Eau de Garamond — a sans distilled from the essence of Garamond



  • The bold /oslash has very small counters, and the top and bottom of the o-shape looks a bit too altered to compensate for the bar. It looks okay in the lighter weight.
    My suggestions:
    > thin the bar of /oslash a bit. The part inside of the counter can be a few points thinner than the outside if needed.
    > Extend the bar a tiny bit, and
    > bring in the top and bottom of the o-shape counter a little.
  • This is a 15-page thread of a designer listening to constructive criticism and making subtle changes to a design over years. The implication I get from your comment that he's blind or careless to subtleties of design really feels out of place to me. 
    Thanks for the support, Craig! But I suppose I've come to terms with that fact that Andreas' advice comes with a side of humiliation; so far the tradeoff has been worthwhile, I would say. :grimace:
  • Thanks, Sander! Does this work? I did the same to the Roman version.

  • Yes, I think they fit in more with the other letters, although the italic /oslash looks less rotund than /o. I would extend the sides of the oval.
  • I’m sorry if I sounded a bit harshly; I only wanted to give the advice which I think is indispensable for a project such ambitious as this one: to proactively watch and ‘scan’ visually everything all the time.
  • I thinned the horizontal bar as per your suggestion, and that significantly improves the letters. Thanks!
    I also reduced the h and balanced the ff just a bit, but it doesn’t look different to me. No amount of looking and scanning will show me these things if they’re below my perceptive threshold. 🤷
  • @Sander Pedersen, like this?

  • Looks better. The distribution of black and white makes it appear similar to the surrounding letters, even though the oval is thinner than that of the /o.

    Some real words with oø/øo combinations, handy for proofs: kommandoøkonomi naboøya portoøkning sjøområde køorden miljøorganisasjon.

  • I've now completed the character set in Black Italic and am hoping to still make it into the Q2 release on Google Fonts... does anyone know when the deadline is?

  • Last-minute change: I always disliked how the staggered f_f ligature plays with subsequent f's... Apparently EB Garamond solves this problem with equal-depth descenders everywhere. Ysabeau does too, now.

  • But that solves a problem that almost never occurs at the cost of making the ff ligature less elegant... Also, is it just me, or is that S way too bold compared to the lowercase?
  • I agree that the previous ligature is more elegant, but even outside of the /f_f/f/ conjunction it often struck me as disruptive, as if it were derailing from the guidelines. I expect the new version to be better for legibility.
    You're right about the /S/. That probably happened in my latest round of tune-ups after Andreas' feedback. Andreas insisted the spine needed more oomph, but I might have neglected to trim away some weight elsewhere to compensate. Does this look better?

  • what I still don’t understand is these differences:

  • … and italic O and S deserve some tuning:

  • Christian ThalmannChristian Thalmann Posts: 1,880
    edited April 2022
    The endings in /c/e/t/ are part of a load-bearing arc whereas those of /a/l/i/ are just off-flicks from an otherwise complete stem. I consider it entirely reasonable that the former should be heavier than the latter. I can try to make the former lighter to see whether that works, though.
    The /pi/ is related to /t/ but could probably bear to be coordinated with /alpha/. The top of /lambda/ is heavy by Irene's request, though. Greek doesn't care for a consistent pen angle.
  • … and italic O and S deserve some tuning:
    Oh, you're right, there are some unsightly red dots in there! How could I miss those...

  • Christian ThalmannChristian Thalmann Posts: 1,880
    edited April 2022
    I tweaked a few letters to bring their offstrokes closer to each other. The thin-toed /t/ (second /t/ in the sample) looks unbalanced to be compared to the previous version, so I'm not using it. The case is even clearer for /c/.
    I think I improved the smoothness of the /S/ (even if it strikes me as very Gill Sans-y with those cuts). I couldn't find anything wrong with /O/, so I left it unchanged.

  • It seems to me like all your caps are considerably bolder than the lowercase. That's fine of course, it's a decision you can choose to make. But for my personal taste, the difference is a bit too pronounced. 
  • You were very right about /S/ before, but I don't see such a general trend in the rest of the typeface.
  • Is the /fa pair a touch loose?

    Also, is the /f_f ligature a touch tight?
  • Last-minute change: I always disliked how the staggered f_f ligature plays with subsequent f's... Apparently EB Garamond solves this problem with equal-depth descenders everywhere. Ysabeau does too, now.

    What about a middle ground between the previous and new position?  It gets to keep a bit of the style, but it doesn't looks too staggered.
    Also, how do they look in text size?
  • kymhfhgdfkymhfhgdf Posts: 1
    I am not sure why I am thinking of Gill Sans while seeing this. Truely interesting!
  • JLTJLT Posts: 6
    I can't add much that others haven't already said, but I really do love this, and the strong Luc(as) DeGroot vibe I'm getting from it. Beautiful.
  • Christian ThalmannChristian Thalmann Posts: 1,880
    edited December 2022
    So apparently my 11 weights of Ysabeau can't be ingested into Google Fonts because the weight classes can only range from 100 to 900 in integer steps of 100. D'oh!
    I'm thinking of either redefining the instances to span the same range with fewer and larger intervals, or decoupling the lightest two weights into a separate typeface. The latter is certainly inelegant but allows me to keep the current instances.
    Typographer Michael H wrote a plea on GitHub to preserve the fine grain of Ysabeau's weight spectrum because it allows it to match the weight of tall serif fonts (like EB Garamond) such that they can be used in an intermixed fashion. I do wonder whether the variable font doesn't already cover role, and in a superior manner.
    Incidentally, I've always thought the current Bold was a bit too light for its own good, so I wouldn't mind the opportunity to readjust at least the naming a bit. Then again, Michael implies the current weights match the weights of EB Garamond, so maybe I'd break something useful in the process.
    I'm a bit stuck here. What would you do?
  • Well, the variable font allows for arbitrary weights! Anybody using the variable font can still get at any weight they like. The redefined weights would only impact the predefined instances—and probably the non-variable fonts derived from them.

    It is unfortunate that Google Fonts has some specific limitations/parameters like that, but to be expected. If there is nothing else that gives you trouble, you have gotten off fairly lightly!
  • If there is nothing else that gives you trouble, you have gotten off fairly lightly!
    Yanone sounded like this was only the first hurdle... I'm hoping it's the largest, though.
  • I've only been lurking during the development of this design, but I signed in to say that it looks great. Congratulations Christian!
Sign In or Register to comment.