Need critique on sans serif Cyrillics

Eimantas PaškonisEimantas Paškonis Posts: 91
edited May 2013 in Type Design Critiques
I didn't put accents here, otherwise it covers whole Extended Cyrillic:
image

Bulgarian and Serbian/Macedonian variants:
image

PDF

Comments

  • Alexei VanyashinAlexei Vanyashin Posts: 5
    edited May 2013
    Hi,

    Here a some quick suggestions.

    — The /Э/ also needs a longer middle bar.
    — Joining in /Ю/ and /ю/ feels too narrow.

    I could give more feedback if you post the Latin and some samples in text?

  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    Very informative, thanks a lot! Here's the text samples:

    image
    image
    image

    Lowercase italics:
    image

    Latin:
    image

    Russian:
    image

    Ukrainian:
    image

    Macedonian:
    image

    Bulgarian:
    image

    Mongolian:
    image
  • The Te with middle hook should have a middle hook like ghe with middle hook, going below the baseline. Ҩ ҩ shouldn't go below the baseline. You're missing some modern Abkhaz letters even though you have some letters that are only used in Abkhaz.
  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    Of all the fonts with these glyphs I looked through, I haven't come up with your combination. Most of the time /Ghemiddlehook goes below baseline there. Less often it sits right on the baseline. So now I don't know what's right.

    Same with /Haabkhasian: some go below baseline, some are more horizontal. Which is official?

    Which letters are missing? Combined ones, like /Гь, /Дә ?

    Also I have no idea how to draw italic /pemiddlehook.
  • Sorry, I mistook the dje for a te with middle hook. But in general the middle hook goes below the baseline. Even though they are not officially used in Abkhaz anymore, some other languages still use some of these letters with middle hook.

    For Ҩ ҩ the preferred form is above the baseline, the Unicode chart was fixed at some point to reflect that. Fonts that have the terminal going below the baseline are based on older versions of Unicode.

    See http://en.wikipedia.org/wiki/Abkhaz_alphabet for the full modern alphabet if you choose to support it.
  • So all the missing glyphs are combines one. What would happen if I didn't generate them? Wouldn't users be able to to just write 2 separate letters?
  • So all the missing glyphs are combines one. What would happen if I didn't generate them? Wouldn't users be able to to just write 2 separate letters?
    Letters with descenders like Ӷ, Қ, Ԥ, Ҭ (which have replaced letters with middle hook in Abkhazian like Ҕ, etc.) cannot be written as 2 separate characters.
    Even if it were letters with detached diacritics like acute, you would still need the precomposed glyph in your font if there's a character for it in Unicode. Unfortunately applications don’t use the parts to display missing precomposed characters if they are available (if you ask for eacute but the font only has e and combining acute, it will not work).
  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    Oh, I have those, just didn't include here because of redundancy.

    Is such /Ha acceptable or should the top terminal be more hooked/horizontal? And can bottom terminal be straight and horizontal? Early version.
    image

    Also which /Ef should is the most common in Russian specifically?
    image
  • John HudsonJohn Hudson Posts: 2,955
    edited May 2013
    The left-most Ф is closest to the typical Russian form. I'd need to see this light weight letter in context to comment on its proportions. The one in you first image looks okay.

    In the Abkazian Ҩ, I would try to better balance the two internal white areas, which usually means making the right loop narrower and a bit shorter. The top left doesn't need to hook over completely, but should probably turn a bit more than you have it. The bottom right exit normally turns up slightly as you have it, although at small sizes it is going to look quite flat.

    Consider making your descending spikes longer and more prominent, especially in the lowercase. Yours are too tentative. This is a lesson I learned from Maxim Zhukov many years ago: start by making them obviously too long, and then shorten them until they don't look too long any more. That's usually the point at which they are just right.
  • John, should decent of the spikes match across weights (or between UC and LC) or should they get longer as the weight increases or proportions change?
  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    @JohnHudson
    By descending spike you mean that tail?
  • John HudsonJohn Hudson Posts: 2,955
    Spike as in Ц. There may be a better term, and of course its shape varies, but spike is how I think of it.

    Jackson, I've not varied the length of the spike much, but then I've not made any super heavy Cyrillic types, for which I can imagine a bit of a longer spike might help maintain proportional relationship with other weights. As a rule of thumb, I'd say that while the spike is usually shorter than the descenders proper, it would behave in a similar way relative to weight. So if your descenders are lengthened a bit for a particular weight and width, then your spike would be also.
  • So you're saying that they should go below baseline? That contradicts Denis here.
  • John HudsonJohn Hudson Posts: 2,955
    I'm not talking about the tail of the Abkhazian Ҩ; I'm talking about the standard spike as in letters such as Ц. Sorry if that wasn't clear.
  • John HudsonJohn Hudson Posts: 2,955
    For notes on the design of Abkhaz letters, see
    http://www.unicode.org/L2/L2008/08134-n3429-abkhaz.pdf
  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    Thanks John, that .pdf is very helpful. For me it's just not clear about /Ghemiddlehook U+0494. Do they want it to look identical to Ӷ, /Ghedescender U+04F6?
  • Suggestions on your italic judging from the text:

    в(U+0423) is too narrow

    д(U+0434) needs a longer ascender, compare to /c/.

    й(U+0439) the cyrillic breve is too narrow.

    л(U+043B) the slope in the left stem should start earlier(same comments as in roman),
    otherwise it is hard to disambiguate from /п/.

    я(U+044F) the bowl feels too small.

    э(U+044D), з(U+0437), є(U+0454) the middle bars should be longer

    The left-most /Еf/ is okay for a Greek Phi. The Russian-style Cyrillic Ef generally has the bar extending beyond the cap height and baseline, and a larger bowl(about the size of a rotated O).

    Rule of thumb for the descending hooks/spikes in цдщ: they should be at least half the size of descenders or longer.

  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    Thanks Alexei, I only don't understand the 2nd point. Italic /д goes to the ascender height and /c (tse?) doesn't have one.
  • Eimantas,

    Sorry to be confusing, what I mean is that the tail of /д/ should have a more closed aperture, just like the top curve in /c/. In other words the curve of the /д/ ascender should be prolonged. The current design looks abruptly cut.
  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    Should the top parts of /д and /л be the same width? As if made /д was made from /л, only with descenders added?

    So it means there's a disagreement between you and John, who says that first /Ef is the most common in Russian. Who's right?
  • John HudsonJohn Hudson Posts: 2,955
    For me it's just not clear about /Ghemiddlehook U+0494. Do they want it to look identical to Ӷ, /Ghedescender U+04F6?
    No. The comment refers to text encoding practice, i.e. that they prefer to use U+04F6 and no longer user U+0494.
    ____

    With regard to the Ф, I don't disagree with Alexei. As I wrote, it is difficult to judge out of context. The bowl size is larger than I would make it for Greek, but could be vertically larger for Russian if you also adopt the convention of allowing the stem to extend slightly beyond the cap height and baseline.
  • Ah, while the bowl is not at cap height?
  • Nick ShinnNick Shinn Posts: 2,131
    edited May 2013
    This is all very well, but is making the Cyrillic version conform to the bland norm in keeping with your typeface concept, as stated in the Latin?

    The path you are taking is a safe one for foreign designers, but it might be better to benchmark your design against contemporary Russian-designed types that have a bit of personality.

    What would happen if you reverse-engineered the Latin from the Cyrillic, according to your present strategy?
  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    "This is all very well, but is making the Cyrillic version conform to the bland norm in keeping with your typeface concept, as stated in the Latin?"
    • So where's the actual question?
    • Yup, but let's also make Cyrillic stylistically different while we're at it.
    • "My present strategy?" Which is what? Asking people questions about letters' proportions?

    All your post is one big off-topic with cute insults here and there. PASS.


  • George ThomasGeorge Thomas Posts: 631
    I don't think Nick was being insulting at all. His comments were appropriate.
  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    If he meant that instead of "crowdsourcing" I should look at native designers' fonts with unique stylistic traits, then it's still "off-topic'ky", because I'm not making a display font. I'm working on a bland sans-serif without any personality here.
  • Nick ShinnNick Shinn Posts: 2,131
    It wasn't an insult.
    In fact, there was a compliment in there — that your original Latin design is more interesting than the Cyrillic.

    Your present strategy is following “what is most common”, is it not?
    That is at odds with your (Latin) design, which is not a clone of Gotham, Frutiger, or Helvetica, but an original design.
    Certainly, some of that originality is carried over into the Cyrillic via common glyphs, but—

    My question is, rephrased, could you make the Cyrillic have some interesting qualities that compare with original features of the Latin?

    If not an exact copy of things like the bottom left of the /b or the tail of the /l, then what?
  • Eimantas PaškonisEimantas Paškonis Posts: 91
    edited May 2013
    If that's so, then allow me to apologize for misunderstanding.

    I know the alphabet, I can read Cyrillic, I know some Russian, I've seen / can access to plenty of it, but I'm not that experienced to spot wrong proportions or incorrect details. That's why I'm here.

    As for the question: I don't know what's more there left to translate. I just have stylistic variants for both uppercase and lowercase /El, /De and 3 variants for the uppercase /Ef. Cyrillic doesn't have that much extra room for imagination (with its straight lines and right angles) anyway.
Sign In or Register to comment.