New to Typophile? Accounts are free, and easy to set up.
Create an account
Typophile RSS | More Feeds
I would like to learn about various aspects of designing Arabic ( Persian or Urdu ) fonts.
What advice would experienced designers give on hinting, ligatures, etc. peculiar to these languages? What problems are beginners (like me) facing?
A great many links and resources on this recent typophile thread.
Thanks raph. I have two specific questions: 1-To make an Arabic font in TT or OT? 2- Best hinting procedure to adopt for a new Arabic font? There is a great variety of curves, and straight lines inclined at various angles, not to mention dots and marks at several vertical heights. I would welcome any suggestions.
1. Given current tools TTF would be esiest... http://blogs.adobe.com/typblography/2007/06/adobe_arabic_sa.html
2. Unless you're making a UI font or targeting a specific rendering environment, then I wouldn't worry too much about this.
Thanks sii, the adobe blog site seems busy or out of order I will try again, but TTF it will be. I worry about the hinting because at 8 or 12 points most Arabic fonts become practically illegible on screen, not to mention extremely ugly. This need not be so...
Try finding it via http://blogs.adobe.com/typblography/
I don't think you can avoid ugliness and illegibility on screen at small sizes. 8pt on screen gives you 11 pixels of vertical space, and the most common marks as well as the tallest ascenders and descenders need to fit in that space. You either simplify the writing system (uglification to most readers) or stick with traditional forms (which render illegibly at that size). Arabic, like many writing systems needs more pixels than the software engineers give it.
It would be wonderful to be proven wrong here, and you could do that by first pushing pixels within the 11 pixel limit, before moving to outlines that work for those shapes - kind of like what Matthew Carter and Tom Rickner did for Verdana.
Thanks. It would be great if there is a font format that uses outlines down to say 11 point and then automatically revert to pixels for a single choice of smaller size- perhaps 8 point, and do that only for screens and not for printers!
I am not sure what you meant by using pixels for Verdana.
In its long history Arabic script has appeared in a marvelous variety of styles ranging from the highly curvatious and decorative to one that is all straight lines and right angles. 11 pixels is quite enough for a variant on the latter. I looked at the UI font and the Arabic in it is pleasing and reads fairly well, I wonder who designed it. The font I am working on now (since 1961 actually but I want to make it into a computer font) is different - inspired by Eric Gill's Sans-Serif and his writings.
> I looked at the UI font and the Arabic in it is pleasing and reads fairly well, I wonder who designed it.
I wonder where I can see a sample of it
> The font I am working on now (since 1961 actually but I want to make it into a computer font) is different - inspired by Eric Gill’s Sans-Serif and his writings.
Since 1961? I wonder what does it look like!
All the Best with Flowers
>I wonder where I can see a sample of it
I meant the Segoe UI font found on Word 2007 and Vista applications.
>Since 1961? I wonder what does it look like!
Well as a student I became interested in the Bauhaus and lettering and at that time I felt Arabic type needed to develop in that direction. I have only recently resumed my interest in font design. Here is a description of my typography work (excerpted from an article "Influences and Motivations..." on my website) , written around 1985
A MODERN ARABIC TYPEFACE
One of my first projects was a study of Arabic typography. Although Arabic calligraphy is famed for its glorious perfection as an ancient art form, its transformation into printing type caused several technical and aesthetic problems, since type makers simply tried to imitate the calligraphic style without making the necessary adjustments. For example, dots, the carriers of important information in Arabic script, became almost invisible. Inspired by the writings of the English typographer Eric Gill, who advocated a simple, clean, sans-serif type design, and by firsthand studies at the British Museum of' an early Koran written in Kufic script, I designed and patented a general-purpose, simplified Arabic typeface that I named after Jerusalem, Al-Quds [1,2] , shown in Fig. 2. Although such 'modern' Arabic lettering without distracting ornamentation was popular for shop signs and product names, a suitable typeface for printing using simplifications of this sort (such as Al-Quds) had not been designed. Eventually, modern typefaces became popular for magazine headlines, but unfortunately Al-Quds is not yet available as an Arabic printing type.
Fig. 2. Sample of Al-Quds Arabic printing font (c. 1972). Only 56 characters are needed to set an Arabic text with this type-face.
The patent referred to is:
It's interesting that you like it, and I've heard that many people prefer it to the Tahoma Arabic. The Arabic in the font was actually taken from Microsoft Sans Serif, and was a last minute addition to Segoe UI. It was added when localization worked out that Microsoft Sans Serif wasn't work well at 9pt.
sii, Does localization refer to how local readers reacted to the font? Analyzing the three fonts you mentioned was educational for me - each has good and bad points.
The Microsoft Sans Serif Arabic has clean letter-forms, but is too condensed, and its diacritical marks (harakat) are messily places in between the letters.
The Tahoma Arabic has an uncluttered airy appearance that make it legible at small sizes, but the letters are limp and characterless - a bit like comic sans. The isolated nun floats incongruously above the horizontal midline, and its harakatb clash between lines. The bold is bloated but the accompanying dots are tiny.
Segoe Arabic with its large shapes has a certain elegance and originality. The variable-thickness lines read poorly at very small sizes, and some letter-combinations (the initial 'ain and middle noon) are much too close together.
All three fonts suffer from two fundamental problems: 1- The dots on the letters are tiny. This is inexcusable, since 15 out of the 26 Arabic letters rely on dots to distinguish them from their dotless twins, and on a monitor dots almost disappear at small sizes. 2- Within the same font some letters (like the lams) are purely equal-width, and some (like the kaf in all three) combine thick and thin strokes. The reader, used to the way naskh or thuluth script changes thickness naturally as it curves, or according to the stroke angle, would feel uneasy with these inconsistencies.
>This is inexcusable, since 15 out of the 26 Arabic letters rely on dots to distinguish them from their dotless twins, and on a monitor dots almost disappear at small sizes.
Remember these are user interface fonts - you're not supposed to use them in documents or Web pages - just in menus, button captions and UI dialogs.
>Remember these are user interface fonts - you’re not supposed to use them in documents or Web pages - just in menus, button captions and UI dialogs.
Sorry if I expressed myself rather strongly concerning small dots - I wrote about the problem in the early seventies, and it persists till now even in print, not just on monitors. I am not at all sure what you mean about not using the fonts in documents or web pages..do you mean embedding ? Is it a legal issue? What about a Word document or a CSS specifying one of these fonts installed on a typical computer?
>I am not at all sure what you mean about not using the fonts in documents or web pages
No, nothing legal they just were not designed for use in documents or web pages.
>No, nothing legal they just were not designed for use in documents or web pages.
I see - do you know of any Arabic fonts that were?
Shukran I signed up for Khatt.net it is very nicely set up. I hope they include an Arabic website alternative as well.
Does anyone know if Typophile encourages or discourages posts in languages other than English- ie Arabic under this topic?
Here at Typophile, you are free to post in any language you want! It may be difficult for you to get a lot of responses if you post in Arabic, because other Arabic speakers will have to find your text first, but we occasionally get posts in French, Spanish, German, etc. As one of Typophile's moderators, I encourage you to feel free to post i any language you see fit!
شكراً دان رينومدز = だん れいのぅづ さん ありがとう = Thanks Dan Reynolds. Of course you are right that using Arabic here limits the search options. Another point is that there are many Arabic fontographers who do not read or write Arabic (Eric Gill was one example), but I just wondered about the possibility. In fact discussions at www.khtt.net a site devoted to Arabic typography, are almost exclusively in English!
Thank you for sharing the information about your UK patent of 1963 regarding al-Quds font. In my upcoming article for "Visible Language", titled "Anatomy of an Arabetic Type Design", I mentioned briefly a similar Patent in 1971 by the prominent Iraqi type designer from Basrah living in Paris: Muhammad Said al-Saggar http://saggar.com He called it "Saggar's Alphabet" or الابجدية العربية المركزة. A font was developed for it and it was used in the seventies by a major Iraqi newspaper called "al-Thawrah". Thomas Milo, a well-known Arabic and Islamic type designer uses same method as yours and Saggar to create most sophisticatedly calligraphic fonts. His latest product is Tasmeem. Here is a gif of Saggar's alphabet: (Notice: the sample text is set with the his alphabet)
Dear Saad - Thank you for the most interesting glimpse of developments of the thanab or tail concept for common endings of Arabic glyphs. Since he tried his utmost to minimize the number of master shapes Saggar could have used one size for the hamza. Apart from the convenience it must have given metal type setters, it may also have helped standardize glyph design within a given font. However these days with the convenience of computerized font design, automated ligatures in word processing these considerations are perhaps obsolete. Saggar's website has excellent examples of calligraphic artworks but only a few words written with his fonts.
It is obvious that there were quite a few people active in Arabic typography in the last fifty years - I wonder it Huda's book covers their work in detail. By now there are so many new and old fonts I feel it is necessary that there be a serious attempt at critical analysis and discussion about what makes a good Arabic font. A designer should feel that the public appreciate good design and that one's fellow designers give honest and useful feedback. Incidentally Basra seems to be a place for creative people now I know of two typographers from there!
>>By now there are so many new and old fonts I feel it is necessary that there be a serious attempt at critical analysis and discussion about what makes a good Arabic font.
Great point, but it is very risky to put it in term of "good" and "bad". We do not need to create rules. Of course, one can evaluate technical quality of font design, but this should not be based on the so called "script rules". What is needed in Arabic typography, in my opinion, is "open-mindedness" and return to our ancestors' tradition of renewal and creativity. We must open the doors wide to options to ensure a decent pool of fonts and let users and market decide. Attrissi, a Lebenese type designer, once wrote that the main problem with Arabic type is availability. I strongly agree. In Latin you have millions of fonts to choose from, while in Arabic we still live in the mentality of "this is allowed and that is not". One can only see how different Diwani calligraphy style is from Kufi to appreciate what openess and renewal did to Arabic a thousand years ago.
Today, we need to get rid of the "mentality of censorship." Believe it or not, most of those in effective positions in Arabic typography today do not only actively censor their colleagues' work, like mine (which is "understandable") but they even censor traditional Arabic calligraphic fonts like Aziz's! I think it is crucial to think that a font liked and used by 1% of users is as important and justified as that liked and used by 90%. Especially in the age of digital typography, fonts are not only about printing.
>>Incidentally Basra seems to be a place for creative people now I know of two typographers from there!
Any great poets, artists, writers!
>>“open-mindedness” and return to our ancestors’ tradition of renewal and creativity. We must open the doors wide to options to ensure a decent pool of fonts and let users and market decide.
Saad - Amen to that. Creativity is important and one should not use obsolete rules to judge fonts. We are living in an era where almost anything goes and people are enjoying unprecedented freedom to make all sorts of fonts. Fine I do not wish to spoil the fun, and as you say the mentality of "this is allowed" or not is stifling. What makes us say a thuluth hand is fantastic, while another gives you an uneasy feeling of imperfection and unease? It is that sense of the "educated eye" that has to be transferred to those who design fonts, use them or read them. Some fonts in online newspapers have letter-dots that are virtually invisible- yet there is no public outcry against that - surely that is a case of a "bad" type in the sense that it is not fit for the use it was designed for. You point out the creativity of the old scripts, but they developed over centuries! Now everything is happening too fast, and I was expressing my worries about the situation in general. How fonts are used is another important aspect. For both text and display surely legibility is a necessary minimum standard? I see that we are not disagreeing on most things Saad, and you know more about the market and current usage.
Vladimir, Indeed, we are not disagreeing at all. I just pointed out these issues so that your statement is not misunderstood. Yes legibility is very necessary for certain applications, but not as necessary in others.
>>Some fonts in online newspapers have letter-dots that are virtually invisible- yet there is no public outcry against that - surely that is a case of a “bad” type in the sense that it is not fit for the use it was designed for.
Yes. This is true. Similarly, I think modern Arabic typography sacrificed plenty of the glory of Arabic calligraphy without real typographic gain. They abandoned calligraphy but still vigorously hold on its old unnecessarily restricting rules. Honestly, I do not think many of the modern Arabic fonts are necessarily more legible just because they follow the attached four glyphs per letter model.
Saad- I missed seeing your last comment until now.
>>legibility is very necessary for certain applications, but not as necessary in others.
Very true. Neither is consistency- in one of your fonts the aleph and lam have totally different shapes I thought that was original and fun because it was clearly a font meant to be used in a decorative way.
>>I think modern Arabic typography sacrificed plenty of the glory of Arabic calligraphy without real typographic gain. They abandoned calligraphy but still vigorously hold on its old unnecessarily restricting rules.
In general I cannot agree with you more, but as they say the devil is in the details. Perhaps one good rule is having good taste in the design, and again that is not at all easy to define. One example of what you call "typographic gain" might be text legibility.
Three suggested 'rules' for legibility may be
1-Continuity in the traditional letter shapes - ie nothing too eccentric compared to traditional letters. Not only will readers feel comfortable with forms they know, but also there is a wisdom in following the 'design logic' of a script honed by generations of writers and readers over centuries.
2-Each letter form should be as different from the others in the same font as possible.
3- The letters should be connected because this gives a unique word-form or pattern for different words.
>>I do not think many of the modern Arabic fonts are necessarily more legible just because they follow the attached four glyphs per letter model.
Not at all - minimizing the number of glyphs for a given letter is the modern trend. But remember that modern Arabic has to written and not just read. For example fonts used to teach Arabic in schools should be easily 'copied' as handwriting. This is one thing I tried to do in my AlQuds font. The initial ain becomes a medial ain and a final in one continuous line. Some modern experimental fonts use the initial ain as a medial one and that makes it awkward and even ugly as handwriting.
>>3. The letters should be connected because this gives a unique word-form or pattern for different words.
Not unnecessarily required. Isolated, well-designed, letters can do same.
>>But remember that modern Arabic has to written and not just read.
Yes, written would make more sense connected, of course. But I think we should not hold progress in Arabic typography and Arabic script learning for the sake of handwriting alone. One can learn shapes isolated first and even write them that way, and learn cursive later. The point is that we need to eliminate restrictions.
>>>The point is that we need to eliminate restrictions.
I think that is happening whatever thoughtful designers like us may plan for. The tools for font design are fairly easy to master and there are plenty of young people eager to try something new. It is not that I am against innovation of course - that can, and will proceed on its own almost. In the end the users will decide which font to use for which purpose, and as you said earlier a large pool of available fonts widens their choice. As for me all this can be baffling. Sometimes I long for the days when we had one or two bad Arabic fonts for everything from schoolbooks to newspapers ;)
Aziz, Saad and all who are celebrating the blessed Eid feast.
Eid Mubark Vladimir!
Nadine Chahine of Linotype has posted advice to beginners at Arabic Typography.
Her blog is also full of other fascinating material on related subjects:
Similarly, I think modern Arabic typography sacrificed plenty of the glory of Arabic calligraphy without real typographic gain. They abandoned calligraphy but still vigorously hold on its old unnecessarily restricting rules. Honestly, I do not think many of the modern Arabic fonts are necessarily more legible just because they follow the attached four glyphs per letter model.
This is very interesting Saad, could you explain better what you mean? I believe Latin should not be strictly identified with a "stranger" culture by Arabic users not using it.
On a side note: I'm eager to have a look to what you are doing with numerals. I just bought one of your faces on MyFonts to have a look…
In another thread, Vladimir wrote:
I guess I am one of those people who feel that the tradition of writing connected words is too important to get rid of, and that connected words create a ’word shape’ that aids legibility.
Talking of Arabic related to calligraphy, I think it should not be forgotten that the Carolingian minuscule (as many so-called National Writings preceeding the renaissance) was almost a connected script. Then, we should remind the Cancelleresca cursive was often as complex as the example of Arabic calligraphy you so rightfuilly cherish.
If we look at what happened in typographic evolution, we can see how Latin evolved its fisionomy through writing in long centuries. I just think such a thing can't be "forced" on Arabic, but I inversely think experimentation by non-Arabic people could be less "compromising" than what it might be felt (in history, any culture has always been enriched by a communion with others).
@ I just think such a thing can’t be “forced” on Arabic
Save in Arabic Flowers where words are made to diverge (from the center to the outside of the flowers).
Arabic culture has always been enriched by a communion with others?!
Then, we should remind the Cancelleresca cursive was often as complex as the example of Arabic calligraphy you so rightfully cherish.
After a year of intense work with Arabic font design it strikes me as strange how we become so sensitive to nuances of design of tiny letters. Script, lettering, type, handwriting...they all evolve and interact over the centuries in fascinating and forgotten ways. Nowadays the changes are much more rapid. Anyway one thing I have been wondering about is exactly what piccic has pointed out: why did Latin languages leave the tradition of connected scripts and adopt separate letters in printing? Or rather, is there no possibility to use the amazing computerized technologies of today (as Tasmeem has done for Arabic) and devise a cursive connected Latin typography...this is all off-subject but then so is Aziz's own carnivorous apple which has been haunting me for days haha.
Vladimir -- why did Latin languages leave the tradition of connected scripts and adopt separate letters in printing?
Maybe this is the wrong question. I doubt that "separate letters" were "adopted" for printing ("adopted" meaning that someone had a choice and made a decision). At the same time it may not be an accident that printing with separate letters (but don't forget the many "ligatures" in Gutenberg's case) was introduced at a moment when blackletter ruled in representative texts. Blackletter was "connected" too in a way, producing a tight web or texture, and it required special markers so one could tell apart an i from an n. But blackletter was "connected" in such a way that it could be broken up into segments which, when put together again, again produced an image of "connectedness". It could have been a combination of (a) Gutenberg seeing a market for multiple copies of representative texts produced in a quicker way than writing or cutting a block per page, and (b) the fact that representative texts were written in blackletter, which allowed him to make his invention in the first place. (I.e. the dominance of blackletter in these texts, which he wanted to produce, made him realize that letters could be separated. Rather than inventing a technology from scratch and then find a matching writing style.) Since printing with separate letters had proven to be the way to go, when moving South, letterforms had to be adjusted so to look more familiar to South Europeans' eyes. As far as I know, there was no sudden jump from using blackletter to using round, light roman type, but rather blackletter shapes were "romanised" in small steps.
What you are saying makes sense: movable-type printing became possible because blackletter consisted of separable letters, a natural and logical metamorphosis. Arabic's experience with type was uneasy at first: the scripts were too complex to chop into units connected horizontally. But then eventually this lead to the new Arabic typography where the letters became simplified in the extreme to suit the technical limitations of printing methods using consecutive glyphs. Meanwhile font designers for Arabic continue working in different directions - On the one hand type imitating calligraphy whether by designing thousands of ligatures, or by magical computerized methods of juggling a couple of hundred shapes as Tasmeem does. Next to these two groups are the majority of Arabic font designers, creating all sorts of designs in a bewildering variety of styles, just like font designers are doing in many other languages. Human ingenuity always finds a way!
why did Latin languages leave the tradition of connected scripts and adopt separate letters in printing?
I don't want to go off-topic, but I think this talk brings us to a good point by where a balanced understanding of what we look as "culturally respectful" may be jointly developed, and to see how the evolution of alphabetic scripts should be considered when each one is related to the others.
For Latin, actually, it was a gradual progression. Karsten points out one of the important reasons, that moveable types were perfectioned by Gutemberg while in Germany the leading style to compose pages in calligraphy was Textura, which, for its peculiarities, proved to be very good for this. It should be reminded, however, that Gutemberg's 42-line bible had a devastating number of ligatures. Modular as it may look, an early Gutemberg page had a complexity similar to the one of certain Arabic calligraphy. Vocalization may be more important in some languages (like Hebrew, where you have mostly consonants), but it shouldn't be forgotten that – as long as conventions take place – it becomes automatic for a native reader to distinguish words in context without the need of expliciting the accents.
To see how the evolution happened it might be enough to look at the books progressively printed by Schweinhem and Pannartz at Subiaco. I have a good iconographic documentation on those, in an Italian book of the 1940s about the history of printing.
For Arabic moveable type, I think it would be great to have a basic history written by some authoritative native designer, but in the meanwhile a question occurring to me is: did you have some kind of exposed writing which eventually led to calligraphic styles in order to write quicker and on different supports (a thing similar to what happened to Capitalis Romana down to Capitalis "Cursiva" (skeletal) and the Rustica)? Does Arabic have a paleography lineage linking exposed writing styles to early notes and manuscripts that could be followed? I am sure of that…
Sorry for bringing too many points to the discussion… We should develop just one of them.
I love typogharphy.
I like jon hudson. He solve problem easly.