The Specimen Sheet: My Design Evaluation Tool

Before I commit to a typeface for any significant project, I make a specimen sheet. Not the foundry's marketing specimen — my own, set with real content from the project at the sizes and weights I'm planning to use. The foundry specimen shows the typeface at its best; my specimen shows it doing actual work.

What a Working Specimen Shows

A foundry specimen is designed to sell the typeface. It shows the most beautiful glyphs, at sizes and compositions that flatter the design. This is useful for discovery — finding faces I wasn't aware of — but it's not a reliable guide to how the face will behave in use.

My working specimen sets: a paragraph of the actual body copy at body text size, the actual headline copy at display size, the numbers and figures that will appear (are the figures oldstyle or lining? tabular or proportional?), and the specific characters that appear frequently in my content.

The Character Inventory

Every project has specific character requirements. A client in the food industry needs good ampersands, since "Salt & Pepper Co." will appear constantly. An arts organization needs quotation marks that don't embarrass themselves. A scientific publication needs proper fractions and superscripts.

I check these specific characters in every typeface I'm seriously considering. A face with a beautiful 'a' and an ugly ampersand is a problem in a project where the ampersand will appear fifty times.

Evaluating Optical Spacing

Type spacing — how the letters sit in relation to each other — is best evaluated in running text at the intended use size. Some faces that look beautifully spaced in a headline setting look tight or loose in running text because the default tracking is optimized for display, not body.

My specimen includes both display-sized settings and body-sized settings so I can evaluate the spacing in both contexts simultaneously. If both work, the face is strong. If one works and the other doesn't, I need to decide whether the tracking adjustment needed to fix the weaker context is something I'm prepared to make consistently.

Testing on the Output Device

A specimen viewed on screen doesn't tell me how the face will behave in print. Different typefaces behave differently under different printing conditions. A face with high stroke contrast — thick stems and thin hairlines — can show beautiful detail on a high-quality coated stock and lose the hairlines entirely on uncoated or recycled paper.

For important print projects, I print a physical specimen on the intended stock before finalizing the typeface choice. What I see on screen is a prediction; the printed specimen is the result.

The Time Investment

Specimen-making takes time — thirty to sixty minutes per typeface, done properly. When I'm evaluating five or six candidates, that's a few hours of work before the design has started. I think of it as insurance. It's far cheaper than discovering three weeks into a project that the typeface has an ugly ampersand that appears in the client's name.