Well, SBL Hebrew should be released next week.

In the meantime, check out Luc Devroye's Hebrew font links

The most popular English-Hebrew word processors are Davkawriter and Dagesh. You can get Davkawriter at www.davka.com and Dagesh at www.jewishsoftware.com . You can get additional fonts also at these sites. By doing a search you can also find free fonts.

The main difficulty with putting a Hebrew word into an English text is doing the vowels (nekudot), which are little dots under and over the letters. Only the special word processors can handle this. Israelis rarely use the nekudot.

Just to be clear: you can paste a word with nekudot into a word processor or page layout program, but to compose with vowels you will need the Hebrew program.

The main difficulty with putting a Hebrew word into an English text is doing the vowels (nekudot), which are little dots under and over the letters. Only the special word processors can handle this.

On Windows 2000 and XP, any word processor or text editor that uses standard system calls for text can position nikud (and also teamin, the cantillation marks used in the Torah). Here's a sample of the new SBL Hebrew font, from Exodus chapter 20, showing OpenType GPOS positioning in MS Word.

>OpenType GPOS positioning in MS Word.
Very interesting. Can you give me more information about Hebrew?

I knew that as of Windows 2000 you could use Hebrew MS Word, whereas before you had to have the special Hebrew Windows. I didn't mention this above as I believe Hebrew Word would be more expensive, and more than what Kyle needed.

I have been using DavkaWriter, but was looking for a better typeface than offered there. One does, I believe, need a Hebrew word processor, because of the right to left text, and expecially for the nikudot. Even Israelis regard putting in the nikudot as incredible hassle. Without a Hebrew word processor, I would think that it would be a nightmare. (Teamim are a double nightmare, but there is only one book for which they are needed, the Chumash - the Torah and Haftarah with the vowel and chanting symbols - and of course the text is already available digitally set.)

I am doing a modern commentary (in English) on a classic Jewish text that I can get on CD rom, and I am also designing and setting the book (to be published by a commercial house). Davkawriter, which I am now using, has their propriatary system, not compatible with others, but I believe the Israeli CD roms with my text will work wherever Hebrew Word works. I suspect Dagesh also has their own system. Will your new typeface, and regular fonts from Israel paste with nidudot correctly into Word?

I am also buying InDesign for the project. I noticed that there is also an InDesign ME (middle east) for Hebrew and Arabic. Will everything work OK for me pasting the Hebrew that goes into Word into regular Indesign, or do I need InDesign ME? Do I need an open type font, or will regular PS or TT fonts also work on pasting the nekudot properly?

You don't need a special 'Hebrew Word'. You can install Hebrew keyboard drivers in any version of Windows 2000 or XP, and just start typing in any recent version of Office. You can also do quite a lot just with Wordpad. The only thing you would be lacking would be Hebrew dictionary support. MS have an Office 'Proofing Tools' pack that includes Hebrew dictionary, sorting algorithm, as well as a whole pile of pretty good OT fonts from Gutmann in Israel.

I have not used DakvaWriter. Is it Unicode? The really nice thing about the MS approach is that it is all Unicode, so the text is interchangeable with other applications and across platforms.

If you want to do Hebrew typesetting, you need the ME version of InDesign. The regular version will not do right-to-left text, and it will not do GPOS mark positioning. [The current InDesign ME will support nikud positioning, but only clumsily handles teamin; I am corresponding with Winsoft, the makers of IDME, about this and hope to have it improved.]

Regarding fonts, you really need OpenType if you want marks to be positioned really accurately. However, IDME has pretty decent algorithms for mark placement, so does a reasonably good job at 'blind positioning'.

>to do Hebrew typesetting, you need the ME version of InDesign.

I am wondering whether pasting from Word (on Windows XP) into regular InDesign will work as far as nekudot, or whether I have to have InDesign ME to position the nekudot correctly. I don't need to compose in InDesign, but I do need to include nekudot as my book is for Americans, and most who know some Hebrew need the vowels. Do you know about this?

Right now I am pasting Hebrew from Davkawriter into PageMaker. Each line becomes a paragraph, and I force justify all of them, except the last line of the paragraph right justified. This works. Davka actually in the past year or so also got Guttmann fonts, so perhaps switching is not going to help me that much, for my purposes. But there was a font on one of Luc Devroye's links - 'Poeta' - that I thought might work better with my design. Your design is nice - I gather it is designed to harmonize with matching Greek and Roman fonts - a feat!

Dagesh and Davka writer are legacies of the time when regular Windows couldn't handle Hebrew. I believe they use little-used slots (is that the right word?) in the ascii table for the nekudot, and then somehow have programming to position them correctly.

However, I haven't heard about people in the US Jewish community who use these programs switching to just using Windows XP and Word. This could be because the dedicated English-Hebrew programs are designed to switch back and forth between languages easily, paste into Word etc. Also they have an on-screen Hebrew keyboard you can click with a mouse, so you don't need to know the Hebrew keyboard. Or maybe people just don't know about the new Windows options.

Regular InDesign does not support any right-to-left text, nor does it support GPOS mark positioning. Only the ME version will do these things.

If you paste Hebrew text into regular InDesign, it gets reversed, left-to-right. Text is always stored in logical order, not display order, so an application needs to be able to apply RTL directionality in order to correctly display the text.

Is InDesign ME just InDesign with an additional functionality for RTL and point placement, so that you can do everything you could with regular Indesign, plus? Does InDesign ME export PDFs that any printer can take?

Thanks!

> SBL Hebrew font

So I guess that's yours, John? It looks really nice.

If it does indeed have any "matching" fonts in other scripts, could we possibly see some samples of those too, especially intermixed and/or used in parallel?

hhp

I've just finished the SBL Hebrew (it should be released in the next couple of weeks; it will be available for free download, and I'll announce it in the appropriate forum). Next up is the SBL Greek, due out later this summer, and then early next year a tri-script version with Latin added.

Interestingly, I think, each script will have a different, appropriate ductus, but the weight and detailing will harmonise. Proportionally, the height of the majority of the Hebrew letters corresponds to the Latin smallcaps. The Latin is a kind of Dutch oldstyle, and the Greek is related to the Clio Greek that I presented at ATypI in Rome -- and hence to the Granjon St Augustin Greek --, although that will be the italic and there will also be an upright version influenced by the later Dutch Greeks. Fun project.

Is InDesign ME just InDesign with an additional functionality for RTL and point placement, so that you can do everything you could with regular Indesign, plus? Does InDesign ME export PDFs that any printer can take?

So far as I can tell, InDesign ME does everything that regular InDesign can do, plus all the other stuff. Frankly, I think it is so vastly superior to regular InDesign that I'm unlikely to purchase an upgrade to the latter again, but will simply wait for the next version of InDesign ME.

I have not sent any PDFs from InDesign ME to a printer, but I have generated PDFs without problems and have not seen anything that suggests they would not work as well as those from regular ME.

> the height of the majority of the Hebrew letters corresponds to the Latin smallcaps

Assuming the smallcaps are somewhat taller than the lc x-height, that makes wonderful sense.

If/when you make a matching Cyrillic, where would/will you put its "x-height"?

hhp

Yes, the smallcaps are taller than the lowercase x-height. I tend to make my smallcaps quite a bit taller than my x-height, presuming the height of the caps allows, because I so often end up designing Cyrillic alongside my Latin. Cyrillic smallcaps need to be noticeably taller than the lowercase x-height, because so many of the forms are easily mistaken for lowercase letters.

Cyrillic x-height = Latin x-height unless I'm making an Old Church Slavonic font. Like it or not -- and I know you don't like it --, the post-Petrine civil type is modelled on Latin proportions, and I'm not in the script reform business.

> I'm not in the script reform business

:-/
Making the vertical proportions different isn't "reform", come on. It simply reflects the desire to make the two scripts of closer apparent size. "Cyrillic x-height = Latin x-height" makes the former look smaller - and for what gain?

How can your logic be different than for the Hebrew?

hhp

Well, you're going to hate this, but like most type designers working with Latin and Cyrillic I use the same glyphs for cognates in both scripts, e.g. a e o p c y x.

Hebrew contains no Latin & Cyrillic cognates. Also, there is an established tradition in scholarly typography -- the target of the SBL fonts -- that Hebrew letters are taller than the Latin x-height but shorter than the caps. This has been the norm for a few hundred years. Interestingly, in Israel, the developing norm for mixed modern Hebrew and Latin text is for the Hebrew to match the x-height of the Latin.

At the moment, there are no plans to extend the SBL fonts for Cyrillic, although it might happen one day. Syriac and Ethiopic are more likely to be added first, due to the nature of the texts under study.

> I use the same glyphs for cognates in both scripts

Well, yes, I dislike that very much - it violates linguistic realities. But really, even a preference for such cognates doesn't have to preclude making the vertical proportions different. You might say it's too much trouble scaling up the Latin cognates to a larger x-height, since you'd have to then make them lighter too; but with a little bit of technique it could be a breeze - here's what I'm thinking:

If you have a weight axis in your font (and one with a light extreme [sufficiently] beyond what you're actually releasing in the Latin), what you could do is simply scale up those cognates for Cyrillic use, but use a slightly lighter weight to match the Latins.

> the developing norm for mixed modern Hebrew and Latin
> text is for the Hebrew to match the x-height of the Latin.

Yes, I've noticed - and it sucks.
I wonder why there's been such a regression. I have to suspect an unhealthy infatuation with the West.

--

I guess what you're saying is that you're simply following established norms, in each script. Norms are important, but functionality itself even more so, no? I mean, if the contemporary Hebrew method were to become the norm, would you then start following it, even though you knew it skewed the apparent sizes? And if you admit that vertical proportions do affect functionality (something the old Hebrew designers realized*), then why not push Cyrillic towards that too? Do you fear you'll be blacklisted by the Russian font mafia or something? ;-) Users don't even consciously realize what's going on, so they'd never complain; but under the surface they'd certainly benefit from better functionality.

*
Here's an example of smart Latin/Hebrew proportions/color_matching, from an old Lino metal catalog:

(Thanks again, Gerald.)

hhp

Oh, I forgot to ask: And what about your SBL Greek x-height? :-)

hhp

The Greek x-height will probably be pretty close to the Latin, but may vary slightly as optically befits. I don't treat Greek lowercase letters as Latin cognates unless obliged to by the design brief. In fact, as you know from our previous discussion on Typographica, I prefer to treat Greek as a Middle Eastern script

>If you paste Hebrew text into regular InDesign, it gets reversed, left-to-right. Text is always stored in logical order, not display order, so an application needs to be able to apply RTL directionality in order to correctly display the text.

John, I downloaded the trial version of InDesign, and tested pastes from the English-Hebrew word processor Davkawriter. It has dual export mechanisms - one to Hebrew programs and one to English programs. The export to English programs goes into regular InDesign fine - in fact better than into PageMaker. It makes each line a paragraph, and I guess reverses the order of storage. (It also has a twin sets of Hebrew fonts for pastes into Hebrew and English programs.) So I can get by with regular InDesign at this point. I'm not sure, but I suspect that the reason people in the U.S. are still using Davkawriter and Dagesh, rather than Word with XP and a Hebrew keyboard installed is because of these kinds of functionalities.

Eek! Well, if you don't mind rendering your Hebrew text backwards in order to make it look okay, I guess that's okay, but I'd say this 'functionality' is a hack. These days, you shouldn't need to destroy text encoding in order to get correct text display.

>you shouldn't need to destroy text encoding in order to get correct text display.
I agree that a more straightforward mechanism is better. However, the placement of vowels is quite good from Davkawriter (it also does the cantillation symbols). They have worked on it for years to get it this good. I would prefer InDesign ME, but it will cost me a thousand dollars more than my upgrade to regular InDesign from PageMaker, for no immediate benefit. When I get richer I will upgrade to ME from regular - a less steep jump.

John,

the sbl hebrew is your design?

just saw that. i need my coffee - But at a glance: there's a resemblance.... the "Barcelona" script ( a Sephardic alphabet though of a more formal, festive nature then the "Sfaradi" script).

David Hamuel

Do you mean the script of the Barcelona Haggadah? I looked at several different Sephardic manuscript examples when I began drawing the type, and it is quite likely that reproductions from the Barcelona Haggadah were among them, since it is so famous. Of particular importance to the SBL Hebrew design was another Haggadah in the British Library (Ms. OR. 1404). A page of this is reproduced in Ismar David's _The Hebrew Letter_, if you're interested. I also looked for a long time at the Soncino types, although SBL Hebrew is much smoother and closer to the manuscript style.

just "rare" to see Hebrew typeface. i speak/write/read etc... Hebrew. and designed Hebrew typrfaces.

>designed Hebrew typrfaces
David, I am doing a modern guide to Pirkei Avot, in English with the Hebrew text also. I would be very interested in your take on the merits using DavkaWriter and Dagesh and pasting into InDesign vs using InDesign ME. I know my publisher is also interested in the best way to deal with Hebrew in English texts for the future, given the changing technology. So if you have a take on the technology used in Israel vs DavkaWriter or Dagesh, I'd be very appreciative to know it. Above you will see my discussion with John on this.

What was the lae great Israeli type designer Gutmann's first name?

Follow the lead http://mail.typophile.com/node/76573

I can see that the letter Lamed extends above the top line by more than accents on capital letters - not even supported by all Latin fonts - do, so that argues for the scholarly tradition. After all, with hot metal, as opposed to TrueType, if you want leading, you put in leads; the tradition has been to fill the type body as much as possible, although slight compromises were accepted when lining type was brought in.

Cyrillic is so similar to Latin, even having lowercase cognates as well as uppercase, that it would seem to make sense for the same typeface to have the same x-height in both scripts, even if the tendency had been for Cyrillic typefaces to tend to a larger x-height.

Of course, an opposite approach would be to indeed use a larger x-height for Cyrillic, and use the glyphs from Latin small capitals for their cognates among Cyrillic lower case. I'm sure, though, that this would be regarded as even worse, and by a considerable margin.

As for Greek, it doesn't have lower-case cognates (unless severely Latinized), just upper-case ones.

@John Hudson:
In fact, as you know from our previous discussion on Typographica, I prefer to treat Greek as a Middle Eastern script :-)

Are you sure you're not in the script reform business?

Of course, it certainly is true that Greek has Phoenician roots, but then why not take Latin back to its Etruscan roots while you're at it?

As we've seen in the other thread on Greek typography, my attitude to Greek is that it missed the Carolingian boat, and so its upper-case looks like it belongs to the same great tradition as Latin and Cyrillic, and its lower-case is jarringly alien. But I also feel that trying to Latinize Greek is the wrong approach, and instead basically one has to leave it to the Greeks themselves to harmonize their upper-case and lower-case into a unified whole. Which may take centuries.

Not that there's anything wrong, in the interim, with designing typefaces which take new and original approaches to such an integration. Eventually, some features from some of those attempts may stand the test of time, and contribute to the ultimate resolution.