On the first day of Ax-mas, we’re exploring: the Weight
Variable Fonts expose at least one axis where characteristics change across a range of values. The most common is Weight.
How bold is bold?
How bold is bold? This question is relevant beyond variable fonts: when students are learning type design, it isn’t uncommon for them to suggest a drawing is a “super heavy version” of their regular, only for the type design instructor to say, “is it really? Can it be pushed further?”
This ambiguity is observable in the weight axis: a Variable Font’s weight can range from 1 to 1,000, but a Weight axis set to 600 doesn’t tell you anything about how bold the drawing is. It tells you that it’s heavier than, say, 400.
The most obvious way a typeface gains weight as it gets bolder is through the thick strokes getting even thicker, but that isn’t the only factor in the weight of a glyph drawing.
Drawings might naturally get wider as they get bolder, so that also changes as you change the weight axis. There might be reflow. Weight isn’t multiplexed.
This makes sense: the weight feature is based on the end typographer’s perspective of typeface weight. Not the type designer’s perspective on how wide the vertical strokes are.
I don’t think there is much of an appetite for setting the numeric value of the weight to correspond to the stroke weight, either (more on that in a future email).
The weight axis can range from 1 to 1,000. If you tied the range to the stroke width, you might end up with a strange range. For example, if the thickness of the vertical stroke of the capital “I” was 220 units in your typeface in the regular weight, and 575 units in the black weight, in theory you could make the weight axis go from 220 to 575.
I haven’t seen many examples of this in conventional typefaces, but let me know if you have. Dmitry Ivanov’s Wavefont and Linefont tie the weight to the stroke weight, but they are graphic not typographic, so this is much more intuitive.
Instead, if you keep the internal and external weight values the same, the convention follows
font-weight values in CSS, where
400 is Regular, and the rest of the scale is built out from there.
It is possible to set up your source files to map own internal design values to the external values that the end typographer interacts with, but that is beyond the scope of this post.
Bold is relative
The inconsistency (and I use that without a negative connotation) of what bold means, existed before Variable Fonts—how bold a glyph drawing in Georgia Bold is has nothing to do with how bold a glyph drawing in Neutraface Bold is. With Variable Fonts, though, people using your fonts are more likely to see a numeric value for the weight. Two fonts with a weight of 400 don’t have the same weight, just as two fonts with Regular in the name don’t have the same weight.
In this regard, the Weight axis has similarities with font sizing: different fonts won’t be visually the same size, at the same numeric fonts size (different drawings, different x-height metrics, etc.).
I don’t think this is much of a concern for typographers, because it is much more common and understandable to want to make something bold relative to the regular, not how bold it is relative to another typeface.
You know a lot of this instinctively while making a font with a Weight axis, but hopefully it’s helpful to make those implicit details explicit.
Until next time,