Color, Palette, and Narrative Clarity
Color communicates analytical category, intensity, and relationship. It is not decoration and it is not personal expression. Use it with intent.
Why color discipline matters
Color is the graphic variable that most powerfully attracts the eye. When it is used consistently and deliberately, it guides the viewer's attention toward the analytical argument. When it is used inconsistently or for aesthetic reasons, it creates noise that competes with the diagram's content. The visual restraint that makes a diagram readable is a professional skill — not a creative limitation.
Color as a variable
In diagram design, color typically encodes one of three analytical variables:
| Variable type | Color approach | Example |
|---|---|---|
| Categorical | Distinct hues for distinct types — no hue should imply more or less than another | Pedestrian routes in blue, bicycle routes in green, vehicular routes in red — the hues indicate mode, not hierarchy |
| Ordinal / intensity | A sequential gradient from light to dark (or cool to warm) for a variable that has a meaningful range | Pedestrian intensity from pale warm to deep warm — darker = higher intensity, lighter = lower |
| Diverging / bipolar | Two distinct hues meeting at a neutral center for a variable with two contrasting poles | Areas of positive ecological value in one hue, areas of ecological stress in another, neutral areas in grey |
Do not mix categorical and ordinal encoding without clear visual separation. If you use hue for category and darkness for intensity simultaneously, the diagram becomes illegible unless the two dimensions are explicitly labeled and visually distinct.
Working with opacity
Partial opacity on diagram fills allows the base plan or underlying layers to read through, reducing the visual weight of fills while maintaining their informational content. It also allows layered fills to compound visually — two overlapping transparent areas create a visually darker zone that reads as an overlap condition without requiring a Pathfinder operation.
Set fill opacity in the Transparency panel (Window → Transparency) or in the Appearance panel per fill. Keep stroke objects at full opacity — partially transparent strokes rarely communicate anything clearer than a lighter solid stroke would.
The base plan and diagram separation rule
The single most common color error in student diagrams is placing diagram content over a fully rendered, fully colored plan. The rendered plan competes at every point for the viewer's attention. The diagram's analytical content is lost in the background complexity.
The rule: the base plan visible under a diagram should communicate only spatial structure, not material character. Grey linework only — no fills, no hatches, no colors. All color in the image comes from diagram content, not from the base plan. If the diagram requires locating specific site features (a building, a path, a boundary), those features appear as grey linework at an appropriate weight in the base, and the diagram content is what gives them analytical meaning.
Building a diagram palette
A restrained palette — three to five colors plus grey neutrals — is almost always sufficient for a site analysis diagram. Start with one primary analytical hue for the diagram's central argument, and work outward from there only if additional variables require it. Test the palette by printing or viewing the diagram at the intended output size, not at screen zoom. Colors that look distinct at 200% zoom may merge at 100% output scale.
There is no prescribed palette for this course — color is yours to develop as part of your personal graphic voice. What is prescribed is that every color in the diagram must be able to answer the question: "What analytical variable or category does this color encode?" If the answer is "it just looks nice," the color is decorative, not informational.
Try this
Build a palette for one of your diagrams. Limit yourself to two analytical hues plus black/dark grey for linework and white/light grey for the base. Complete the diagram entirely within that constraint. Then evaluate whether the analytical argument is clear. If it is, you do not need more colors. If it isn't, the issue is almost always in the diagram construction or the analytical clarity — not in the color palette. Add a third hue only if a genuinely distinct third variable requires it.