Primate Panic Logo

I think logo design is one of the hardest single tasks in creating a game. Companies like Hamagami/Carroll (http://www.hcassociates.com/) make it almost their entire business. Any gamers will immediately recognize their work for companies like Blizzard.

For me, the toughest thing about designing logos is dealing with abstract forms and realistic forms at the same time. In most traditional logos, for example, the Nike swoosh, the logo is only ever intended to be displayed in one or two solid colors. Logos like this are almost 100% abstract.

However, in most games, like Warcraft III (http://www.blizzard.com/war3/), logos are often executed in full-color, with shading, highlights, and reflections. Essentially, the abstract forms of the logo are treated as real, physical, objects.

The challenge in this is that game logos have to follow all the rules of normal logos, such as looking good in silhouette, using unique, simple shapes, and being very easy to read, while also being rendered in a style that is consistent with the game it represents. The artist has to be a graphic designer and a painter at the same time!

For Primate Panic, the process of coming up with a logo was lengthy and a little frustrating. I started of with pages and pages of sketches. These are just a few:



I liked the idea of the somewhat primitive, 3D letters (upper left.) I had the idea of rendering them in a stone-like texture. I was hoping that this rendering style combined with the perspective would make them the alphabetic equivalent of the giant head statues on Easter Island. Here’s what I came up with (all the extra lines are temporary construction guides):



As much as I liked this idea in theory, I ran into some trouble when I started executing it:

1. The perspective made for a totally unreadable silhouette.
2. The rough stone texture caused the letters not to “pop” as much as I’d like.
3. Because the letters were so blocky, there were hard to read at small sizes.

I wasn’t ready to give up on this idea completely, but I decided to give my second favorite idea a try—the beveled letters over the banana leaf (lower left sketch).

My process in creating the logo took five steps.



First, I sketched a mockup of the logo in full color in Photoshop. This was encouraging, so I moved onto blocking in the basic form of the leaf and letter in solid colors. Early on, I decided that the letters looked better in yellow than in blue. After finalizing the shapes, I blocked in shadows using more solid colors. After I was happy with the general pattern of highlights and shadows, I began the tedious process of blending the shadow tones with intermediate colors. I don’t use the smudge tool or blur tool because I think they produce unnaturally perfect results, and they are also hard to control. Finally, I used Photoshop’s layer styles to add drop shadows and dark outlines.



Here’s the final logo. I’m pretty happy with it. It reads well, and also seems to evoke the themes and mood of the game.

Despite that, I’m glad I won’t have to design another logo for at least a few more months!

I have consistent taste in corrective eyewear…

…But my eyes do not have consistent taste in corrective lenses.

Melton in Oils

Melton
18 x 25 Oil on Board
2007 Jacob A Stevens


My latest painting is a portrait of my painting teacher Philip Melton. Yes, yes, I know, painting one’s teacher is the ultimate brown-nosing technique, but really I did it out of necessity: Nobody else in class felt like sitting under lights for half an hour while I tried to take a reference photo decent enough to paint.

Choosing a better palette

Green was always my favorite color growing up. Because of this, I was frustrated when I followed the standard recipe for making green by mixing blue and yellow, and got a grayish muddy color that looked more like canned peas than the bright, vivid, color I had been hoping for. I had been taught that the primary colors red, yellow, and blue, could be used to mix any color. That just didn’t seem true though—the colors I mixed using the primary colors were dull and muddy compared to the pre-mixed colors I bought at the store.

Still though, traditional painting is taught with the basic assumption that red, yellow, and blue are used to make all other hues.


Traditional primaries

The above diagram shows a very basic primary color palette consisting of cadmium yellow light, cadmium red light, and ultramarine blue. It is arranged in the traditional manner, with the three primary hues equidistant from each other. So-called secondary colors have been mixed from the primary colors. Notice how dull the purple and green mixes are.

Although this palette is the basis for almost all traditional painting palettes, and it can certainly be used to produce beautiful paintings, it has some limitations:

1. Secondary colors, especially greens and purples, are very dull.
2. Neutralizing a color (darkening it or graying it) requires the use of not 2 colors but 3, because red, yellow, and blue are not complementary (opposite) to each other.
3. A full range of dark colors is difficult to achieve, since all dark colors must be made using the basic recipe of red, yellow, and blue, which do not span the entire range of possible hues.
4. Mixing a color is non-intuitive because the triadic system does not reflect the physical reality of human vision.

A common solution to the above problems is to use a “split primary” palette.


Split primaries

This palette relies on the idea that more vivid secondary colors can be mixed by splitting each primary into a “warm shade” and a “cool shade.” Yellows are split into Cadmium Yellow Light (cooler and more green) and Cadmium Yellow deep (warmer and more orange). Red is split into Cadmium Red Light (warmer and more orange) and Alizarin Crimson (cooler, tending toward violet.) Blues are split into ultramarine, which is warmer and has more red in it, and phthalo blue, which is more green. The idea is that to mix a green, you’d use the more greenish blue (phthalo) and the more greenish yellow (cadmium yellow light.)

This works fairly well. The mixed green and mixed purple are more vivid that with the simple primaries. However, they are still not as bright as they could be. Additionally, this palette relies on the popular pigment alizarin crimson, which, despite its widespread use, is not lightfast. The palette also still has the limitation that no two colors are complementary. Three pigments (plus white) must be used to mix any dark or grayed colors.

In order to address these limitations, artists usually start adding more and more colors to the palette. Muted and pure greens, various purples, and earth tones are used to expedite color mixing and expand the available color gamut.

As a beginning oil painter, this philosophy leaves me frustrated and confused. How do I choose what colors to add to my palette? If I have various shades of reds, yellows, and browns, how do I know what to start with when, say, mixing a flesh tone? Do I have to buy dozens of colors to make paintings as vivid as professionals’?

Something about the whole system doesn’t sit right with me. Being a digital artist, I know that I can make any color on the computer screen by using various mixes of red, green, and blue light. No matter what shades of brown or olive green or turquoise or chartreuse I’m looking for, I can always find it simply by adjusting those three factors. Can painting be made so simple?

My goals in experimenting with alternative color palettes are the following:

1. Expand the range of colors I can produce, including vivid greens and purples, and the full range of dark colors.

2. Reduce the number of tube colors I have to buy.

3. Simplify color mixing so the process is consistent and uses as few colors as possible per mix.

Not knowing much about oil colors, I started off with a subject I’m more familiar with: mixing colors on the computer. I’d been told that mixing on the computer and mixing paints is totally different, but I decided to operate on the assumption that they are actually closely related if one could find a mapping between the two. After all, we look at paintings and computers with the same set of eyes.

Computer color works on the principle of additive color mixing. We have three different types of color receptors in our eyes. One type is stimulated most by blue light, one is stimulated most by red light, and is stimulated most by green light. Colors in between, like orange or purple, stimulate two or all three types of receptors.



additive and subtractive mixing

Therefore, when a computer screen displays a color, all it has to do is stimulate the right levels for each receptor. Rather than having to produce every hue directly, it uses combinations of red, green, and blue light to “trick” the eye’s receptors into believing that they are seeing other hues.

Printed color works much the same way, but instead of producing light in a dark environment, they absorb light in a white environment (the paper.) While a monitor emits red, green, and blue, light, printer inks absorb cyan, magenta, and yellow light. Cyan is the opposite of red, magenta is the opposite of green, and yellow is opposite of blue, so printed colors act sort of like negative numbers in math.

Since paintings absorb, not emit, light, my first approach was to use a palette based on printed color synthesis. This palette contains hansa yellow (chosen for transparency), phthalo blue (chosen for its cyan-like green tint) and quinacridone magenta (chosen as a lightfast alternative to alizarin crimson.)



CMYK palette

As you can see, this palette produces *slightly* more vivid greens. The problem, however, is that no paints, and especially not oil paints, are truly transparent like inks. Colors don’t just subtract from each other, they also mix. This limits the color range considerably.

The best solution I’ve found to this problem so far is to use six evenly spaced hues, where each hue has a complementary hue on the other side of the color wheel. This includes the colors red (cadmium red light), magenta (quinacridone magenta), blue (ultramarine), cyan (phthalo turquoise), green (phthalo green yellow shade), and yellow (hansa yellow). This is called the complementary color palette.


complementary palette

You can see from this diagram that using this palette dramatically expands the range of colors you can produce:


comparing gamuts

The inner triangle represents the colors that can be mixed with red, green, and blue. The outer hexagon represents the colors that can mixed with yellow, green, cyan, blue, magenta, and red. Clearly, the greens and purples are much more vivid.

Another great property of this palette is that each color has its complement “built-in.” You no longer have to mix yellow and red to get orange, and then mix in blue to get gray.


mixing neutrals

Overall, I’m much happier with this palette than those I’d been using before. It’s intuitive, has a great color range, and you don’t have to buy much paint. However, there are a couple major things to be aware of when using it:

1. Hansa yellow is both transparent and has low tinting strength. It would make sense to add cadmium yellow light when you need to make opaque, light, yellows.

2. Conversely, phthalo blue has incredibly strong tining strength (but is still relatively transparent). It's so strong that it's somewhat difficult to use. You only need a little bit compared with its neighbor ultramarine.

I've found both these shortcomings fairly easy to overcome, and am happy to continue practicing with this palette.

The site handprint.com has a lot of great info on this subject.