vocab.design

typography

X-height

also called lowercase height, mean line

The height of a lowercase x in a typeface, measured from the baseline, which is what makes two fonts at the same point size look different sizes.

The x is used because it is flat top and bottom, so it has no overshoot: round letters like o and e are drawn slightly taller and deeper than the lines they sit on, or they would look smaller than their flat neighbours. The line the x reaches is the mean line, and the distance from the baseline up to it is the x-height. Everything else in the lowercase is measured against it: ascenders on b and h rise above it, descenders on p and g fall below the baseline, and the ratio between them is most of what gives a face its texture.

This is the answer to the complaint that two fonts at 16px look nothing alike. Point size is the size of the em, an invisible box inherited from the body of metal type, and it says nothing about how much of that box the letters fill. Verdana spends about 55 percent of the em on its x-height and Times New Roman about 45, so at one declared size Verdana looks a size or two larger. Designers handle this by optically matching rather than numerically matching: when you swap a typeface you almost always have to re-pick the size, and a font stack whose fallback has a different x-height will jump on the frame the real face arrives.

A large x-height is not simply better, though the last two decades of interface faces have leaned that way. It buys legibility at small sizes and on screens, because the part of the letter doing the identifying is bigger, and it holds up under low resolution. What it costs is word shape: with short ascenders and descenders, words flatten toward uniform rectangles and are harder to tell apart at a glance, which is one reason ALL CAPS (no ascenders or descenders at all) reads slowly. Large x-height faces also need more leading, since the lines are visually heavier and crowd each other sooner.

The measurement is available in CSS as the ex unit, one x-height of the element’s first available font, and in the font file as the sxHeight metric. The related unit cap is the cap height. Both are useful for the specific job of aligning something to the visual top of lowercase text, an icon beside a label being the usual case, where an em based value drifts every time the face changes.

Which word?

If you wantsay
explaining why two fonts at the same size look differentx-height
aligning a heading optically instead of by its boxcap height
explaining why 16pt in two fonts looks differentpoint size
naming the strokes that reach above lowercase heightascender
you need caps that match x-height exactlypetite caps

Sources