Freitag — toying around with a geometric display sans

13567…10

Comments

  • Alexander Stetsiuk
    Alexander Stetsiuk Posts: 36
    I have no Cyrillic expertise, but all the /Ш/ш/s look wide to me.
    You right
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited January 2016
    OK, thanks, I'll work more on б and У and narrow down the Шш.

    For б, I was going for something like FF Tisa and FF Profile:

    https://www.fontfont.com/news/challenges-of-designing-non-latin-typefaces

    The doubly-bent form of б goes against the geometric philosophy of my typeface, but I suppose I have to make sacrifices for legibility. How about this, though? I've now reduced the height of the ascender with respect to the Latin lowercase; I guess I don't need them to be much larger than the caps, given that there are almost no ascenders in Cyrillic.



    BTW, can I get away with a Roman-style д in my Unicase-style cut, or would it have to be an Italic-style one (similar to ð or g) in order to fit the other letters?

  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    I've changed г and д to be cursive-style in the unicase cut after all.



  • Alexander Stetsiuk
    Alexander Stetsiuk Posts: 36
    How about this, though?

    Yes, that looks so much better.

    BTW, can I get away with a Roman-style д in my Unicase-style cut, or would it have to be an Italic-style one (similar to ð or g) in order to fit the other letters?

    I think this is fine, but that does not mean I'm right. However, if you go this way, I advise you to keep both versions.


    Bulgarian Cyrillic


  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    Thanks Alexander!

    I was going to add a Bulgarian mode anyway — it's distinct from the unicase/roman distinction of my SS02. I've now implemented Bulgarian as the SS01, which can be combined with SS02 for «grown-up» capitals:


    The /Л/л is perhaps still not Bulgarian enough in the Bulgarian mode, though? I'm tempted to leave away the tail and keep the round dome. It would feel wrong to me to use a triangular apex when even the /A does not have one.
  • Alexander Stetsiuk
    Alexander Stetsiuk Posts: 36
    It looks good.  Of course, you are right: the El are not necessarily triangular in shape.
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    Alexander, the strongly curled delta-like /д felt out of place to me in the context of flowing text, especially in the cursive. Is a less strongly curled form such as this still acceptable?





  • Alexander Stetsiuk
    Alexander Stetsiuk Posts: 36
    It is perfectly, Christian! 
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited January 2016
    Good to know! :)

    So... how about Greek? Is this too latinized?

    EDIT: Forgot the SS02 caps:

    Does the tonos have to be before the caps rather than on them? Glyphs puts it on top, but that might just be because I don't have a /tonos.case yet.
  • Craig Eliason
    Craig Eliason Posts: 1,436
    Have you seen precedent for rounded /Δ/Λ/? I suspect those won't read.
    Generally yes, I think there's too much latinization going on here. 
    As starting points I think your /Υ/γ/ need to be structured not as a U-shape with a "handle", but rather as a stroke that loops through the "handle". Similarly, the alpha should have more of a continuous looping character rather than simple shapes stuck together. (I recognize the design premise of this typeface is simplified and geometric of course, so that's rightly steering some of these decisions.)
    /δ/ needs to end up heading the other way up top!
    Yes, I think cap tonos goes before the letter.

    Are there existing unicase Greek fonts you've seen? 
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited January 2016
    Hi Craig, thanks for the insight! I did intend to allow myself some degree of abstraction (especially in the unicase cut, which I think I'm going to rename the Titling cut) to make the geometric concept of the typeface work, but I suppose abstraction in the direction of Latin geometric tradition might be particularly problematic. I do have triangular versions of /Δ/Λ/ in SS02, which will become the default cut of the typeface.

    I've seen /a-shaped /alpha in professional typefaces before; I think it's a reasonable geometrification of the cursive form.

    I did draw a continuous-stroke /γ/, but didn't like the look of it. I suppose I'll have to go back to it anyway. Maybe I can keep the more geometric current version for SS01, though?

  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited January 2016
    Could this be an acceptable /xi...?  I guess not...?

    How about this one, though? Does it absolutely need a crossbar at the top to match /zeta?


  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited January 2016
    First full set...  I rather dislike the /Sigma.ss02 in the second line — can I get away with using the lunate /Sigma everywhere...?




  • Craig Eliason
    Craig Eliason Posts: 1,436
    /rho shouldn't have a stem on top. Maybe /Rho too if you're going for a unicase look?
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited January 2016
    Is that even true for non-cursive fonts? I do have a spurless alt for SS01 (see last line); I suppose I could promote it to default if necessary. I think it looks kind of goofy, though.

    EDIT: Of course, that was entirely because I drew the /rho.ss01 goofily. I like this version better. I've promoted it to default now.




  • Simon Cozens
    Simon Cozens Posts: 741
    It took me several seconds before I recognised those lambdas as lambdas - have you seen any precedent for that? I've only ever seen the second stroke go south-west. The only variant I've seen is the cursive lambda which is a bit like a connected /idotless or /l.


  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    I have seen precedent for such a lambda (e.g., PF DIN Display Pro, designed by a Greek), but I suspect it's not considered good form...

    Anyway, Gerry Leonidas has been tweeting in no uncertain terms that my Greek is lacking in many ways, so I've been reworking it a bit. Still work in progress, of course, but I hope this is going in the right direction.



  • Dan Reynolds
    Dan Reynolds Posts: 176
    Well, for what it is worth, I think that Gerry IS right over on Twitter. What is the rush, anyway? Why not just release Freitag as a Latin-script family first? Then, at a later date, you can release an new, or updated version that includes Cyrillic and Greek.
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited January 2016
    Yeah, I'll probably delay the Greek and publish the rest (unless the Cyrillic ends up having serious problems, of which I haven't heard so far).

    I wish Gerry would give some indication of what features are weird and which are wrong, though. His «homework» is practicing handwriting and looking at a lot of other people's stuff. The first will presumably just lead to me founding my own heretical chirographic tradition, and the latter might just wash away what little advantage I have as an idiot savant in type design; namely, not being preconditioned by mainstream type design.

    I'd rather be given the chance to understand the issue intellectually rather than having to guess at it through a long and aimless walk in the dark. Some people want to learn languages like that, through endless trial and error and such. Me, I just want a gods-damned reference grammar to churn through. Case in point: I drew my first broad-nib font in Glyphs before ever having held an actual broad-nib pen. Now that I have, I'm still rather satisfied with how it turned out.
  • Alexander Stetsiuk
    Alexander Stetsiuk Posts: 36
    I drew my first broad-nib font in Glyphs before ever having held an actual broad-nib pen.

    I thought I was the only one.
  • Dan Reynolds
    Dan Reynolds Posts: 176
    Well, as far as I am aware, there is not a reference grammar for Greek type design that is available for you to churn through. Nor is there one for Cyrillic, at least one that is in a language other than Russian. But there is middle ground between endless trial and error on one’s own, and a reference grammar. For instance, in July, the great ICTVC conference will take place again in Thessaloniki. There may likely be a few workshops that one could register for, to learn more about Greek type design – both by doing and by exchanging ideas with designers who are expert with the Greek script. At the very least, if you attend a conference like that, you can show your print outs to designers who are expert on the Greek script. That kind of in-person exchange just isn’t going to happen on Twitter, and it is honestly very unlikely to happen on a forum like this. Your best best is to seek an opportunity like that out, in my opinion.

    There are also occasional Cyrillic type design workshops at conferences like the ATypI, etc. And there are type conferences in Russia, too, that one could attend. Is that not an acceptable middle ground between working in the dark without much external input, and having a manual that spells out the rules in detail?
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    There might not be a reference grammar on Cyrillic, but there is a nice concise primer pointing out common misconceptions, which is helping me a lot:
    http://leksandra.livejournal.com/115861.html

    Similarly, there's an extremely useful page explaining how to Polish Accents: http://www.twardoch.com/download/polishhowto/kreska.html

    ...and several pages on how to get the Icelandic /Þ/þ/Ð/ð right. So far I haven't found an equivalent for Greek. Gerry's PDF primer on Greek typeface design contains a lot of font samples, but is often unclear about whether they are good or bad examples, and about what design aspects are good or bad about a particular font. According to the text, many if not most commercial fonts have bad Greek, so I find it hard to figure out the rules.

    I certainly agree that Twitter is a bad medium for such things. A conference or workshop would be delightful, and on my wishlist. I have two day jobs who take preference, though.
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited January 2016
    Out of interest, to what degree does this work? I have no way of telling.  :grimace: 

  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    edited February 2016
    Alright, I've started working on the promo materials. At this point, I think I will release Latin, Cyrillic, Hebrew, and the current preliminary Greek. The latter will essentially be considered a public beta, and I will supply a more sophisticated version when/if I make it, as a free update on MyFonts. My impression is that lack of explicit feedback is what's currently holding me back on that front, so I'd like to solicit that by having people play with it. I'm even considering handing out a Greek-only demo version for free for that purpose.

    Meanwhile, here are some pictures. I am definitely going with the name Quinoa for the family.

    BTW, is the Dutch correct? Hulp! :grimace: 






  • Martin Silvertant
    Martin Silvertant Posts: 166
    BTW, is the Dutch correct? Hulp! :grimace:
    Very cool design. I love the unicase letters.

    As for the Dutch sample, the correct expression is 'Groetjes uit het leuke Katwijk aan Zee'. On the other hand, 'leuk(e)' sounds a bit forced in this context. What's the expression you want to make in English?
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    Thanks Ofir, that's fantastic feedback and exactly what I needed.  :blush: I do have a few questions of my own, though:

    Round vs fig-shaped Samech: I do have a stylistic set (SS01) that replaces many default shapes in the Latin with spurless, rounder versions. I could do the same thing with Hebrew, with the default glyphs being more traditional and SS01 supplying a more geometric interpretation (including an /O-shaped Samech). SS01 is going to be offered as a free-standing font, so it will be usable everywhere.

    For the traditional approach, I would still prefer a «sans-serif» design philosophy to match the Latin. I like Samech C. Would it be acceptable to remove the top left «serif» of Tav entirely...?

    For Ayin, wouldn't a descender on the non-final version get in the way of Nikkud? Or would you recommend a contextual alternate to be used when a Nikkud follows?

    For Pe, it looks to me like the forum font's rendition is even more «stingy» than mine... but if it's the first thing that leapt to your eye, I suppose I'll have to change it (although I really like the current appearance). I think I prefer your version B, and I'll use version A with the central dot. I find myself rather opposed to C...  :grimace:

    For Kof, I'll use your approach A. The alternative Shin is not much to my liking, but I suppose I could offer it as a character variant.

    Thanks again!  :smile: 
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    Martin: «het leuke» is what I had originally written, but apparently I overthought and malcorrected it. :grimace: It's only the indefinite article that triggers the form «leuk» with neuter nouns, right? Should have trusted my instincts... thanks for the correction!

    I was going for something like «nice» or «lovely». Should it rather be «mooie»?
  • Ofir Shavit
    Ofir Shavit Posts: 397
    edited February 2016

    My pleasure :)

    Indeed you can remove the Tav's serif entirely and kern it with those opened on the right side.

    I think that descending the Ayin was done in first place to reduce the huge gap on the right plus make it bigger because when it touches the baseline only on the left corner it seems too small and "on high heels". These are the considerations leading to determine how much to descend the Ayin and normally it shouldn't descend enough to conjunct with the Nikud (Hebrew punctuations), but if needed you can shift the Nikud to the right and it'll be better than alternating when Nikud is added because of the above mentioned problems.

    The forum's Pe is stingy too but it uses another possible structure in which the roof of the Pe is slanted (rotated actually), this is easing the density of the 3 horizontal strokes of the normal structure as well as making room for the Nikud's dot in the middle of the letter (I wouldn't alternate for the Nikud as well, I think that consistency is important), so it is pointing a bit upwards and maintaining a flat forehead  which is somewhat important. In case you like the "leaf shape" I recommend using B5 (and apply it to the final Pe (8) as well).



    :)

    Pey.png 34.1K
  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    Thanks Ofir!

    I do like the looks of B5, but it doesn't look very amenable to a central dot. It also feels a bit haphazard compared to the other letters, which are solidly grounded between the horizontal bounds of the cap height. I think I will settle on the «A» version in your first post; that should fit the geometric flavor of the typeface best and cause the least problems with a central dot.

    I'll try a descending Ayin with a bottom anchor point shifted to the right. How common it is to use nikud anyway these days? I guess Yiddish needs the ones under Alef, but those are unproblematic.

    If I use the type-C shape of Samech, do I have to adapt Kaf as well, or can it remain circular?

    Would a completely square shape be acceptable for final Mem in the geometric stylistic set, or is that a crime?  ;o)

    I assume it would be unacceptable to clip off the right-hand serif of Dalet?

Categories