Creating a Complicated OpenType Font for Biblical Hebrew

gohebrew's picture

In general, a set of typeface editing software programs are used to create a well-crafted professional-quality Biblical Hebrew typeface which fully supports the rich features of diacritic placement in the popular standard Microsoft/Adobe font format.

One program tool in particular is essential to create an "intelligent" or "smart" OpenType Biblical Hebrew font. That is "Microsoft VOLT". Although the user interface is graphical, and after one learns how to navigate it, is rather easy, it appears to be complex, difficult to understand, and user-unfriendly. Only a team of Microsoft programmers could have made it.

If Apple engineers would have created "Apple VOLT", every two-bit self-proclaimed type designer would be knocking out hundreds of passable OpenType fonts, ruining it for the rest of us. It's in our best interests that the wizard hides behind the curtain, if you know what I mean.

Until Apple gets back at Microsoft and Adobe, and creates a new font format, called "SuperType" (like TrueType was Apple and Microsoft's way to screw Adobe and PostScript [did John Warnock really cry, "Now, everything Getchke and I planned for PostScript is ruined!"] - ah, remember when Apple didn't have Steve Jobs?), let us explore why Biblical Hebrew is best suited for OpenType.

First, it takes the brain work about correct Biblical Hebrew typography out of the hands of page layout professionals, and into the sweaty palms of qualified type designers - yes, those fearless few. Second, it borrows fascinating rules of Hebrew grammar from Hebrew school teachers, and lends it to the masses. Now, those teachers can keep their students awake and actually learn these rules of Hebrew grammar by using fun exciting neato groovy and keen software tools.

A sample of Biblical Hebrew text from Joshua 1:10-11 in the OpenType Biblical Hebrew font, FrankReuhl GH. Frank and Reuhl were two Germans in the early twentieth century, who created this popular standard Hebrew type design. It was the Hebrew version of Times-Roman, in that it replaced the famous Vilna type design standard of the 19th century.

Contact FontWorld ( for more details.

Note the special Hebrew glyphs for the shva-na, final nuhn with patach, the folded lamed, the unfolded ayin, and the various final chofs. Featured also is John Hudson's automatic furtive patach, highlighted in blue.

gohebrew's picture

OpenType is an excellent engine for the requirements of Biblical Hebrew. It is missing only a few things needed to produce perfect Biblical Hebrew.

First, Biblical Hebrew characters, like many other languages, require the correct placement of many diacritic marks.

These include:
1) the dagesh mark, a center point in a Hebrew letter;
2) the nikkud vowel symbol, which either appears under a letter (usually), or as a point above and on the left side of a Hebrew letter (known as a "cholam haser" or a missing cholam), or as a centered on the left side of a Hebrew letter (only if the letter is a final nuhn [plus kamatz], or a final chof [plus either a kamatz, dagesh kamatz, shvah {and I think there's also a very rare occurance of a dagesh shva}, and a final mehm with a chirik on the extreme right side under the letter), or the furtive patach, placed on the extreme right side, under the ayin, dotted hei, and chet, when these letters appear at the end of a Hebrew word, to indicate that the patach should be pronounched before the consonant, unlike how it is usually pronounced;
3) the cantorial music note, known as trup, or taamei mikra, or taamim for short, placed either above or below the Hebrew letter, indicarting how to chant the syllables;
4) the mysterious meteg (a short thin horizontal line, placed on the left side of the other lower glyphs, or by itself.

So, in all, there could be 0 to 3 upper markings, above the Hebrew letter, or 0 to 3 bottom markings, below the Hebrew letter, and another one smacked in the middle, or 0 to 2 diacritic glyphs placed beside the Hebrew letter.

Without trying to sound arrogant, I think Biblical Hebrew typographic conditional placement is the most complicated in the world!

gohebrew's picture

OpenType is really great. Like TrueType before it, it supports Unicode.

If a text document is encoded with Unicode, then every character and glyph in the world has its unigue Unicode value. This then avoids the confusion when data is transfered from one program to another, or from one platform to another, and different sets of encoding are used or misused. People who remember the nineties will never forget - what a mess!

OpenType is cross platform, so if the font can be installed on the Macintosh and work, it can also be installed on Windows and work, too. No more file conversions etc.

But do you know what is the greatest thing about OpenType that never existed before in fine Hebrew typography?

Accurate placement.

This no longer is in the field of the operator of the page layout software program or super word-processor - that guy is a good graphic artist at best... it is in the field of the talented type designer.

Accurate placement is now in the intelligent OpenType typeface.

Two examples of the popular FrankReuhl font. One with intelligent OpenType character/diacritic placement, and one without (ah, that font doesn't even have taamim!).

gohebrew's picture

So, how does OpenType do it?

Well, there are features of OpenType, basically speaking. Really, they are called 'features'. Just like in the movies.

CSUB and CPOS were friends. Te'wrain de'lo'meesh'pa'rin. Two friends that can't be separated.

The snail and his shell, as they say.

C stands for character.

SUB stands for a conditional substitution. And POS stands for posee - round 'em up, rawhide.

No, no, no, POS stands for accurate positioning.

So, the most basic aspects of OpenType is the ability to do contextual replacements (CSUB), and carefully position diacritic non-width glyphs upon a base character.

This means that a sequence of Unicode values can be substituted for a different set of font glyphs - that's CSUB.

And a number of diacritic glyph markings can be accurately placed or positioned upon a base character - that's CPOS.

Now, what about seeing and placing multiple diacritic glyph markings, when MS VOLT can only show us and place a single diacritic glyph marking at a time?

Biblical Hebrew could have a whole array of diacritic glyph markings on a single base character.

John, what do you say?

How about Simon, what do you say?

Chajmke's picture

A sample of Biblical Hebrew text from Joshua 1:10-11 in the OpenType Biblical Hebrew font, FrankReuhl GH. Frank and Reuhl were two Germans in the early twentieth century, who created this popular standard Hebrew type design.

Please allow some corrections. The name of the type is Frank Rühl, or Frank Ruehl: Frank stands for Rafael Frank, Chazan and teacher for religious education in Leipzig. Ruehl was the name of a local type foundry.

See a picture of Rafael Frank in my blog:

For those who are able to read german, some additional information:


gohebrew's picture

Sample of CSUB and CPOS in action. See how they run!?

Various different Hebrew letter featuring different kinds of diacritic glyph markings, even multiples carefully placed automatically by OpenType.

gohebrew's picture


Thank you for the correction about Mr. Rafael Frank, the co-creator of the popular modern Hebrew type design.

Do we know biographical information about him? The name of the temple where he was chazzan, for example.

Regarding the discrepencies in my description of him, is simply due to what little details I heard about him, and the person or firm which cut his designs into metal type. From your account, they weren't details, but inaccuracies.

About the spelling of his typeface, I used to spell it Fronk-Ruehl, simply in order to make a distiction between my font, and other people's fonts of similar names, and somewhat similar designs.


I see that this reproduction of this famous design is very similar to my [first] version. Interestingly, Merganthaler/Linotype and Varityper used different designs. There are at least five different designs.


Germany not only produced the popular standard text font of Mr. Rafael Frank a"h, but also one of the designs popularized by Mr. Tzvi Narkis a"h of Israel.

When I lived in Rochester, NY, I had the good fortune to visit the Graphic Arts library and type design collection at R.I.T., the Rochester Institute of Technology, which houses one of the finest and largest collections of type designs in the world.

While there, I happened across a poster about a play that took place in the Warsaw Ghetto or a concentration camp with beautiful hand-drawn Hebrew lettering on it. It was exactly the design later popularized by Mr. Narkis, a sans-serif Hebrew typeface design. I was mesmerized by what I saw.

There, in German occupied Poland was the birthplace of the first sans-serif Hebrew typeface design. And in the ashes of mankind's murderous holocaust was an expression of wonderful lively creativity!

Germany also homed the popular display Hebrew typeface design, Chaim, or Haim Hashamen (the 'fat Hayim') [because generations later a narrow version of Chaim was created, known as Haim Tzar (the 'skinny Hayim')].


Coincidentally, the library's assistant librarian, Mr. Perkow (I think) grew up in Crown Heights, Brooklyn, to where I later moved. Small world, isn't it?

Chajmke's picture

My (german) blog article was some kind of small biography. A very interesting document in this regard is
אותיות דפוס וגופנים

It is astonishing: Rafael Frank's influence on jewish culture is not tiny, but he is almost unknown...

It is a small biography and a small book written by Rafael Frank about hebrew typography. I have no idea whether it is available in english, there is a german one, of course.

The story of Narkis is very interesting. Are photos available? That would be interesting.


gohebrew's picture

Mr. Rafael Frank wrote a book on Hebrew typography? This is amazing, as I believe that it was the first of its kind.

Beforehand, Hebrew type design was limited to a very small set of designs.

I am only aware of five sets of designs.

First, there was the design family of ancient Sephardic origin, popularized by the famous Romm family Talmud publishers in Vilna, Lithuania. Their huge printing establishments dominated Jewish religious and secular publishing in most of the 19th and early 20th centuries, until the Nazis destroyed their existing inventory and burned their original Hebrew book plates and their two establishments. Later, the premises were entirely conviscated by the Communists, and nothing remained but history and dying memories.

Like the simple biography of Frank has been lost to a great part, the story of the illustrious Romm family printing establishment is very limited.

The patriarch of the Romm family traveled to Italy with very old drawings of Hebrew letter forms, according to an account byFrederick Goudy, the great American type designer, in his small book on the evolution of type design. This Romm family patriarch went to the Bodoni school of type design in Italy.

Interestingly, the resulting Hebrew type designs are similar to some of Bodoni famous works, prompting one to question which came first, the chicken or the egg?

I believe the answer rests in reproductions of hand-drawn books which feature this very same Hebrew type designs, dating from periods which preceded the invention of the printing press. This clearly indicates that English type designs popularized by Bodoni were inspired by old Sephardic drawings of Hebrew letterforms.

What is fascinating is that those very popular Bodoni English type designs do not have design origins in other older type designs, and stand out in the crowd, so to speak.

Another style of Hebrew type design is non-Jewish in origin, from the famous priest Bomberg and [businessman ?] Soncino, based upon their interpretations of drawings of old Sephardic and Ashkenazic letter forms, as seen through European Christian eyes.

I was told recently that John Hudson's Adobe Hebrew and SBL-Hebrew were inspired by these sources. Is that accurate, John? In we study this design carefully, we see elements of Mr. Henri Friedlaender a"h popular Hadasa typeface (which seems a cross between the first and second designs).

The third kind of design is Ashkenazic in origin, and similar to the design popularized by the Israeli typographer Mr. Eliyahu Koren a"h.

The fourth style of design appears in the Dead Sea Scrolls and other archaelogical findings. This ia crude script-like design, that looks like it was formed from the speed of the scribe's hand, and not due to any design issues.

The fifth design style is found in the Torah Scroll. It is block-like, squarish, ornate, and clearly rich with design significance.

I agree that with such an important role that the FrankRuehl design has played in modern Hebrew publishing, much too little is written about this very talented designer who revolutionized modern Hebrew type design with his single significant contribution.

gohebrew's picture

Thank you very much for the Hebrew language article in PDF about Rephael Frank.

I read and understand Hebrew, but speak like I'm breaking my teeth. :)

Ironically, it appears in a lighter and modern version of the FrankRuehl typeface, popular in inexpensive scholarly publishing systems in Israel. It loses much of the flavor and originality in the homogenized details. As they say, Frank would have 'puked'.

gohebrew's picture

Wow, what a fascinating article.

Mr. Frank reproduces samples of a few other type styles than I mentioned above.

Also, his classification of various different type styles is in deed most revealing. There are two kinds of Rashi-like script, and a third which is partially derived from it.

It's amazing that this variety of Hebrew type design existed in pre-Nazi Germany.

I heard that secular Jewish culture was very advanced in Germany at that time, perhaps more than in America today. Sadly, the Nazis succeeded in gassing even the remnants of that culture, but thank G-d a little bit remains of its legacy.

gohebrew's picture

How does your blog article neo biography differ from the PDF article in Hebrew?

gohebrew's picture

John Hudson and Simon Daniels,

Now, what about seeing and placing multiple diacritic glyph markings, when MS VOLT can only show us and place a single diacritic glyph marking at a time?

Biblical Hebrew could have a whole array of diacritic glyph markings on a single base character.

John, what do you say?

How about Simon, what do you say?

How would you handle it?

gohebrew's picture

> Now, what about seeing and placing multiple diacritic glyph markings, when MS VOLT can only show us and place a single diacritic glyph marking at a time?

s h o u l d r e a d

Now, what about seeing and placing multiple diacritic glyph markings, when MS VOLT can only show one base character together with a single glyph in a font (be it one diacritic element, or a ligature made up of many diacritic elements) and place a single diacritic glyph marking at a time?

Chajmke's picture

The hebrew was the only non-german version I know ;-) There is also a scanned version available: The picture quality is much better.
My blog article gives a short sketch of his biography and tries to explain the great importance of his work and wonders why he is unknown. There isn't even a hebrew wikipedia article about Frank?!

Yes, the printing and typographic activities were amazing. Do you know the Jewish Encyclopedia article about typography? It is full of pictures and examples:

It also features the Letteris Tanach from Berlin/Vienna (interesting font) and some more examples... it could be of advantage to "collect" examples of Tanachprintings before starting off with a new font ;-)


gohebrew's picture


Thank you for the links. I will review them soon.

david h's picture


Why unknown? what about:

1. The huge exhibition in New York -- 2,000 Years of Hebrew Books (october 15, 1988 -- january 14, 1989; The NY Public Library)

2. An essay by Moshe Spitzer The Development of Hebrew Lettering (Alei Ayin, 1952, Tel Aviv)

3. An article by Ittai Tamari Digitization of Hebrew Fonts (Raster Imaging and Digital Typography, 1989, Lausanne)


theplatypus's picture

Nice bit of information. Full of things I haven't a clue what you speak of. But I will proceed and learn. ha!


gohebrew's picture


Thank you.

I very much liked your clean and deceptively simple web site. The content also reflected this clarity of design. I have not seem something this great in a long time.

Jewish mysticism praises deceptive simplicity as the ultimate expression of the most complex image.

For example, the Creator of the world is described as a "achdut hapashut" - a 'pure and simple' unity.

We see that a person is most effective at communication by saying something simply and directly to the point.

King Solomon, one of the wisest and most intelligent people of all time, was able to describe the most complex concept in very simple words for even the simplest person.

gohebrew's picture


Do you have these two articles (#2 and #3) to post here at the blog on Hebrew Type Design, and can report upon the exibit?

Is it too much to ask you? I hope not.

What's more?

david h's picture


I don't scan + post Copyrighted Materials (#3) — Cambridge University Press; maybe to post my own summary. I don't scan old stuff (#2) especially when there's not a chance to find another copy; again my own summary — different story.

> can report upon the exibit?

Report what? what I'm Diane Sawyer of ABC News?

> What’s more?

More samples of course.

gohebrew's picture


Is the copyright on its content or appearance? If I reset it (in theory of course), is there a violation?

Read it, analyze it, and present a lecture.

Is Diane Sawyer related to Tom...uhh, uhh...Brokau? :)

Samples of different versions of FrankReuhl, or his original kind? Please post it all at the Hebrew Type Designs blog with your comments. Why is it so popular? What did Frank do? Can the Vilna or Romm family "Meruba" font have a come back to replace FrankReuhl?

PS I didn't give up on an automatic shvana etc.

gohebrew's picture

The creation of Shvana into a
furtive patach-like automatic substitution

I reviewed in depth the Nusach Ari, the rendition according the Ari zal by Rabbi Shneur Zalman of Liadi, the first Lubavitcher Rebbe, Birkat Hamazon, the booklet of blessings after a meal.

This booklet is special because it features a graphic symbol for a shvana.

I looked for a pattern when a shvana appeared, and when an ordinary shva(nach) appeared. I asked myself under what consitions did it always appear, and under what conditions it never appeared.

I found that there are a few rules. Therefore, an OpenType font can include CSUB contextual replacement for each rule.

For example, the shvana always appeared over the shva after the first letter of a Hebrew word, if that letter is a shooruk, a vov with a dot on it middle left side. One rule.

There are other rules as well. Stay tuned for more.

david h's picture

> This booklet is special because it features a graphic symbol for a shvana.

Israel, do you have a scan/sample?

Chajmke's picture

@ David: Of course Rafael Frank is not unknown among the few specialists, but his cultural significance is completely underestimated. Both: in Israel and in Germany.


John Hudson's picture

My own view on this topic is that sheva na is properly an encoding issue. It may or may not be possible to automate a glyph-level solution based on the encoding of sheva, but even if possible this does not mean that this is the best or most appropriate way to handle the distinction. As with the qamats qatan / qamats gadol distinction, this is something that belongs at the plain text level so that users can reliably interchange text while maintaining this distinction. The furtive patah is not a good parallel, because that is a standard grammatical rule, while visual display of a distinction between two different pronunciations of sheva, like the distinction between two different pronunciations of qamats, is a user/publisher-specific preference. For users who wish to make the distinction, we can assume that they would like the distinction to be retained when they send text e.g. by email or published over the web; ergo, the distinction needs to be made at the plain text level and not rely on a glyph-level display solution.

gohebrew's picture


Do you want to have your cake and eat it too?

If the glyph is in deed based a major grammatical rule, like the kamatz gadol, then it should be data transferable, and hence, deserves a Unicode value.

The furtive patach then is a minor grammatical rule, then it does not have to be data transferable, and hence, does not need its own Unicode value.

If so, a hataf kamatz kadol and a shvana (both which we implemented already), should have a Unicode value in the next Unicode standard.


I agree that for data transfer purposes, the implementation in an OpenType font as an automatic feature like the furtive patach is unnecessary.

The advantage of implementation in an OpenType font as an automatic feature like the furtive patach is that existing Unicode data of the Hebrew Bible, Jewish prayerbook, poetry, children's primers etc., can have these symbols placed automatically, correctly, without keying in anything.

[Can a user search and replace it for a Unicode value? or is a "virtual" glyph?]

gohebrew's picture


> can you start a new thread to post your VOLT questions?

> can an MS Volt and an OpenType source file, or a font to ship, have a reference to a complex exception table for look ups?

For example, can I designate a CSUB to occur whereby one string “a” is replaced with a different string “b”, except in the cases listed in the exception table.

Another question about MS Volt/OpenType support wild cards in the CSUB strings, whereby one string “a” with a wild card or more in between is replaced with a different string “b” with the same wild card or more in between?

Michael Cunliffe Thompson's picture

I'm interesting in creating a SIMPLE open type font. I'm looking for a relatively gentle way to ease myself into the
process. I'm a user of Font Creator and so was very interested
to come across this recent post:

Adding OpenType Features

Is this suited to working on a in Hebrew font?


Typograph's picture

using VOLT makes no difference what peogram you used to produce your TTF file, as long as the font is a working TTF Volt will handle it.

Michael Cunliffe Thompson's picture

The article I referred to suggests using the OTComp tool rather than Volt.
Any opinions on that?

Typograph's picture

Hi mike.

I saw your font way bak that you hav put a link to on TAPUZ for download.
its very nice.

So i understand that you wish to produce a hebrew font as opentype.

if you wish, you can send me your miketompson font and i will send you back an open VOLT project that supports nikud, alef-lamed ligature, and wide alternates [alef-dalet-he-kaf-lamed-memfinal-resh-tav]

these are the main functions needed in a hebrew font.

this would be a good starting point for you in developing hebrew opentype fonts.

just make sure that the font you are snding me is your standart for all your fonts are using this same format.

Glyphs req:
All Diacritic glyphs
Wide alternates

eli fried
my mail is

Syndicate content Syndicate content