Type Design Technique & Theory — TypeDrawers

Subscripts vertical alignment

pereelmagne — Fri, 04 Sep 2026 08:24:09 +0000

I see different implementations as for the vertical placement of subscripts. Noto Serif places the figure centered to the baseline, whereas IBM Plex Serif makes figures rest entirely on the baseline. See the image.

I haven't found any "authoritative" reference regarding the vertical alignment of subscripts. Does anyone know of any? Maybe in chemistry the IBM Plex solution would be preferable, but Noto Serif's would be the general implementation. I don't have a clue.

Circle in Font not completely round when font is used

Rae2J — Tue, 01 Sep 2026 19:46:25 +0000

Hello there,

i am completely new to Font design. I work in IT and in my free time i am designing and producing adapted products for blind and visually impaired people. My wife is blind, so i have some experience in that field.

There is a Truetype font for wrtiting in Braille. That means i type the text like here and the font creates the dots according to the Braille dots.
I am based in Switzerland, so my native tongue is german and the font i use is the german standard Braille Truetype font.

The problem i am facing now, is that if i write with that font, the dots in the font are not perfectly round. If i want to use them in a 3D Software i face some problems.

So i thought i am a smart guy, and downladed FontForge. installed it, opened the Braille font and saw that the circles where not really exact circles. So i made a test with the letter A and replaced the dot (circle) with a circle i made, holding the shift key while drawing it. So i assume it must be precisely round. I checked it with the position coordinates i saw in Fontforge and it seemed very round. I made a Screenshot and tested it within a Vector Program with an overlayed circle and its completely round.

Then i saved the font and installed it.

When i now use the font in the vector program, the dot (circle) is not a 100% round anymore.
When i overlay it with a perfect round circle i see that there are (very tiny) differences between the font and my circle.

The picture shows the fonts circle in yellow and my reference circle in the back in red.

I know this is practically no difference, but to my workflow it really IS a big difference.
In other words - the 3D software still gives me a hard time using the font and much time of work to correct this.

So with no clue where to start i searched the net and found your page here.
And now i am hoping that you guys know whats wrong and how to correct it.

I am not bound to FontForge. If there is a program that does the job, that would be good too.
I only rather fix the font, than to fix every dot manually in the 3D software.

Thank you in advance for your help and your time. I really apreciate it!

Greetings
Reto

Math: horizontal delimiters, vertical alignment?

Andreas Stötzner — Thu, 27 Aug 2026 12:19:12 +0000

My question is about the pairs of horizontal delimiters in the 2300 Unicode block. This overview is taken from the UTR#25 (technical report):

I wonder about the difference in the vertical positioning of the first pair 23B4-23B5, in comparison to the other pairs, where the ‘top’ sorts sit below the baseline but the ‘bottom’ sorts sit very high. If that is intentional, I couldn’t find an explanation anywhere, why this is so. It seems illogical to me but I may be missing an important detail. – I also noted that in the STIX Two Math font, differently to the Unicode materials, all top sorts sit on top and all bottom sorts sit at the bottom.

Has anyone ideas about this?

Seeking Examples of Typeface Design Documention

Daniel Yacob — Thu, 27 Aug 2026 11:17:31 +0000

Greetings,

I'm looking for examples of documentation of a typeface that would cover the initial objectives of the design, the intent of its application, style and shape influences, obstacles and challenges encountered, and the decisions made to overcome them, along with the tradeoffs considered.

Documentation like this would be more like a project report than a design brief, which in turn is more focused on eliciting requirements from a customer.

Please let me know if there are any classic examples of documentation like this that I should review. Thank you!

Bold assignment for weights other than Regular?

Christian Thalmann — Tue, 14 Jul 2026 16:36:04 +0000

One of my users said they used Ysabeau Medium as their go-to font in Word, but found it unfortunate that Word's bold function doesn't work on it. I've assigned the Bold weight as the bold version of Regular, of course, but I haven't made any such assignments for the other weights. Should I, or will that cause more trouble than it's worth?

Does Surrealism have a Typographic Language?

Arne Freytag — Tue, 07 Jul 2026 07:36:25 +0000

I have been reflecting on the connection between typographic styles and broader movements in art and architecture.

For many historical periods, there seems to be a fairly clear typographic counterpart: for example, Renaissance and humanist typefaces, Modernism and geometric sans-serif typefaces, and so on …

But what about Surrealism?

Are there any typefaces, type styles or typographic movements that could seriously be described as ‘surrealist’? If so, what characteristics would distinguish them? Are there historical examples from the Surrealist period itself, or later typeface designs that embody Surrealist ideas typographically?

Or does Surrealism differ fundamentally from movements such as Art Nouveau, Art Deco or the Bauhaus, in that it manifests itself primarily through visual language, composition, collage and visual context, rather than through the design of the letterforms themselves?

Would you consider Surrealism a missing category in typographic classification, or simply a movement that never developed its own recognizable typographic language?

On atavism and cultural authenticity

Nick Shinn — Tue, 11 Aug 2026 21:39:13 +0000

Over centuries, the Latin script became simplified to adapt to various constraints, notably those of typographic technology.
Ultimately, even italics, which best represent the Western identity in scribal form—with all the attendant complexities—dropped the fancy ligatures and flourishes, and became, in the 20th century sans, a sloped version of the roman.
But there is a different approach now, for other scripts, which is to have the technology adapt to the script qua script, i.e. covering both its meanings: that which is handwritten, and also a language encoding system:

• Simplified Arabic was rejected as being culturally inappropriate (type tech being Western no doubt played a part).

• The same phenomenon has occurred with Greek, in which a cursive ductus in the minuscules has been identified as being culturally significant; the cock-ups made by Western type designers, such as Eric Gill’s heavy-handed Latinization, may have had some influence.

• And again, users of the Inuktitut script like to see it looking very geometric-sans, as that is how the language was first rendered in a unique manner, rather than piggy-backing on the Latin alphabet.

Ironically, tech progress in the West has since the early 19th century resulted in a growing complexity of letter forms, eventually, with OpenType contextuality, being able to render the old scribal styles, as preserved in the joining-behaviour of formal calligraphy and vernacular handwriting. This effect was identified by Marshall McLuhan as the third law of his Tetrad, Retrieval:
What the new medium retrieves that was previously obsolesced.

Aksharaya calligraphy manuals

Simon Cozens — Wed, 24 Feb 2016 12:18:45 +0000

In the "shut up and take my money" department, Aksharaya have prepared six calligraphy manuals for the upcoming Typoday 2016 conference in Bangalore - introductory manuals for Devanagari, Nastaliq Urdu, Gujarati, Gumurkhi, Odia and Kannada.

If you email info@aksharaya.org, they'll courier copies to you after the conference.

Figure colon - Stylistic set or Contextual alternate

Ermin Međedović — Wed, 29 Jul 2026 15:49:28 +0000

Most of my fonts have a figure colon, which I find useful as an alternative to the divide sign, and it also makes time expressions like 16:30 look better.

Until now, I have made it available as a Stylistic Set, but I’m now considering putting it into calt so it would be enabled by default.

Any thoughts or advice on whether this is a good idea or something to avoid?

FYI: guidelines for Omani rial sign

Peter Constable — Sun, 26 Jul 2026 13:24:43 +0000

The new Omani rial sign will be encoded in Unicode 18.0, which will be published a couple of months from now. The Central Bank of Oman had published guidelines for its usage, and has just posted updated guidelines that allow for a narrow aspect ratio glyph in fonts, suitable for tabular layout:

https://cbo.gov.om/Style Library/CBO/downloads/The Omani Rial Sign Guidelines.pdf

diverging from conventions for Arabic joining

Peter Constable — Sat, 11 Jul 2026 20:54:19 +0000

This is a question for Arabic script experts:

How strong or intrinsic are the conventions for which letters of Arabic script join on the left or the right? Is it conceivable that a designer could create a design setting some Arabic text in which some letters don't follow the normal conventions — e.g., having a beh not join on the left or on the right, or having an alef or dal join on the left?

Is it plausible that, given enough time, conventions could evolve so that common practice for joining (perhaps only in non-liturgical / Quranic text) is not what it is today?

The Place of the Low-waist 'a' in Modern Latin-Based Fonts

Dsche — Wed, 08 Jul 2026 09:27:23 +0000

First of all, I want to apologize for any possible mistakes in my English, but I hope I can still formulate it clearly.

In his latest work, Carl Dair depicted the lowercase 'a' as follows:

We observe a narrow sign with a diagonal stroke connected to a vertical stem noticeably below the optical midline. The back is slightly, but still noticeably curved. In other places the glyph is depicted in a similar way:

However, different designers, in different years and with different goals, approaching Dair's legacy, chose for their versions of the font a more familiar typeface, with a connection at least not below the optical midline:

My question is this: is there a significant reason for such a derogation? A significant reason rooted in the expected form of the modern Latin alphabet, maybe?

The difficulty for me is that I'm not familiar enough with the Latin alphabet, and I can only seriously consider the place of such a glyph in the Cyrillic alphabet. And if we were to create a Cyrillic font inspired by Dair's work, then...

Well, in Cyrillic, such a glyph looks archaic, strange, and becomes the narrowest glyph in the entire alphabet (except for Cyrillic alphabets that use 'i'). Archaism is no vice. The strangeness is noticeable at first glance, but when reading, the eye quickly adjusts and doesn't stumble over every letter. The narrowness is actually a blessing; Cyrillic is tragically overloaded with wide glyphs, and this fence simply needs to be broken up with something.

So, for Cyrillic, I would leave this version at least as an alternative, since modern fonts allows alternative glyph shapes.

But what about Latin? Is it possible to do this in a modern Latin alphabet, or will it definitely be unpleasant/inconvenient to read?

Difference between Cyrillic Ef Ф and Greek Phi Φ

Wei Huang — Mon, 25 May 2026 10:41:57 +0000

What's the recommendations on the design difference between the uppercase Cyrillic Ef Ф u+0424 and Greek Phi Φ u+03A6? I've noticed there’s 3 main differences:

1. Cyrillic Ef is sometimes more square than Greek Phi:

Helvetica World Regular

2. Size of the bowls are different:

Theinhardt Pan Regular, Cyrillic Ef’s bowls are wider

Sharp Earth, Cyrillic Ef's bowls are narrower but taller

SF Pro Text, Greek Phi’s bowls are bigger

Graphik Regular, Greek Phi's bowls are bigger

3. One may extend beyond the cap-height and baseline while the other does not:

Zed Display Regular, Cyrillic Ef’s bowls are larger and the stem extends beyond cap-height.

Neue Helvetica World Regular, Greek Phi’s bowls are larger and the stem extends beyond cap-height.

Also see SF Text Pro above, both extend beyond the cap heights but Phi is slightly taller but probably on account of the bigger bowls.

Would love some insight into these differences.

Best practice for kerning?

benwiggy — Sat, 20 Jun 2026 17:53:08 +0000

The biggest issue in font design seems to be kerning. I spend about eight times longer messing around with kerning than I do with actually making the design. Kerning is always the first thing that gets criticised; and the inevitability of 'missing one' is a huge frustration.

What's the best way to ensure decent kerning? FontLab's automatic kerning of certain pairs (but not all pairs) is 'non-optimal' – but it strikes me that kerning is exactly the sort of thing that computers ought to be doing for us, yet neither FontLab nor Glyphs do this well, if at all, without third-party python scripts, etc.

Currently, I'm just setting large swathes of text in my font, or creating random combinations of glyphs, looking through it, and then manually adjusting by eye in FontLab when I see something that looks wrong.

Am I missing something? Thanks.

Vertical metrics standards

Tofu Type Foundry — Wed, 23 Apr 2025 19:46:44 +0000

I have some naive questions. This article from the Glyphs team dives into the complex issue of vertical metrics; even after reading it numerous times I still don’t fully understand the intricacies. In the past I’ve gone with “the Google strategy” outlined in the article, since it appears to be a well-rounded solution.

What I’d like to know is: Do you use one of their recommended methods for setting hhea, typo, and win values? Is it necessary to use a different method because of the needs of the typeface? Does each foundry have a standard so all their fonts work well together in their own library?

Circled zero. Any real-world use cases?

Ermin Međedović — Wed, 24 Jun 2026 10:18:25 +0000

Most of my typefaces include circled numerals in the character set. I’m currently drawing them for an upcoming family and started wondering about the usefulness of a circled zero.

Has anyone encountered a real-world use case for it?

The only one that comes to mind is when a font provides only single-digit circled numerals (①–⑨) and someone needs to compose numbers such as ①⓪, ②⓪, ③⓪, etc. by combining a circled 0 with another digit.

Curious whether there are practical applications I’m missing.

[Do]lorem ipsum

michele casanova — Mon, 01 Jun 2026 06:01:15 +0000

This video was pointed out to me as interesting because it reconstructs the history of "Lorem ipsum".

https://www.youtube.com/watch?v=kL1PDqzqhM4

"Lorem ipsum" was apparently created in 1966 by James Mosley (1935-2025) from four pages (36, 56, 70, and 118) of this 1914 edition of Cicero's "De finibus bonorum et malorum": https://archive.org/details/bub_gb_ufOZBzV878IC/page/36/mode/2up

The text (specially modified to be meaningless and to have a letter frequency similar to English, for example, by reducing the number of Qs) was requested by Letraset to create "dummy text" sheets.

The "Lorem Ipsum" text was then used in Aldus PageMaker 3.0 (1988 ca.) with some changes and the introduction of obvious errors ("nibh", "zzril").

Was all this information already known?

Honorific ligatures for Islamic texts: gaps and open questions

husada — Sat, 06 Jun 2026 09:27:56 +0000

We are building a digital library platform for classical Islamic texts, and in the process of evaluating fonts for the platform, we ran into a challenge.

One of our goals is to offer readers, particularly those already familiar with classical Arabic book printing, an experience that feels both authentic and comfortable to read, while remaining clean and modern. For that, we needed a font that is a clear, readable Naskh; well-supported; and importantly, one that includes well-crafted honorific ligatures.

We chose to limit ourselves to OFL because we need a font we can freely use, host, and potentially extend. Among the OFL options we evaluated, Scheherazade New stood out as the strongest candidate overall. However, its honorific ligatures did not match the aesthetic we were looking for.

The best honorific glyphs we found were in KFGQPC Arabic Symbols. The visual quality and calligraphic weight there are excellent, and clearly informed by established printing tradition. But KFGQPC has several problems for our use case: its glyphs are placed in a separate symbols font using Private Use Area codepoints rather than the assigned Unicode honorific slots, which means they are not text-searchable and interoperable; its coverage is also incomplete, with many honorific expressions common in the broader Islamic scholarly tradition simply absent; and finally, its EULA does not permit modification or redistribution.

So we ended up handcrafting them ourselves: 61 honorific ligatures as clean SVG artwork (union paths), drawing from the visual standards of classical Arabic printing, with reference to the visual forms and compositions of KFGQPC but not directly derived from them:

GitHub: https://github.com/baalwi-id/arabic-honorific-ligatures

Coverage spans 11 groups (alh, sal, slt, slm, slw, rad, rhm, rmt, qds, hfz, mix). Most have assigned Unicode codepoints; a smaller number are PUA candidates.

Since we are graphic designers rather than font engineers, we are limited in our ability to take this further on our own, so we raised the issue with the SIL team. The discussion is here: https://github.com/silnrsi/font-scheherazade/issues/15

The conversation has evolved in an interesting direction: it seems what would be most valuable for Scheherazade New is improving the consistency of its existing honorific letterforms with its own Naskh base, rather than replacing them with our Thuluth artwork. Our Thuluth collection remains available as a standalone open-source set for anyone who needs it, and we are also separately exploring the possibility of building a Scheherazade New derivative that incorporates the Thuluth honorifics as a distinct font, though that is still in early consideration.

This feels like the right forum to get informed feedback on all of this: the artwork, the approach, the SN proposal, or anything else. Very happy to hear from anyone with experience in Arabic font development or Islamic publishing typography.

Anyone paint as an outlet?

Adam Ladd — Fri, 08 May 2026 15:33:39 +0000

Kind of specific, and opened a discussion similar to this a couple months ago...

But I'm curious, given the element of drawing involved in type design, if there are others out there that paint (or other forms of art) as an outlet?

I've tried to pick up a brush again from time to time, but tough to get myself to dive deeper into it and enjoy it as a way to get off the screen and have a more tangible expression. I think it's partly committing to the time and then cleanup involved (considering watercolors partially for that reason).

Similarly, curious if there are certain styles/movements that most inspire you in that area or even inform how you view type design and the shapes? For me, I've found myself drawn to impressionism and the semi-freedom of it to reinterpret but still be understandable.

Textura Eszett

Nick Shinn — Tue, 12 May 2026 21:52:36 +0000

My latest project, Puffery, updates the “Olde English” style of blackletter with Roman-based capitals—providing better legibility for those unfamiliar with the classic Fraktur capitals, and also to enable a neat all-caps setting. So I had to include a suitable capital Eszett, giving it some commonality with the lower case ß.

There are two variants of the typeface—Vintage Capitals (above) and Regular (below).

Open type Glyph table version 1

mitradranirban — Thu, 26 Mar 2026 03:10:55 +0000

Is the proposal to include cubic bezier in .ttf fonts accepted by OpenType?
If enabled, what will be its pros and cons over cff/cff2?

Postscript name issue.

ok_locksmith_8414 — Sat, 25 Apr 2026 15:31:20 +0000

I am testing my typeface through Fontspector, and this FAIL message comes up. (I included a picture of it below.)

The message pretty much explains what the issue is. In some apps, instead of just “Rolesville,” I get Rolesville-Thin. It’s saying the PostScript name is wrong. I am making a 9-weight typeface (100-900). My font file has three masters: Thin, Medium, and Black, with interpolated instances in between.

I tried to research this, but I don’t understand it.

Can someone tell me what settings I need to add or adjust in my font file to fix this issue?

I am using Glyphs 3.4.1

Thanks.

Which form of the Cyrillic lowercase ghe with a stroke (uni0493) is preferable in the italics?

Stefan Peev — Tue, 17 Mar 2026 14:43:57 +0000

Here are some semples of the Cyrillic lowercas ghe with a stroke (uni0493). Which one is more correct for the italic style?

Redraw vs adjusting font (which is faster/easier?)

Carl Enlund — Wed, 01 Apr 2026 13:56:49 +0000

This is me thinking out loud:

Isn't it sometimes faster/easier to redraw your font from scratch? When you realize a lot needs to be changed in the design.

Rather than trying to make adjustments to existing glyphs/nodes.

Redraw = throw away what nodes/paths you've drawn and draw them from scratch (perhaps keeping old drawing as visual reference).

Factors that can influence whether it's worth the effort to redraw font:

- current glyph count (maybe redrawing falls apart at >50-100 glyphs depending on complexity)
- how big the change is
- master compatibility
- if glyphs use simple construction that can be derived from a recipe, e.g. use /d to make /p - then you just follow that recipe to quickly redraw whole font from scratch
- how strictly you want to keep shapes/motifs consistent across glyphs (strict consistency could favor a redraw)

Types of changes I'm thinking of below. Would any of these be sufficient for redrawing from scratch:

- vibe feels off (biggest reason to redraw i believe)
- change genre
- proportions feel off across glyphs
- bad curve structure (tooling for harmonize and extremas are useful here)
- make font bold (tooling exists to avoid redrawing. And redrawing will break master compatibility)
- change spacing (use tooling duh)
- make italic
- make oblique (good tooling exists to avoid redrawing)

What are your thoughts? When to redraw font from scratch?

/Carl

How did you used to do interpolations without middle masters?

Ray Larabie — Sat, 04 Apr 2026 02:52:09 +0000

Not that long ago, before Glyphs, FontLab only allowed two masters per axis. For example, if you had a weight range between thin and black, you couldn't put a medium master in the middle. Rather than use MM interpolation, I used to use the blend tool, make manual adjustments to the interpolation, then subdivide and repeat. Nowadays, I can add as many intermediate masters as I want. So how did you handle it?

What is the South American "satillo" in shape and use?

Chris Lozos — Tue, 17 Feb 2026 06:50:41 +0000

I am adding the South American glyphs to a sans font and have never designed a satillo and am looking for advice from a native speaker.

Best practices for multi-script families

Tofu Type Foundry — Tue, 24 Mar 2026 19:31:52 +0000

Is it recommended to separate font families by script? I see some multi-script families choose to split into different fonts per script (Futura Latin, Futura Cyrillic, Futura Georgian) and other times everything is under one umbrella (Futura, which supports Latin, Cyrillic, and Georgian in the same package). I’m guessing the difference in vertical metrics of certain scripts is a leading reason behind splitting families. Alongside file size for web fonts, perhaps?

For those that do create separate families, do you include other script’s characters in that font? Example: Does Futura Arabic only contain Arabic characters? To type an English word would you need Futura Latin? Or are there some common characters that all scripts will include, such as a basic ASCII character set?

[I’m not actually discussing Futura, just using it as a generic typeface name]

What hobbies or activities do you do for a break?

Adam Ladd — Tue, 18 Nov 2025 16:04:55 +0000

When the tedium parts of the process set in, or your eyes and head are fatigued from concentrating on the screen and details for too long... what do you like to do to help 'reset' and refresh?

Just curious to hear and compare. I find I need to be off anything screen-related and do something more tangible or loose, like a bike ride.

Transliteration into Latin characters of Cyrillic characters

michele casanova — Mon, 09 Feb 2026 07:44:13 +0000

Is it reasonable to include two character variations (cvXX) in a font to transliterate Latin characters into Cyrillic and vice versa? Or is it just a waste of time?

I know there are numerous variations, but the ISO 9 table appears to be bidirectional; there seems to be only a case issue for ъ and ь.

Halfwidth and Fullwidth Forms

michele casanova — Sun, 25 Jan 2026 08:27:48 +0000

Does it make sense to add elements from the "Half-Width and Full-Width Shapes" block (FF00-FFEF) to a font that doesn't contain Asian characters (Chinese, Japanese, or Korean)? When might these characters be useful?