How much bigger do you like to make your x-height for Bold?

Options
Adam Ladd
Adam Ladd Posts: 276
edited July 28 in Type Design Technique & Theory
Just curious others preferences/techniques between your lightest master vs boldest master.

Do you typically like to add to the x-height for the bold, or keep it the same as the light?

Example: Light = 500 units / Bold = 510 units

I've done both.

Does it vary based on the typeface proportions, or do you always default to making it larger for optical reasons (by a little or a lot)?

Comments

  • Mark Simonson
    Mark Simonson Posts: 1,758
    edited July 28
    I usually only do it with lower contrast faces if I do it (I often don’t). Beyond that, it’s fairly subjective how much to do it and the concept of the design.

    In Mostra Nuova, it’s pretty extreme, going from 423 in the Thin to 517 in the Black (122 increase). (The cap height and ascender are 685.) This is mainly because I was adding lowercase to a formerly all-caps font that gets extremely bold. I wanted to keep the x-height smaller in the lighter weights, following Koch’s Kabel as a model.

    In Viroqua, it’s more subtle, going from 459 to 498 (108% increase). (The cap height and ascender are 681.) In this case, the face had a fair amount of contrast, but I thought it looked better with the adjustment since both the thin and thick strokes get pretty heavy on the extreme end.
  • John Hudson
    John Hudson Posts: 3,451
    I usually add a little height in the bolder weights, the amount depending on the stroke contrast.

    More importantly perhaps, I make my italic lowercase slightly shorter than the roman, because the slant makes baseline-to-xheight distance appear longer.

    A really tricky thing is coordinating a typical Latin lowercase height with a traditonal Greek, because of the different stroke modulation patterns. As they gain weight, the Greek will start to appear shorter than the Latin because of the verticallt comressed counters. I’ve not found a particularly good solution for this.
  • Ray Larabie
    Ray Larabie Posts: 1,463
    The design of the capital T can be a factor in that determination. There are also practical considerations with the 'f' and 't' crossbars.
  • michele casanova
    michele casanova Posts: 54
    Thomas Phinney said:
    Robert Slimbach made the caps smaller in the bolder weights instead of increasing their x-height.
    As an example, I'd like to point out the "Titillium Web" font. In a recent redesign (as "Titillium Sans Pro"), the proportions were changed.

  • Kent Lew
    Kent Lew Posts: 1,004
    I always make my bolder weights slightly taller x-ht than the lighter weights. How much difference depends on the style (degree of contrast) and scope of weight change. Typical amounts for me are in the 10-20 units range. 
    However, a recent sans-serif in development runs a whole 40 units increase in x-ht from Thin (48 stem) to Black (204 stem), with 24 units x-ht between Regular and Black. But some of that extra increase is more about enabling room to keep counters open while still packing on weight, rather than apparent height.
    Ultimately, the eye is the arbiter — what appears balanced when weights are set inline with each other. Considering both internal and external spaces.
    I was taught the received wisdom of italic slightly smaller than roman, but have mostly not abided by it. The few times I tried it, the italic just appeared smaller, contrary to the rationale. Probably because I favor pretty slight italic angles, usually <10° off vertical. 

  • Adam Ladd
    Adam Ladd Posts: 276
    edited July 29
    Some really interesting nuggets and insights here.

    In the past, doing some graphic and logo design work, it was hard to not have the heights align (when pairing weights/styles) for graphical/grid reasons. Though I've found I'm leaning more towards the increased height with the type designs (but it's case-by-case).
  • Jasper de Waard
    Jasper de Waard Posts: 647
    When you think about it, it's actually a little odd that we increase the x-height but not the cap height in bold weights. 

    In my mind, that is because usually no caps exceed the cap height, and many straight strokes run along it (BDEFPRTZ), such that messing with it would be visible and likely appear like an inconsistency to the casual viewer. The x height is kind of obscured by the many ascenders, and only ftz run along it, such that you can get away with raising it quite a bit without anyone noticing.

    To me, much of type design is exactly that. How much can I push this thing, or that thing, to create more space/character/balance without anyone noticing?
  • John Hudson
    John Hudson Posts: 3,451
    When you think about it, it's actually a little odd that we increase the x-height but not the cap height in bold weights. 
    I don’t think so.

    a) Most text consists of lowercase letters.

    b) The x-height area carries a large amount of information used in reading, including some relatively complex shapes and counters that need to remain legible as they gain weight.

    c) The taller cap height area contains relatively simple shapes.
  • Dave Crossland
    Dave Crossland Posts: 1,486
    michele casanova said:
    Thomas Phinney said:
    Robert Slimbach made the caps smaller in the bolder weights instead of increasing their x-height.
    As an example, I'd like to point out the "Titillium Web" font. In a recent redesign (as "Titillium Sans Pro"), the proportions were changed.

    Is that redesign the one at  https://github.com/chialab/titillium_pro_zanichelli ?
  • Dave Crossland
    Dave Crossland Posts: 1,486
    John Hudson said:
    When you think about it, it's actually a little odd that we increase the x-height but not the cap height in bold weights. 
    I don’t think so.
    I agree; I also think this messes with the font-size setting, since fixing cap height allows all other proportions to float in relation to that as an anchor, but if you just scale the upper and lowercase along an axis, you are effectively picking apart the internal consistency of the font-size progression. Not sure I am explaining this well :smile:

  • michele casanova
    michele casanova Posts: 54
    edited July 30
    Dave Crossland said:
    Is that redesign the one at  https://github.com/chialab/titillium_pro_zanichelli ?
    I used this version: https://github.com/chialab/titillium_pro but they are probably very related (I opened some issues in this repository)

  • Jasper de Waard
    Jasper de Waard Posts: 647
    John Hudson said:
    When you think about it, it's actually a little odd that we increase the x-height but not the cap height in bold weights. 
    I don’t think so.

    a) Most text consists of lowercase letters.

    b) The x-height area carries a large amount of information used in reading, including some relatively complex shapes and counters that need to remain legible as they gain weight.

    c) The taller cap height area contains relatively simple shapes.

    Sure, all true, but I think of increasing the x-height primarily as a way to create a visually equal size, not a way to fit more stuff in. Bold shapes tend to appear slightly smaller because their counter space is decreased, and this applies to lowercase as well as uppercase.

    This is Avenir Next. The bold o's look like they have the same size as the regular o's (but they are bigger), while the bold O's look smaller than the regular O's (they are the same). 
  • John Hudson
    John Hudson Posts: 3,451
    The bold o's look like they have the same size as the regular o's (but they are bigger), while the bold O's look smaller than the regular O's (they are the same).
    I understand the theory, but I actually don’t see them that way. The bold lowercase o is obviously slightly larger, and the regular and bold uppercase O align. But I am oddly attuned. :) 

    I would include counter size in the ‘more stuff’ to be fitted into the x-height area, especially in a and e.

Categories