, on the arm layer to select the colored area. We then go to the Selections menu, to Modify, and expand it. How much is dependent a bit on what you're working with. For small detailed work, like hands often are, a few pixels is sufficient. For larger stuff, 10 pixels is usually good. In this case, I have some of both, so I'm going to say 5 pixels. You'll find out why a few steps down.

Use the eye-dropper to grab the color of the arm. Next, grab the paint bucket tool, with match mode set to NONE so it'll fill anything. Fill the selected area with the selected color.

Next we need to select the colors we'll be using to shade with. We'll need to select a range of shades based on the same color. Normally I choose five, but here I'm shading a large area, and my selection border isn't very large, so I'm going to go with seven instead for a smoother transition. (Again, the reason will become apparent in a bit.) Remember that just making something lighter or darker by altering the luminance value isn't quite true shading; you have to adjust the saturation, and sometimes the hue as well. Notice how I made the darkest shade a bit redder. Here's the range I ended up with, with the base value in the middle.

It's best if you keep this color palette in a separate image window. You'll be coming back to it a lot.

Now that I have my colors selected, I start painting them in the logical places. It doesn't have to be an exact science, just approximate where the shades should be dark, and where they should be light. You may find it helps if you mark a spot in a top layer as a light source. I sometimes do without though, and for expediency's sake, I'm just going to wing it here. When coloring, be sure to color in the "bleed" that we created as well; just match it to the area it's adjacent to (which will generally be dark). The result will be something like this:

Now that we've got this, how do we turn these approximate shades into something a little smoother? We let our graphics program do the work for us. Go to Adjust --> Blur --> Average. This is why we wanted that bleed; if we were to use this without it, it'd blur in some of the transparency, which would be undesirable. We get the following dialogue box:

The filter aperture is how wide of a range the blur will have. The bigger the aperture, the more things are blurred. A smaller aperture means less blurring and preserves more of the original delineation. If we had a really small detailed thing, we'd want a small aperture. For a larger one, a larger aperture. Here, we're mixing a bit, which we really should have avoided, because hands are trouble enough, but oh well. We'll just have to settle for less-than-optimal shading on the hand. The goal is to blend the colors together and not turn things into a muddy mess. In this case, I'm going with an aperture of 19. You can fiddle with it to get things to just the look you want.

When you've got that, deselect by pressing Control-D. Go back to the copy of the layer, and use the magic wand to select the original boundaries again. Invert the selection, and go back to the shaded layer. Delete, deselect, and now's a good time to save your progress.

Here's what we've got:

Next we use the same colors and technique to shade the rest of her flesh. You will use different apertures in different places, but the rest is pretty much the same. You may sometimes find yourself using the Hue/Saturation/Lightness controls after you've shaded an area to make it match up better with the other areas. That's all right.

Go ahead and shade the leaves, vines, and so forth as well. Just use the same technique, with varying shades of green and brown. We'll get to the hair in a bit. For the vines and leaves, small apertures are usually best; for the lips, go with the smallest of 3.

When we're finished with everything but the hair, we'll have something like this:

Now it's time to color her hair. Hair is one of those things that really has a different color depending on where the light hits it. So here's the color palette I'll be using:

With the dark color, we're going to trace Quinton's lines. With the yellow, we're going to fill in the bright areas, with a bit of the orange-yellow around it to facilitate blending. We'll also use the orange-yellow in semi-bright areas. We'll wind up with this before we blur:

Then, as with the rest, we let our Average Blur command do the work. I used an aperture of 5.

Our final result looks like this:

Now, we're not quite done yet. With Lesson 3, we'll finally take care of that dazzle-ball.

[Chaomancer Omega]

Lesson Three: Light casting

Welcome back, class. At long last, it's time to handle that dazzle-ball. At the end of this lesson Ivy will finally be fully colored.

We're going to start our little light show off a little different than we have with the rest of the coloring job. This time, we're going to create the new layer on top of the lines, not below it. And when we create this new raster layer, we'll be choosing a different blend mode: Overlay.

Hard Light does just what its name implies: It makes the layer act as though it's a strong light source. No wimpy little flourescent bulb for us, no sir! (That's Soft Light). This is a strong-as-the-sun light source if we need it to be.

Hard Light blends the colors used on this layer onto the colors of the lower layers. Thus, if we were to color in yellow, we'd see the yellow, but it'd be blended into the colors below. We could paint a yellow square over her head, and we'd be able to see the flesh tone and red hair beneath with a yellow cast.

Mind you, we're not going to do that.

No, what we're going to do is paint a green sphere. Take note that the light color will be brighter than the color you picked to paint with. Adjust accordingly.

We provide the dazzle-ball with a certain amount of color variance as well. We will, of course, let the Average Blur handle things for us later.

In fact, this is as much as we want to do with the ball itself. We once again use Average Blur, and we set the aperture as high as it'll go, to 31. Notice that this time we're not worried about it blurring into transparent pixels; we want that: it'll give it a nice glow effect.

Now we want to do some highlighting. First, we create another layer on top of the dazzle ball layer; again, it will have a Hard Light blend mode.

First, since this dazzle-ball is creating the vines, we will take the green color we used earlier and paint over the vines using it. As we get further from the dazzle ball, we'll use a smaller brush, until we're down to one pixel, and then, nothing at all.

We use an Average Blur with Aperture of 13, and this is our result:

The glow on the vines is more subtle, but it's there. That's all we really want.

We create yet another Hard Light layer for the reflections of the dazzle-ball's light off of Ivy's body. We use both the green and the yellow for creating the resulting highlights. We want to use a very small brush, though; one or two pixels only. Before we blur, we have this:

And after an Average Blur with maximum aperture, it becomes this:

Are we done yet? Not quite. We could be if I wanted to be. But I'm going to give this a nice dark background. And to make things continue to blend well, some of my Hard Light layers will be changed to Screen, which is always a brightening effect. Here's what we end up with:

And so, at long last, we have completed our Poison Ivy coloring. It isn't as high a quality as a true comic-book colorist could do (well... some colorists anyway), but it still is something I don't mind having my named attached to.

Future lessons in this class will be a bit smaller in scope, and also will come on a sporadic basis unless people have questions. (Roughly translated, if you want to see more, talk!)

[Slovman]

Swiggity-sweet, man! Great final product.

I would show future coloring projects I'll do with your lessons as a guide, but they're of an...shall we say...adult nature.

[TheRidge]

G6BH's??

[Leeds Knight]

What does that mean exactly?

[Chaomancer Omega]

You'll find out when Slovman gets around to posting the jargon lesson in the Halls 101. Whenever that is.

[DrQuinzel]

Woohoo!! Very, very nice. Although I know I'll never be able to do that myself, even with the lesson to guide me.


I want more pretty pictures!!!!

[Thor665]

Prof Omega?

In the final image Ivy's hands almost look like they've been dragged back out over the top of the dazzle ball layer - is this simply the effect of switching the dazzle ball lighting effect from hard light to screen, or was something else involved?

[Chaomancer Omega]

Just a side-effect of switching to screen.

[Slovman]

My first attempt, Professor!



Not quite finished with the car...as I'm a bit puzzled as to how to color the interior of an engine...but any suggestions other than that?

[Chaomancer Omega]

Not bad, not bad. The engine I'd probably make dark grey. (I'll include a tutorial on metal shading some point in the future, but "cartoon style" is quite all right as well.) Other than that, I'd consider adding shadows to the cape; add a bit of a silhouette behind her to give the image more texture.