Getting Started

Items you will need:

• A computer to view this tutorial. I mean, create digital art.
  
  
• A CG program for your computer. Among the most popular are Photoshop (any version can work reasonably well) and Painter. Other programs like Illustrator and Flash can come in handy depending on your needs. My program of choice is  Easy Paint Tool Sai which has a reputation for being inexpensive and extremely easy to use. It is also capable of providing beautifully smooth lines and coloring.
  
  Some more obscure options include programs like Nekopaint (free), Manga Studio, and IllustStudio (Japanese only at the moment), none of which I can recommend personally but have heard decent things about.
  
  
• A digital drawing tablet; preferably one with pressure sensitivity. I use Wacom Intuos3 and 4, both medium sized; easy to carry around but not so small that they're hard to draw on. Both are great but the Intuos4 has much higher pressure sensitivity, a big plus.
  
  



Wacom Intuos4

With everything in order, I open Sai to begin my drawing. The process is generally broken down into the following steps:

1. SKETCHING
    1.5. ROUGH COLORS (interchangable with step #2)
2. LINEART (interchangable with step #1.5)
3. BLOCKING IN COLORS
4. SHADING
5. FINISHING TOUCHES
6. SAVING THE IMAGE

One more thing to keep in mind: this tutorial is for beginners and beyond with an understanding of digital CG and Photoshop terms/tools/functions. For example, knowing what a layer is and what they do is vital. Photoshop tutorials are everywhere on the internet. For things that aren'teverywhere on the internet I'll try to give more detailed explanations.

*Some images can full-viewed!

easily the most fun or frustrating step, depending on how well your vision matches what comes out on screen. simply open a digital canvas in your program of choice and start sketching.

I usually start with a high resolution, 300 dpi A4 canvas (2480x3508 pixels, a default option in both sai and photoshop) which is a personal preference. 300 dpi is the standard for print production (more info here:http://en.wikipedia.org/wiki/Dots_per_inch) and working at high resolution means you can shrink the image down later and the lines will look much finer and smoother. Not to mention you can fit a lot more details in!

Gesture drawings (“Quick scribbles to show movements of the body that quickly records an entire image”) are helpful for creating movement, composition, and feeling in your image.

1. the flexibility of the canvas size also should not be overlooked. This is one of the big advantages of working digitally, since it's not easy to change a canvas size when working traditionally. I may start with a vertical A4 canvas, but if i run out of space on the right side of the image, i won't hesitate to expand it.
   
   

   1.5. ROUGH COLORS:
   

   optional; you can quickly lay down the colors first, with any tools. I create a new layer underneath the sketch (or you can place it above and set the layer property to darken/multiply) and color away. Sometimes blocking in color helps find flaws in a composition that might not not be so obvious with just black lines and a white background. Also a good time to explore different color schemes.
   
   

   
   
   

   I save my files after sketching as a .psd--even in Sai--which makes for easier transition between programs. Saving in .sai format is recommended if you intend to use Sai's vector tools (which I won't be covering). The vector lines produced in Sai lose their vector property if you save the file as .psd and then close it.
   

   
   

   
   
   
   
   step #2: lineart/inking
   
   

   one point that i will stress again (and that people keep asking me about) is that working at high resolution (big canvas) allows you to shrink down your image later so that the lines will become much smoother, finer, and neater. this is one of the "secrets" that makes my lineart appear so smooth.
   it helps to have a steady hand and confident strokes when you ink your lineart, but shrinking the image helps a lot.
   i prefer to ink my images freehand for a more natural look, using the vector tool in sai only in very rare cases. 

   
   now, create a new layer over the sketch and after choosing your brush of choice, ink neatly over the sketch. The opacity of the sketch layer should be lowered, to around 5~20% so that the sketch lines don't confuse you and interfere with your inking.

1. 
   ink layer over a sketch layer at low opacity
   

   The brush size is really up to you; in photoshop i wouldn't go over 3 pixels unless the inking style calls for thickness variation. In sai, 0.7-2.0 pixel brushes are my most frequently used. 0.7 becomes very fine when the image is shrunk, so it works if you don't want your lineart to be so apparent.
   

   
   this is inked with a 2 pixel brush in sai, at 100% view

   Inking neatly:  it's important to make sure that the lineart ends meet where they're supposed to and leave no gaps. This is for the next step of filling in flat colors, because any gaps or breaks in lineart will cause the magic wand to “leak out” and select whatever's outside of that gap. using a paint bucket to dump the color will then fill in the entire selected area. It's not a huge problem but can be annoying (and slightly time-consuming) to fix.
   

   
   (in sai, the magic wand selection area is dark periwinkle blue,
   the lineart gap is circled in red)

   
   For scanned lineart/lineart without a transparent background: this requires cleaning and a quick step with masking in photoshop to remove the white background, but won't be covered in this fully-digital style tutorial. I'll have to make a separate tutorial for this some other time~

After completing the lineart, i either hide (also known as turning off layer visibility) the sketch or delete the sketch layer completely if i have it saved somewhere else already. Now it's time for color.

step #3: flat colors

get your magic wands out! before you start though i'd recommend locking the layer transparency of the lineart layer, which i usually keep at the top. This is to avoid accidentally filling color onto the lineart layer, which i've done before and it wasn't fun. Always keep the lineart layer separate.

a quick overview on locking layer transparency (in photoshop):

^this is how you do it


^this is what it does

this step can apply to any transparent layer that has pixels on it, including the lineart (if you lock the lineart, you can use a brush to color the lines themselves).
locking a layer's transparency is a pretty straightforward concept that is rather hard to explain, so i recommend trying it out in photoshop.

here's my attempt: in the example above, the first image shows a patch of light yellow, which are the "opaque pixels" on that layer. in picture 2, i colored messily using a pinkish beige without locking the transparency, causing the colors to go out of the lines. picture 3 is what happens when i do the exact same thing, only now with transparency locked.

here's the opacity lock checkbox in sai:

a new canvas' transparency in photoshop vs. sai: it should be noted that the black/white checkered background (indicating transparency) only applies to photoshop. new photoshop canvases are NOT transparent (they have white backgrounds) but it can be removed with the background eraser tool in the toolbar.
new sai canvases are automatically transparent (they don't have the white/gray checkers to show for it)

 

anyway, back to blocking in flat colors!
here's the flat-color breakdown:

the above was done in sai, which has a sensitive magic wand that can select very precisely the entire area inside lines as long as the lines are smooth. If you're filling out flat colors in photoshop and the magic wand is leaving a slight gap between the lineart and your selected area, go to the top menu and choose select->modify->expand (usually 1 pixel will do the trick) and then you can use a paint bucket to fill in colors just as well.

the thing with magic wands is that there will always be nooks and crannies that it won't be able to reach, so once you fill in the colors, deselect it and use a pen to manually fill in the spaces that still need filling in. pretty basic logic that goes a long way:

when i finish laying down the flat colors, i lock the transparency for every layer to shade directly on it; creating new layers for shading is unecessarily bothersome for me and limits the amount of control I have over how the shadows turn out.

step #4: shading
(and color application)

without a doubt what people are most curious about, shading is where you really begin to show the moods and colors of your piece. There's no correct way to go about shading and no rules about how shading should turn out. even with an anime-style lineart, you can color it to the point that it looks photographic.

simply choose a layer with flat colors (with locked transparency), take a pen or brush of your choice, choose the color you want for shading, and add shadows directly onto the layer (like in the transparency lock image 3 pictures above).
i tend to use light flat colors and add darks from there, but the other way around definitely works too.
I use a combination of the pen tool and watercolor (sometimes airbrush) in sai to shade, depending on the look i want. The blur tool in sai is also helpful for getting gradients/very soft shading.

in this tutorial i won't be covering individual topics like how to shade an eyeball, color a face, or how to do lighting, since they all pretty much use the same technique and are learnable by observation. here are a few examples, all using the same technique (in sai) with the pen tool & watercolor tool for blending, with the occasional airbrush:

naiyami giving you the look. yes, that look, the "did you eat my nachos?"

Colors are too complex for me to cram into a few sentences in this tutorial (that, and i'm still trying to grasp colors myself). There's too much to talk about, and in some of the feedback readers have mentioned this color section has only skimmed the surface but it's all I got for now, so bear with me!

Having a good understanding of the color theory (http://en.wikipedia.org/wiki/Color_theory) is helpful to the n'th degree. that doesn't mean you need to know all the terms, but rather how colors interact with one another. For example, in the rubik's cube illusion below, the two squares in the center are actually the same shade of brown.

The reason one appears almost orange is because the surrounding colors are relatively darker, while top center square appears as a darker shade because it's surrounded by lighter colors. this touches on value and contrast, which you can read about here. When cging, you're creating an illusion with colors that give depth and form that in turn evoke emotions, so know your colors and how to use them.
it takes a lot of practice to become familiar with working convincingly in color, but it'll all be worth it when you suddenly realize that your illustrations are turning out the way they're supposed to!

here is a home-made color wheel for your reference:

notice how all the colors directly opposite from each other look really nice together?

 

COLOR RELATIONSHIPS (these images explain it a lot better than i could ever. source is listed in the credits)

Monochromatic Relationship: Colors that are shade or tint variations of the same hue.

Complementary: Those colors across from each other on a color wheel.

Split-Complementary: One hue plus two others equally spaced from its     complement.

Double-Complementary: Two complementary color sets. (by the way, it can be any two pairs any distance from each other)

Analogous: Those colors located adjacent to each other on a color wheel.

Even the most basic ideas like complementary colors are important to keep in mind, as they can make for solid color schemes. For example, shades of green and red are complementary, or 'opposite as can be' and can create a visually absorbing image.

varying shades of complementary colors

Also, colors are not black and white! (i kind of like that pun actually :P)
there are many ways to shade a color. here's an example done with bright fuchsia pink:

and another example shading an arm; skin tones aren't only two colors, they can be lots of colors; a tinge of orange inbetween light and shade, or a light blue from a secondary lighting source.

 

Working digitally is very forgiving so there's nothing to lose from trying different shading styles and color combos; i've mimicked everything from watercolor to oil (to varying degrees of success..), all with the same tools in sai.
Brush settings (ex. percentage of persistence or color blending with the watercolor tool), do make a difference so tweak them around and find what works for you.

step #5: finishing touches

(aka. Post processing): this is also quite an important step for me, and is where i move from sai to photoshop. photoshop is THE program to use if you want to wrap your cg up into a nice little package complete with a bow on top. This step in particular requires a good handle of photoshop's functions, but experimenting can always lead to new discoveries and unexpected results.

Here are some of my most frequently used processes:

a. lineart manipulation
b. color adjustment
c. color layering
d. Texture

*some of these effects can also be done in sai, but photoshop is definitely recommended for this.

a. lineart can hold an image together, so it's best not to overlook it:

b. color adjustment is optional and pretty straight forward, employing use of contrast changes, color balance, select color, curves, etc. you can adjust the layers individually or flatten the entire image and adjust from there.

c. Layering is a helpful technique to adjust colors if they didn't turn out the way you wanted. Layers of color can change shades, tones, brightness, contrast, etc.:

above is an example of layering over a grayscale painting with color layer set to multiply. There are many layer settings you can try, some more useful than others.

d. textures are like seasoning on pasta and can really spice up an image. I use it when necessary and usually try not to let the texture overpower the image.
Below is an example of basic texture application:

texture layers, like everything else, are fully customizable. If you want texture on a certain part of the picture, erase the texture everywhere save for where you intend to leave it. The opacity of the texture layer can be changed, the contrast can be upped for more noise, the color can be changed--basically anything you like.

Textures are available from many users on deviantart (reading their policies is common courtesy) as well as from sites like cgtextures.com. many are royalty-free so you can use it for whatever you want, but it's always nice to credit and spread the word about someone's effort to bring to you quality textures.
so, be sure to credit the maker of the texture when appropriate. to make your own, you can take a camera and go out to snap walls and sidewalks, or scan your own crumpled paper and watercolor washes.

step #6: saving the image

believe it or not this is also an important step to do correctly, as the saved image will be what everybody sees. or maybe i'm just nitpicky, i don't know!