How to compare xheight between thin version to black version

let’s just say I’m going to interpolate and instance to create a font family, I’ve already drawn the light version and will draw the extrabold version
what I’m asking is is there a difference in xheight between the light and xtrabold versions when drawing it?
I checked some of the fonts that have been released, some are different, some are not… whether. is there a standard rule?
Thank you

A lot of times, the Bold needs to be a bit bigger than the Light. Type out some lines of Light text and add some words in Bold and print it. Maybe play with the font size in the app where you set the font.

No rules, have to trust your eyes.
A big factor is whether the bold is high contrast or monoline–if it’s pretty thick at the meanline and baseline, it will likely have to be noticeably taller. Your eyes don’t just compare the top contour, they also are reading where the “heartlines” of the strokes are.

The more the contrast, the less x-height expansion is needed. Bodoni less than Helvetica.


I understand now, optical is more important
just my thoughts, draw the boldface first or the thin line first?
or make 3 masters with thin, regular, and most bold, my assumption is regular for text needs, and boldface for display
from the instance I can’t directly determine the ratio of stem thickness and xheight

Unless your light or bold aren’t particularly light or bold, very often three masters will be helpful. I’m not sure if the order you draw them in matters all that much, though it can be helpful to sketch out some glyphs in all three early on in order to uncover issues in one master that might change your thinking about the others.

I usually draw all my masters at the same time. There are so many parameters that need to worked out early on and that influences both ends of the axis (contrast, horizontal and vertical proportions). And if makes it possible to use the RMX-tools

I would suggest that the eye perceives that in linear typefaces the counters usually are less high in the bolder weights. This causes bolder weights to look smaller than lighter weights. As we do not expect bold weights to look smaller, it seems logical to compensate for this.

This effect is similar in typefaces with relatively heavy horizontal strokes (broad nib contrast, Garalde, old style aka Dynamische Serifenschrift or even Renaissance-Antiqua among the Germans in the audience).

Agree. On rereading my first post, I said your bold might have to be “noticeably taller,” but “noticeably” is really the wrong word here–the whole idea is finding the heights at which the weights read as the same. If the reader “notices” a difference, you’re not there yet!

From the font that I’m developing geometric semi grotesk, I draw a double story, thin stem = 20 and regular stem 75 seems fine, but for black type with stem 166 with complex images like ‘a’ it’s difficult to make the same xheight, instead ‘a’ is like short and wide
and for the horizontal stem it is no longer the same between n and a in the black
is this what is meant by “This causes bolder weights to look smaller than lighter weights”
Screen Shot 2022-08-19 at 07.30.10

You might also want to consider the height of the stroke’s centerline. When thickening the strokes , the centerline may drop significantly, making the letter appear smaller at the same x height, because the centerline is now lower than in light or normal weights. I’d suggest to not only refer to the topmost points in your letters, but the centerline of the stroke, as well. Your „n“ gives a good example of what I mean, I guess.

1 Like