Guides for designing Cyrillic letters Њ, Љ, Ћ, Ђ, Џ

Igor Petrovic

For those who are interested, here is the article I wrote on how to design Cyrillic letters Њ, Љ, Ћ, Ђ, and Џ.

https://nostalgicdolphin.com/blog-post-1/

This is PART 1 which deals with Serbian Cyrillic upright. PART 2 is about true italics and it is on this link:

https://nostalgicdolphin.com/blog-post-3/

----------

NOTE: This is the updated link (July 1, 2024) because I moved the article from Medium to my blog.

UPDATE: September 25, 2024 — Added Џ (Dzhe) letter

----------

Initially, I took these notes about three years ago when I did research on this topic—in order to respond to the request of my friend who asked me to review the Cyrillic set of his font. I complemented the findings during the design of my font Naslof last year.  

Cheers!

Andreas Stötzner

Thank you Igor, a very profound article!

Vasil Stanev

Andreas Stötzner said:

Thank you Igor, a very profound article!

Igor Petrovic

@Andreas Stötzner @Vasil Stanev Many thanks for your comments, I appreciate it much!

Nick Shinn

Bookmarked!

Christian Thalmann

Thanks Igor!

Does this work?

James Puckett

Thanks, Igor. Don’t stop here, write an eBook about the entire alphabet!

Rafael Saraiva

@Christian Thalmann I'd double check the instroke of the "El". 


Also, in the heavier weight, bowls of Њ, Љ feel a bit funky / disruptive, but maybe Igor could confirm that as native speaker.

Christian Thalmann

Rafael: My left leg is thin at the top and heavy at the bottom, so it does follow that recommendation. A blunt cut like in your drawing would be out of character in my typeface. 

I agree there’s something inconsistent about the bowls; will take another look. 

Rafael Saraiva

the doodle was just a casual illustration, not meant as reference for your design

to my eyes, it does not feel native... you could keep the upward curve, but the stroke modulation does not feel right. Have you consulted with locl type designers?

 

Christian Thalmann

Rafael Saraiva said:

Have you consulted with locl type designers?

Yes. I have native readers among the regular contributors to my development thread (https://typedrawers.com/discussion/1631/eau-de-garamond-a-sans-distilled-from-the-essence-of-garamond/p4). None of them took issue with /El-cy/. I also use the same tail in /Ia-cy/. 

Igor Petrovic

@Nick Shinn @James Puckett

Thanks for the support and encouragement! I am glad to hear there is interest to elaborate Cyrillic type design further, and now I see there is a demand for such a book, thanks for the suggestion. It's actually quite possible. I usually beat my fear of the new by taking notes, so beyond a ton of other type design notes, I have covered all Cyrillic caps already. However, they are not refined, so will require some time to sort everything out and stylize it in common sense language, as well as add notes for lowercase.

@Christian Thalmann @Rafael Saraiva

You are welcome! And thanks for sending the preview. I would say that from the perspective of Cyrillic design—and the article I wrote—this is pretty much ok. Here are a few details:

- I think that the left portion of the top bar on Ђ and Ћ could be shorter. In thin version especially, and in regular to some extent, while black is fine. Basically, the asymmetrical position should be obvious. In addition, shortening the left part of the bar makes spacing easier. 

- I would say Ђ in thin weight could be a bit wider.

- Bowl of the Њ in the black version opens the question. While it doesn't look consistent with the Њ bowls in regular and thing weights, this bowl shape is possible in the case where Љ, Ь or other letters in the set (Latin lowercase b i.e.) have it shaped like this. This somewhat compromises the idea of one smooth bar-to-bowl stroke in Њ design, but it is still ok, as long as the bar and bowl joins appear at the same height—which is most important. So, in that case, it's up to you to decide would you like to have the bowl shape consistent with the rest of the typeface, or would you like to have smooth Њ.

- There is no special preference for how the terminal of Ђ descending tail should look n terms of width and cut angle. I don't know how the rest of your typeface looks, mentioning this just in case.

At the end, a note induced by the El-Cy tail question. In the article, I assumed that it is already designed, so I just commented on it in the sense of composing Љ from it. As you probably already know, on this link there is a closer look at tail shape/issues:

https://leksandra.livejournal.com/115861.html

I agree with all of the shown. There is a suggestion about the thickening of the tail, but I would say it is more related to high contrast typefaces, and because your typeface is generally low contrast, there is no need for sudden thickening of the Љ tail in the black version. To demystify: When it comes to Cyrillic design, the general skeleton shape + general typographic rules would be just fine most of the time Especially in modern/low contrast/sans typefaces. On the other side, there is a point to check thin-thick relation in high contrast typefaces, as well as terminal/serif shapes, since that kind of typefaces shows stronger "expectations".

Christian Thalmann

Thanks Igor!

Updated version: Wider and more asymmetric Dje and Tshe, flat bowl in Nje. Is that better?

Comparison of /El-cy/ with related glyphs.

Ray Larabie

Your tip for optional forms of Ћ & Ђ was especially helpful. Deciding where to bend the rules for technical, progressive and ultramodern themes is one of the hardest parts of my job.

John Hudson

There is a suggestion about the thickening of the tail, but I would say it is more related to high contrast typefaces, and because your typeface is generally low contrast, there is no need for sudden thickening of the Љ tail in the black version.

I have a slightly different take on this. Contrast is contrast, and it doesn't really matter whether the contrast is high or low: what matters is where the thick and thin fall relative to each other. There can always be room for experimentation, innovation, and freedom, oc fourse, but these things need careful craft as much as convention, tradition, and restraint do.

I do find Christian's Л a bit strange because so much weight is being added in the descending stroke, which then tapers out towards the upturned tip. The sag on the underside of these strokes is good (a lot of people make the turn into the terminal too sudden), but there's so much weight above it, and so little in the terminal. The logic of where the thin and thick fall doesn't change because the design is low contrast.

Be cautious of comparing Л and Я as 'related glyphs'. They can be related, but when that happens — more often in the lowercase and most often in italics — it is because the Я takes the contrast pattern of the Л, not the other way around.

Christian Thalmann

My El-cy and Ia-cy do bend pen rules, but that‘s hardly avoidable with Cyrillic (at least not without going full-on calligraphic), given that it’s designed for vertical stress. 

As for character architecture rules, I do have a light stem with a heavy foot, it’s just a very gradual transition between the two. Don’t mistake the offstroke for the foot; it’s just the toes. 

Adam Jagosz

@Christian Thalmann I think Nje and Lje should have visually compatible bowls.

Christian Thalmann

So should I make both flat or both organic? I guess consistency with Softsign is not necessary?

Rafael Saraiva

Hope you don't get this personal, Ysabeau is a lovely design. But to be honest, some Cyrillic letters do not look as good the Latin at the moment. I find that's tricky to trust solely on native readers feedback because they don't have the background to judge. If they can read it, they will say it works well. So I find more efficient to involve type designers or experienced typographers. 

There are other letters that could inform the features of Л and л. At the moment the terminal looks like the flipped Latin LC t, and alien to other key letters. 

Also, I find that the right diagonal of Ucyr too heavy compared the Latin V and Y. And its tail could (not should) be informed by LC Latin y.

Igor Petrovic

@Ray Larabie

Thanks Ray, I am really glad to hear that. If I manage to cover the whole alphabet, I will have that on the mind, paying special attention on places where such optional forms are possible.

@Christian Thalmann @Adam Jagosz

I would say this is good I still feel that the tail of the black Љ is too thick on the turn. It's up to you to decide do you want to have it like this because of the general typographic style of the typeface, but there is no need to make it so thick because of the request of Cyrillic design.

As for the bowl of Њ, on the second look I agree with Adam. Љ and Њ should have consistent bowl shapes, and I will amend the article with this note. In addition it should be consistent with the Ь bowl, as well as other similar bowls, like lowercase b for example. I don't know how Ь or b bowls look in your typeface, but looking at the shape of R leg, I see that it has a humanist touch so maybe it is better make Њ bowl similar to Љ. Just make sure that joins on Њ appear to be at the same height.  

@John Hudson @Christian Thalmann

I agree with you John that generally contrast is contrast. But there are a few examples where the logic of thin/thick distribution is different in classic high-con than in classic low-con typefaces. For example, capital N usually have thick diagonal and thin stems, while classic low contrast usually has stems thicker than diagonal. Z sometimes also has thin diagonal and thick horizontals in hi-con, that's why I made that distinction.

The other thing is that hi-contrast typefaces have that "situation" at the ending of a thin stroke. It's much easier when it's a serif typeface, because serif doesn't leave thin unnatural ending. But with hi-con sans typeface, terminal usually widens to play "pseudo-serif" role. I think that Leskandra in her article covered this case when she said that left stroke of the Л should have thickening at the end, and that @Christian made his Љ tail too thick following that advice.

What I wanted to say is that I think there is no point at all to make Л ending thicker if the font is low contrast (except the typeface concept/design suggests that). My stance is that in that case weight distribution should be default (vertical>diagonal>hotizontal), and that's how I made my Л i.e. (I am not trying to say my typeface is perfect, just use it to depict my stance).

I basically agree with all Leksandra wrote, and situations are very wisely captioned. But it would be good to expand it with comments for special context that examples appear in. 

For example, I think that the first "Hockey Club" shape could be fine sometimes if crafted carefully, and with the consideration that it could be too close to letter П. If these things are taken into account, it could work, especially for geometry fonts (though I still prefer classic sloping variant). That shape is natural in handwriting, and if you would ask people in Serbia to write the letter in the sand (skeleton) most of them would probably draw a shape like that. Maybe the perception of the default is different in Russia, someone could comment.  

But this "Hockey Club" shape influenced my design, because I like to have left stroke of the Л starting with the straight segment.

Christian Thalmann

Rafael Saraiva said:

I find that's tricky to trust solely on native readers feedback because they don't have the background to judge. If they can read it, they will say it works well. So I find more efficient to involve type designers or experienced typographers.

This sounds as if native Cyrillic readers and type experts were mutually exclusive sets. We have our fair share of Cyrillic type experts here on the board.

There are other letters that could inform the features of Л and л. At the moment the terminal looks like the flipped Latin LC t, and alien to other key letters. 

Also, I find that the right diagonal of Ucyr too heavy compared the Latin V and Y. And its tail could (not should) be informed by LC Latin y.

I agree that the /U-cy/ is unfinished. I think I reduced the contrast there because a thin left diagonal ending in a thick foot looked weird. I might try something broken (thick-thin-thick) as in the Latin /y/, but with a hooked foot.

I don't think the /El-cy/ looks alien to the font; I like it this way. As your sample shows, it resonates with my /Ia-cy/ and /U-cy/, which also have tapering leftward offstrokes. I don't know what the /nine/ and /j/ are doing there; their structure is very different from /El-cy/, at least in the Garamond style that I'm pursuing (even if they may all have ball terminals in a modern). I will try lightening the foot as per Igor's suggestion, though.

Looking at a large number of fonts on MyFonts, I find that there are almost no contrasted humanist sans typefaces that offer Cyrillic... and the most commonly used solution for non-contrasted sanses (hockey stick) won't translate well to my contrasted humanist approach.

BTW, we should move the discussion of Ysabeau to its proper thread; we're off-topic here.

Christian Thalmann

Since this is still partly a propos, though, here's my reworked version of the glyphs in question. I've made the foot of /El-cy/ lighter as Igor suggested, and toned down the size and bounce of the offstroke in general. Quite a bit better, I think!  

Rafael Saraiva

I don't know what the /nine/ and /j/ are doing there

The Latin design and its features inform a lot your decisions on the Cyrillic. As you might know the Latin /j is recycled into the Cyrillic letter je. At the moment when I see both the el and je side by side, I don't see harmony.

I like it this wayIt seems that you're attached to your ideas at the moment, which is fine by me I think I've made my point already, so won't bother you anymore. I sincerely wish good luck with the Cyrillic development!

Vasil Stanev

There are Cyrillic workshops organized from time to time where I'm at and I think also in Moscow and elsewhere, perhaps it's best to check them up.

Consult with type designers with a native language that uses Cyrillic would be the easiest IMO. But the matter is not how you construct this or that letter but to have the eye to recognize which part is getting heavier, which one clogged, which one too anemic and so on. For example, I was taught as a greenhorn that I should push the top left oval of the Я a bit to the right. But when you realize that this is done to compensate for other parts and for it to not appear falling to the left, and that with a different shape different treatment should be used, you train yourself to do any national script. I used the same principles for Latin and Cyrillic that I used for Cherokee and I would use for Lao if I made Lao. Doing this, you also remember that sometimes a lettershape has to not only comply with basic rules, but also technicalities like interpolating the font or matching the Latin or some other script.

In the end, you are making functional art - it should feel, work and look good. This or that part can be discussed ad infinitum.

Igor Petrovic

@Christian Thalmann Low turn on the left stroke of the Љ looks better now  

Igor Petrovic

 Just a quick note, I updated the first post in this discussion with the new link because I moved the article from Medium to my blog. 

Also, a heads up for @Thomas Phinney and @Justin Penner (and maybe someone else) who were so kind to list the article in their type design resources repositories. Thanks!

John Savard

Rafael Saraiva said:

I find that's tricky to trust solely on native readers feedback because they don't have the background to judge. If they can read it, they will say it works well. So I find more efficient to involve type designers or experienced typographers.

You are absolutely right that it takes an experienced typographer to provide advice on many aspects of a typeface.

But it is also true that there are cultural aspects to what is usual for typography in a particular language that usually only native speakers will have any acquaintance with.

So there are three workable alternatives, in my opinion.

• Rely on an experienced typographer who has made a study of typefaces designed by native speakers, or that were traditionally in use in the area;
• Make use of the expertise of experienced typographers who are also native speakers of the language in question; or
• Seek inputs both from experienced typographers and from native speakers.

And all three of these add up to the same point: it's not "either-or", it must be "both-and".

Also, I have found this page online:

https://nostalgicdolphin.com/blog-post-1/

which directly addresses the question being asked.

I also recall that in the context of discussing differences between Russian and Bulgarian italics, an excellent resource on Cyrillic typeface design was mentioned, but I haven't been able to turn it up with a search.

EDIT: Oops. I see this thread began with that link, not with a request for help with those letters.

Tural Alisoy

I use this link often. Thank you very much, Igor.

Igor Petrovic

Thanks @Tural Alisoy, glad to hear that. Now that I have a blog, I hope to find time to cover some other Cyrillic topics (hopefully soon)

@John Savard Would like to take a look at that Russian vs. Bulgarian italics resource if you happen to find it. Maybe @Stefan Peev wrote or referenced it, since he collected an extensive set of Cyrillic resources on his site ( localfonts.eu )

John Hudson

Would like to take a look at that Russian vs. Bulgarian italics resource if you happen to find it.

Perhaps Krista Radoeva’s articles?

Cyrillic Script Variations and the Importance of Localisation

Extended Cyrillic: Bulgarian

Igor Petrovic

Looks great. Thanks, John!