My text face wish list

13

Comments

  • my critique was for the bold, right. It may have become better in the last sample, however, they still look a bit too martial to me. The upper and lower endings of [ ] and { } are too thick.

  • So should I add intermediate layers to keep the Regular at its previous level?
  • So should I add intermediate layers to keep the Regular at its previous level?
    I like the original regular, as of course the heavier brackets/braces are more distracting if drawn too heavy in the heavier weights.
  • With intermediate layers:

  • Jan, I find your points very interesting. I’m always interested in matters like you mention. Just for the records, Andron is intended (mainly) for scientific editing in the humanities in particular and, yes, it has italic small caps.
    I would be interested to give one or two of my typefaces a test run with a small piece of sample text, if you’d care to provide one.


  • I still regularly encounter text typefaces coming from renowned designers and foundries with improperly drawn diacritics in letters of the Polish alphabet.  Common offenders include generally-off-looking stroke in /ł and well-drawn, but improperly-centred kreskas.
    I’m curious about the above. I know that the kreska is usually steeper than the acute accent, but does it have differing placement rules as well?
  • Jan PietkiewiczJan Pietkiewicz Posts: 17
    edited May 2021
    Andreas, I'm well-aware of your great work on Andron and I will definitely consider it for my future projects (especially if I'm ever trusted with a book that could make generous use of its multi-script capabilities). I will try to come up with a sample text that could serve as a crash test for all my requirements, but this will take me some time. The samples you provided are of course very promising.

    I’m curious about the above. I know that the kreska is usually steeper than the acute accent, but does it have differing placement rules as well?
    I'm not sure what constitutes the mainstream opinion among the Polish type specialists in 2021, but personally I'm perfectly happy with kreskas just as steep as acute accents. What I do consider unacceptable, however, are kreskas that seem to precede or run away from their letters (i.e. ones that are positioned too far to the left or to the right from the optical center). It's very possible that the typefaces that (in my estimation) suffer from this issue are characterised by the same treatment of a-acute, e-acute etc., but it's the Polish letters I look at in the first instance. Later I will try to provide some examples.


    For an additional piece of trivia, I just recalled that an editor friend of mine recently showed me a fragment of a highly complex text where it would be necessary to distinguish roman slashes from italic ones. This runs counter to advice I encountered in many places, according to which italic slashes should not be made any more slanted than their roman counterparts.
  • I'm not sure what constitutes the mainstream opinion among the Polish type specialists in 2021, but personally I'm perfectly happy with kreskas just as steep as acute accents.

    Just for clarification, does “perfectly happy” mean you accept these equally, or that you prefer the more acute-like form?

    It’s my understanding (which admittedly might be wrong) that in Greek the steeper form of the tonos common in older fonts is now discouraged, and I’m wondering if the same sort of shift might be happening in Polish.
  • ok, I throw in my 2 pence. Please judge harshly:


  • Nick ShinnNick Shinn Posts: 2,131
    I tested my neighbour, a Polish author, on the design of /ł, with texts that featured either a vertically centred crossbar, or one centred on x-height—on a typeface with quite a small x-height. I didn’t tell her what the issue was, so it was “spot the difference”, which she couldn’t, and read both just fine.

    Admittedly, this is entirely anecdotal, but it did confirm my belief that such matters are of absolutely no significance to readers, only to type designers, and perhaps the occasional enlightened typographer!
  • Christian ThalmannChristian Thalmann Posts: 1,938
    edited May 2021
    ok, I throw in my 2 pence. Please judge harshly:
    Good idea!

  • Thanks for your insight, Jan! :grimace:
    I find the self-assured "true kreskas" in Cormorant almost intimidating

    If it's any consolation, they're not even different from Cormorant's regular acutes. It's part of Cormorant's Stainless Steel Victorian Murderspider aesthetics. :smirk:

  • Nick ShinnNick Shinn Posts: 2,131
    edited May 2021
    I suppose a comparison to the height of the /ł crossbar would be that of the f’s.
    I was surprised, when teaching type design to a class of university graphic design students, how many of them placed the f crossbar way down low, clearly wrong, not at x-height.

    But then again, Hermann Zapf did the same thing in Palatino, and Roger Excoffon in Antique Olive. In both these examples, it works for the typeface. Also, nobody dare question the Masters!

    So, rather than a set pattern applicable to all typefaces, I prefer a general principle: What I consider when designing /ł is the adjacent characters in the particular font I’m working on, especially when it’s followed by /u or /y, e.g. Batłtykiem, odwołuje. Again, it’s not an issue for readers (who have dealt with the keming of r_t in Helvetica for decades), but I do try to avoid unintentional ligatures!


  • Jan PietkiewiczJan Pietkiewicz Posts: 17
    edited May 2021
    Nick, thank you for your very welcome warning against the blind orthodoxy of universal patterns – it certainly made me reconsider the strength of some of my judgments as expressed in this thread.

    I now realize that while I was trying to make a very similar point with regard to kreskas (the angle as such doesn't matter and it's not necessary to follow any rule regulating it; what matters is how the kreskas hang together with everything else), apparently for some reason I'm eager to make the crossbar in /ł my hill to die on. I generally tend to assume that type designers know what they are doing, and for instance I think I understand what was the intended effect of the low crossbar in Berling Nova; still, because of the said crossbar I wasn't able to bring myself to use this particular typeface in a recent project of mine and my (over-trained) eye still stumbles on its /łs whenever I read a book typeset in it. Similarly, I simply had to ask for the custom replacement for the /ł shown above – not because I didn't find it pretty, but because I deeply felt it was wrong and that some imaginary opinionated reader could accuse me of using an uncalled-for non-conventional form of the letter in my work. Despite all such anxieties and prejudices on my part, I absolutely have to concede that in general there's obviously much more space for flexibility and experimentation with any letterform than I might have suggested before.


  • Hi there,
    This is gentle reminder to stay on the original topic. If an interesting tangent subject is being discussed, please open a new thread for it.
    TIA.
  • Ramiro EspinozaRamiro Espinoza Posts: 839
    edited May 2021
    >> I wouldn’t say this is a tangent, it’s more of a specific granular discussion
    Maybe, but a member taking part in this debate mailed the moderators asking for our intervention and this is why I posted the reminder.
  • Nick ShinnNick Shinn Posts: 2,131
    Fair enough, but I have three “Agrees” to your one!

    (However, I do tend to go off on tangents.)
  • Fair enough, but I have three “Agrees” to your one!

    (However, I do tend to go off on tangents.)
    I also think this derails the original discussion Nick, not because it’s not relevant but because it’s mostly on the design of diacritics rather than the inclusion or exclusion of relevant glyphs.
  • Great thread!
  • edited June 2021
    This caused a problem because these brackets in CJK have a large blank space at their left (3008) and right (3009) sides to fulfill the square area of CJK characters. This doesn't work in math and computing notation, when the excessive space in not needed or desired.
    Both 3008 and 3009 should not fill the square area in CJK context when kerning or proportional variants are enabled, or in Korean.
    If 3008 and 3009 have proportional variants in a font, this is usually accessible through the 'palt' Proportional Alternate Widths feature which should be activated when the 'kern' feature is enabled or sometimes the 'pwid' Proportional Widths feature, which should activate the 'kern' feature itself instead.
    In some Korean fonts, the proportional glyphs are the default, since white space in half of the square area would look like a space, which would break Korean grammar as some word compounds are attached to other bracketed word compounds.

    So in theory they should have the appropriate shape as canonically equivalents of 2329 and 232A in those contexts, along with non CJK context. In practice, unfortunately that's often not the case.






  • I just wanted to say that I returned to this very valuable thread, as I am currently doing a revision of Andron. A lot of requirements described here are already met, but some of the suggestions will make it into the new version, I’m pretty sure.
    If anyone has further ideas about what could get improved, now is the opportunity to speak about it and to eventually see them materialized in a typeface.
  • Igor Freiberger said [last year]:
    … The 27E8 and 27E9 are mathematical operators so they will most probably not match the text style.
    why should it be likely that “they will most probably not match the text style”?
    My idea of a typeface is that everything matches anything.
  • This is also my idea. I didn't say they should, but in most cases mathematical operators are draw in straight, linear style regardless the typeface style. Andron is an exception.
  • John SavardJohn Savard Posts: 1,088
    My idea of a typeface is that everything matches anything.

    So that explains the New Shekel symbol in your first attempt at Hebrew.
    Well, this isn't an idea to be rejected out of hand. Such symbols as $, &, and %, for example, are made to match the style of the letters and numerals in the typeface to which they belong, and so why wouldn't that go on to include all the sorts?
    In the past, only a small handful of typefaces were used for the setting of mathematics, and so people have come to assume that this is what mathematics has to look like.
  • I would like to make a pitch for the inclusion of the underdot H, in all cases (U+1E24, U+1E25). It is used widely in the transliteration of the Hebrew letter chet, and also in the transliteration of modern South Arabian languages and, natively, in Asturian.

    Another thing I’d like to see is upright (i.e., roman) brackets in italic fonts, if not as a replacement, then at least as an alternate in a stylistic set. I don’t believe that italic brackets should really exist outside of the relatively rare case of entire texts set in italics. 

  • Nick ShinnNick Shinn Posts: 2,131
    Italic parentheses are a typesetting issue, not a font issue.
    Consider these options for italic as a subsidiary within roman body text:

    1. I’d like to see upright (i.e., roman) brackets in italic fonts, if not as a replacement…
    2. I’d like to see upright (i.e., roman) brackets in italic fonts, if not as a replacement…
    3. I’d like to see upright (i.e., roman) brackets in italic fonts, if not as a replacement…

    Option (1) doesn’t require any extra typesetting, as all the typographer has to do is select the letters to italicize, the parentheses are already roman.
    Option (2) seems perverse, as text surrounding the parentheses is all italic—and if the parenthesized text begins or ends with f, the effect will be nasty, even if kerned.

    Therefore, slanted parentheses in italic fonts are the best practice.

    However, there are also type genre issues, which is why I included roman parentheses in my Scotch Modern italics (via a stylistic set). Upright Scotch parentheses, being a fine line (like the plus and multiply symbols), do look appropriate with italics, in this historical style.

Sign In or Register to comment.