Glyphs thickness

mauro sacchetto
mauro sacchetto Posts: 371
edited September 16 in Font Technology
While fixing an old font (a Simoncini), I'm having a lot of doubts about how to establish a harmonious proportion in the small caps between the lateral thickness of the curved glyphs (|O|, but also |D| or |P|) and that of the oblique strokes with different angles (|N|, |V|, |W|).
Which ones do you find too discontinuous in this test?
Already here, the thick right diagonal stroke of |M| appears much thinner than that of |N|.

Is the only final criterion the eye (because, moreover, when printing a .pdf, the appearance changes a bit depending on the zoom...)?
Thanks
m

Answers

  • Yves Michel
    Yves Michel Posts: 247
    You should read Karen Cheng's "Designing Type". This 'invaluable book' as the editor qualifies it (and he's absolutely right) will answer your questions and many others.
    And, as you mention, don't base your design only on calculations but trust your eyes.
  • mauro sacchetto
    mauro sacchetto Posts: 371
    Thanks for the tip; I’ll be sure to get hold of that text.
    In the meantime, what are the most obvious flaws that your expert eyes spot?
  • Nick Shinn
    Nick Shinn Posts: 2,398
    As this is small caps, I suggest you examine it at small cap size, in text settings combined with capitals, lower case, and numerals. Especially italic lower case.

    In other words, you won’t know how well the automobile will perform unless you take it out of the garage.
  • jeremy tribby
    jeremy tribby Posts: 296
    edited September 17
    Is the only final criterion the eye
    yes. true of everything in type design :smile:  but there are principles to explain this.. when you slant a vertical stroke, its horizontal width stays the same, but the shortest distance between its edges gets smaller, so you need to compensate for that. as yves suggests, going mechanically by the cosine of the slant angle won't always look right, though - you need to use your eye and consider the broader context. the diagonal of the N stroke might become a bit heavier because it's uncrowded in a field of white, for example, while W might be lighter due to its density. also agree with nick! since small caps can set a little wide, you might make different choices, it will be helpful to look at a few pages set the way you imagine the letters being used
  • Craig Eliason
    Craig Eliason Posts: 1,525
    To my eyes, second stroke of /M, first stroke of /W, thick stroke of /X, and tail of /R feel a little thin. Big rounds /C/G/O/Q might get a touch too dark.
  • Yves Michel
    Yves Michel Posts: 247
    @mauro sacchetto
    Another good reference for your library: "Type Tricks" by Sophie Beier. (www.bispublishers.com)
  • Craig Eliason
    Craig Eliason Posts: 1,525
    edited September 17
    For the printing inconsistency issue, you could consider printing the font at a larger size (so rasterization is less apt to cause distortion) and viewing it from farther away. (Nick Sherman's Size Calculator can even calculate how to reproduce a conventional text-font perceived size at different distances!)
  • mauro sacchetto
    mauro sacchetto Posts: 371
    Thanks for the bibliographic references.
    Cheng's text can be downloaded from the Internet, but it is the first edition.
    I have ordered the more comprehensive second edition and Beier's text.
    This is a second draft with some corrections:

  • mauro sacchetto
    mauro sacchetto Posts: 371
    They are 95pt at the widestpoint vs 78 of the stems.
    Thanks for the opinion, now I do some retouching and then I also check optically

  • Kent Lew
    Kent Lew Posts: 1,055
    The bowl of /D looks fairly balanced with the stem. I would aim to get /O, /C, /G, /Q looking similarly weighted. (Which might or might not be a matter of the same actual measure.)
  • mauro sacchetto
    mauro sacchetto Posts: 371
    The thickness is the same for all the large curves; therefore, if some glyphs appear darker, it is not a matter of actual measurement, but rather the resulting optical effect.
  • John Hudson
    John Hudson Posts: 3,807
    edited September 23
    I usually start by making the thickest part of my bowls about 110% of the thickness of my straight stems.
    In this example, the straight stem is 220 units (on a 2000 UPM) and the thickest part of the bowl is approx. 242 units.

  • Craig Eliason
    Craig Eliason Posts: 1,525
    John Hudson said:
    I usually start by making the thickest part of my bowls about 110% of the thickness of my straight stems.
    In this example, the straight stem is 202 units (on a 2000 UPM) and the thickest part of the bowl is approx. 242 units.


    Isn't that closer to 120%?
  • John Hudson
    John Hudson Posts: 3,807
    Mistyped stem thickness, which is actually 220, not 202. Fixed in the edit.
  • Kent Lew
    Kent Lew Posts: 1,055
    mauro sacchetto said:
    The thickness is the same for all the large curves; therefore, if some glyphs appear darker, it is not a matter of actual measurement, but rather the resulting optical effect.
    The nominal thickness might be the same at the widest point. But that doesn’t mean that the weight is equal in all. It may be a matter of how the two bounding curves interact. 
    In the /D the transition from horizontal to vertical seems about right, and thus the weight feels balanced. With the other rounds mentioned, it seems like more weight accumulating in the “corners”, leading to an overall heavier feeling. 
    So either adjust the interaction of curves through adjusting control points; or if you want such broader curves in these glyphs, then adjust the overall thickness from the heaviest point to reduce the overall feeling of weight relative to stems.

  • mauro sacchetto
    mauro sacchetto Posts: 371
    Thanks for the very useful observations. I’m working on it.

    PS
    Today I received Cheng’s book; to me—a "font designer" in the loosest sense, not a professional and quite inexperienced—it seems really useful. Beier’s book is on its way.
    If you have any other titles like these ones, I’d appreciate it if you could let me know.
  • John Hudson
    John Hudson Posts: 3,807
    edited September 24
    How to create typefaces: from sketch to screen, by Cristóbal Henestrosa, Laura Meseguer and José Scaglione, is a very good book.
  • Yves Michel
    Yves Michel Posts: 247
    Hi John,
    "How to create typefaces" seems unavailable. I searched Amazon FR, US, UK and Googled the name.
    Do you know a reliable source?
  • Erik
    Erik Posts: 33
    Yves Michel said:
    Hi John,
    "How to create typefaces" seems unavailable.
    Disclaimer:  Iʼm not John, nor do I play him on TV.  —⁠But Iʼm curious as to how you concluded the book is unavailable.  Did you try the Item.zero shop (whose link is buried in the page John linked to)?
  • Yves Michel
    Yves Michel Posts: 247
    @Erik
    Buried is the term. So, I didn't see it. Thanks for the help.
    And I'll watch you on TV! 
  • Erik
    Erik Posts: 33
    📺👍
  • Thomas Phinney
    Thomas Phinney Posts: 3,192
    How to Create Typefaces is just back in print with a second edition, after a long break. I bought five copies to save on shipping to the USA, because I can sell or gift them, and the original was already my favorite type design book.
  • mauro sacchetto
    mauro sacchetto Posts: 371
    edited September 26
    From what I’ve seen, the second edition of the book is available only in the English version, not in the original Spanish.

    I am posting a further, more extensive draft, incorporating the initial corrections:

Categories