Source Sans Hebrew

1,140 views
Skip to first unread message

Ben Nathan

unread,
Nov 17, 2015, 8:45:50 AM11/17/15
to Google Fonts Discussions
Hi All !

I would like to share with you my latest"Latin to Hebrew font - Source Sans Pro
It's a real challenge and i've been sketching different approaches for the past weeks.

I started with trying to create a round hebrew sans serif (the first sketch).
This version works fine with the latin but unlike the latin it lacks personality and looks too generic (a mix of the hebrew fonts Oron / Narkis Tam / ...)

My new sketch is less round, I wanted to use the straight lines as much as I can with a curve as a contract. 



Source Sans Latin + Hebrew:



 

Some text:



I got more work ahead of me:
I think some letters are not right - ק, מ, ת, פ

I'll post updates on this project and would love to hear your thoughts.

Cheers!
Ben


Ben Nathan

unread,
Nov 25, 2015, 12:54:55 PM11/25/15
to Google Fonts Discussions
Quick update:

Almost dont with the regular, and still working on the Bold


More updates soon! would love to hear your thoughts


Cheers,

Ben

Dave Crossland

unread,
Nov 25, 2015, 1:06:30 PM11/25/15
to googlefonts-discuss, Paul Hunt
Hi Ben

I'd love to hear what Paul thinks! :)

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.
To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/0b28fd0e-b0ac-46e2-873e-c365a527deba%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Cheers
Dave

Meir Sadan

unread,
Nov 25, 2015, 1:35:52 PM11/25/15
to Google Fonts Discussions
Hey Ben,

Big improvement! I actually liked your first פ pe and ף final pe (with the round tops) better.
And I think the ג gimel still looks a bit strange in the ensemble – maybe try a different leg shape?
There's something weird in the bold ס samekh – the inner outline looks a bit boxed comparing to the outer outline which is rounder - the sample isn't too big so I can't really judge.
Also the ל lamed may need to be more similar to the ק qof, or just a little less steep leg?
It's exciting to see everything coming together :>

Paul Hunt

unread,
Nov 25, 2015, 10:09:01 PM11/25/15
to Google Fonts Discussions, da...@lab6.com
Thanks for looping me in, David. I guess I hadn’t been getting the updates before. I think that before I can really try to evaluate the design too much, I would like to be able to do some text setting tests. As such, It would be nice for this project to be getting regular pull requests so that I can practice building the fonts and running my own tests. For example, I would like to try to ascertain as to whether the Hebrew character height is tall enough compared to the Latin. Similarly, I would like to evaluate the color by setting text blocks and comparing/contrasting with the Latin. Once I am able to do these things, I can hopefully provide more in-depth input into the design.

P

Dave Crossland

unread,
Nov 25, 2015, 10:11:13 PM11/25/15
to Paul Hunt, Google Fonts Discussions
Hi

Ben, are you ready to make pull requests on the source files? :) 

On 26 November 2015 at 09:56, Paul Hunt <phunt...@gmail.com> wrote:
Thanks for looping me in, David. I guess I hadn’t been getting the updates before. I think that before I can really try to evaluate the design too much, I would like to be able to do some text setting tests. As such, It would be nice for this project to be getting regular pull requests so that I can practice building the fonts and running my own tests. For example, I would like to try to ascertain as to whether the Hebrew character height is tall enough compared to the Latin. Similarly, I would like to evaluate the color by setting text blocks and comparing/contrasting with the Latin. Once I am able to do these things, I can hopefully provide more in-depth input into the design.

P



--
Cheers
Dave

Paul Hunt

unread,
Nov 25, 2015, 10:20:02 PM11/25/15
to Google Fonts Discussions, da...@lab6.com
If so, I would greatly appreciate first running UFOs through the UFO normalizer tool before making commits to the project.

https://github.com/unified-font-object/ufoNormalizer

Ben Nathan

unread,
Nov 26, 2015, 4:17:18 AM11/26/15
to Google Fonts Discussions
Hi Paul, Dave and Meir !

Thanks for the input and i'll create the UFOs. 
I still have some more on this type family, so i'll finish this sketch (on the weekend) and upload a version you could play around with

Cheers!
Ben

Ben Nathan

unread,
Nov 29, 2015, 5:05:19 PM11/29/15
to Google Fonts Discussions
Quick update - made good progress over the weekend, did 2 main things:
- made my x-height taller
- made it narrower

* Paul i'll send the normalized UFO tomorrow

some pics:


On my to do list:


- Finalised the light style - some letter adjustments, proper spacing, nikud.

- Design the Bold master.


More updates soon.


Cheers!

Ben

Ben Nathan

unread,
Nov 30, 2015, 9:18:47 AM11/30/15
to Google Fonts Discussions
Hi Paul, attached is the normalized UFO.
Source Sans Pro-Light_normalized.ufo.zip

Dave Crossland

unread,
Nov 30, 2015, 7:51:45 PM11/30/15
to googlefonts-discuss, Paul Hunt

Hi Ben

Please could you send this as a pull request via github? :)

Cheers
Dave

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

Ben Nathan

unread,
Dec 1, 2015, 1:03:17 PM12/1/15
to Google Fonts Discussions, da...@lab6.com
I think I did it (i'm new to github) :)

I would download the Glyphs file if you wish to see and test the font.

Paul Hunt

unread,
Dec 1, 2015, 7:29:02 PM12/1/15
to Google Fonts Discussions
I plan on showing this (the ExtraLight weight) to my colleagues tomorrow during our meeting where we look at current design work. I will post back with feedback afterwards.

P

Ben Nathan

unread,
Dec 2, 2015, 11:12:20 AM12/2/15
to Google Fonts Discussions
Hi Paul, thanks, waiting to hear your team's thoughts.

Today's update - 
I kept on working on the Light and started designing the Bold master, I did the interpolation to match Source Sans (6 styles overall).
Still got work on the Bold master, but I feel it's in the right direction.


I'm attaching a PDF test I did and here'e an image of the family: 


Cheers!
Ben
Hebrew4.pdf

Paul Hunt

unread,
Dec 2, 2015, 5:34:59 PM12/2/15
to googlefon...@googlegroups.com
Okay, here is the consensus, please note that none of us are Hebrew typography experts per se:
The current design feels very rudimentary and in general not a great match for Source Sans (SSP). In order to get to a design that will be a good match for SSP, a considerable amount of work would need to be done, beginning with establishing a design direction that we feel will be a good match for SSP. I am not sure how invested you are in putting in the amount of work that we feel would be required to arriving at a design that we would want to put the Source name on, or if you want to just complete the design and be done with it. In the latter case, I don’t think that we would be comfortable making the design, as it currently stands, a part of the Source brand. If you are interested in making the invemstment into doing what we feel to be the right thing for the design, please let me know and we can devise some first steps to get things started in the right direction.

Best regards,

Paul Hunt

Liron Lavi Turkenich

unread,
Dec 4, 2015, 4:42:30 AM12/4/15
to Google Fonts Discussions
Hi Paul, 

Could you please elaborate a bit on why you (and your team) don't feel that the proposed Hebrew is a good match for SSP?  
I will share here my opinions on it: I think that those days there are many many "sans"-monolinear Hebrew typefaces that are very generic and look too similar to one another. I feel that in Ben's design he managed to incorporate some features that makes it look much more personalised than other typefaces. 
It is always a challenge to harmonise Hebrew with Latin because of the basic different letter structure, and usually what happens is that the Hebrew looks to stiff and square compared with the more flowing Latin. Here- Ben added "friendlier" and warm features such as the small curve in the Vav and Yod, the middle of the Peh and more, which make them go well together. 
I also think it is looking very well next to the Latin in terms of color (on a block of text). 

So if you could provide some more explanations on your thoughts it would be great. I think this is a valid discussion and a learning opportunity for all of us. 

Have a lovely day,
Liron

Omer Ziv

unread,
Dec 4, 2015, 3:22:31 PM12/4/15
to Google Fonts Discussions
Thanks Liron for your thoughtful response!
Hebrew really is very different from English - our letters are square and heavy :) It seems to me that Ben did a lot to try and match SSP's designs.

Paul, I'm sorry the process wasn't more inclusive, and would love to hear more about where you feel Ben's work strays from your original intentions.
I've seen some very interesting, constructive conversations on other threads in this group - and even if Ben's work doesn't end up published as a part of SSP (although, I really hope it will be), your thoughts would be greatly appreciated!

Cheers!
Omer

Paul Hunt

unread,
Dec 4, 2015, 3:45:31 PM12/4/15
to Google Fonts Discussions
Of course I can share a more in-depth analysis of why I, as the designer of Source Sans, do not feel that the current design is not a good match for SSP. To start off with, the brief for SSP was to create a clear, simple sanserif typeface that would be pleasing in a user interface framework and would also work in running text on screens. To achieve these goals, I wanted to stick fairly close to traditional proportions, but with gently compressed forms. Additionally, I tried to reduce each letterform to it’s most basic structure while still keeping some details that added warmth and aid in legibility, eg the flick on the outstroke of l adds a bit of personality, but on a functional level it serves to differentiate between I (capital i) and l (lowercase L). In a way what I tried to do is to create a type that traces the bones of a traditional book face–the prototypical forms–and treats them with a touch of naïveté that lends the appropriate amount of warmth and character to the design.

In looking at the current state of Ben’s design, I do not see the same considerations influencing the design. The proportions feel like they have been made to adhere too strictly with the proportions and rhythm of the Latin, which I feel is probably a mistake to try to do so closely. The designer has tried to reduce the forms to something very simple, but I feel that he has gone too far an they no longer have much soul: the design feels very ‘ball-and-stick’ to me, a term that we use in Latin type design to refer to designs that are overly constructed, as if they they were made by piecing geometric balls and sticks together. For example, look at the Alef: it’s literally 3 straight lines arranged in together. You can say the same of many letters in the Latin: k, v, w, x, z. I could have also reduced l, t & y to straight stick forms, but I went with more writerly forms to give the design a greater sense of being informed by the human hand. I feel that Alef is in the same category as my latter examples of construction, some humanizing details can and should be kept to give life to the design so that it does not feel as sterile and rudimentary as it does now. Likewise, I feel that Yod and Vav are overly simplified and would benefit greatly from a treatment that helps these shapes to relate better to their more traditional, calligraphic forms.

Lastly, I find letterforms as designed to be a hodgepodge and not well harmonized. Perhaps the main thing that jumps out and bothers me is that the diagonals are all over the place. In particular, I am disturbed by the discrepancy between the Lamed and Qof. Lamed feels again like a stick letter, and Qof feels like it has been too rationalized away from its prototypical form in a way that is more appropriate for a display design and less so for a text design. Ideally, I would like to see diagonal forms of Tet, Lamed, Samekh, Ayn, Qof & Shin harmonized. (I was going to try to link to the Toby lettering book page for ספרדי, but the flickr link seems to be broken: https://www.flickr.com/photos/yaronimus/314322954/)  But the diagonals are only one area where I feel that harmonization is lacking, there are other details that should be handled in a more consistent way. I feel like overall effect of the deficiencies listed above is that there is not enough tension in the design to make it look interesting and give it life, as I hope that I have been able to do successfully with the Latin, Greek, and Cyrillic.

This summarizes my concerns and I do not feel that these can be overcome with a few small tweaks. The design needs to be rethought and redrawn in many places so that the spirit and use cases align much more closely to Source Sans pro instead of focusing on the wrong things that may questionably make the current design feel superficially like a match for SSP.

Dave Crossland

unread,
Dec 4, 2015, 5:38:50 PM12/4/15
to googlefonts-discuss

Hi Paul!

Thanks for taking time to write up your in depth analysis :) 

I want to apologies for not keeping on top of this project, as Ben did agree at the outset to discuss the project with you, and you suggest using the Github issue tracker at https://github.com/adobe-fonts/source-hebrew-sans/issues/1 for this, but I didn't really keep track that the discussion didn't happen there and only realised that I hadn't heard from you regarding this when I looped you in via CC'ing you here. I wish I had forged a more in-depth discussion of the design direction earlier. 

I would like to suggest that to move this forwards, we three (you and Ben and I) schedule a hangout-on-air early next week to discuss the design in detail. (I'll be in India so early morning or early evening PST is best for me, and Ben is in Israel and so in-between us.)

I think there are two possible outcomes ahead: the first, and my preference, is that you and Ben work together such that you are happy with his design; I wonder that this may involve wider native-reader-type-designer review of the design to comment on how well the two scripts are working together. As you know I have a number of Israeli designers working on various projects (including Liron - thank you Liron for your email earlier in this thread :) who I hope will be happy to comment. 

The second outcome is that Ben renames his project to not use the Adobe trademark and OFL RFN name. It might be wise to do this as a next step, as a provisional 'code name' for development purposes, anyway, even if we arrive at the first outcome in the end :)  With this Plan B, it may be that Ben then starts on a new design that eventually becomes the Source Sans Hebrew. 
 
--
Cheers
Dave

Paul Hunt

unread,
Dec 4, 2015, 5:47:07 PM12/4/15
to Google Fonts Discussions, da...@lab6.com
David,

I’m open to having a discussion on this next week. Robert Slimbach (designer of Myriad Hebrew and my lead designer) is wanting to have some discussion with me about design direction, but that will likely not happen until Wednesday next week, so I do not know how much I will have to say specifically regarding direction on a call earlier in the week, but can try to talk about things generally. If you think this will be helpful, let me know and I shall try to make time for this. Otherwise, I am currently in the process of relocating to Austsralia and will be taking some time off work until my work visa comes through (hopefully in early January). Depending on if we want to continue with developing a Hebrew counterpart for SSP, we may need to put that work on hold until after my visa is approved and I can get back to work.

Regards,

Paul

Meir Sadan

unread,
Dec 8, 2015, 5:23:02 AM12/8/15
to googlefon...@googlegroups.com
Hi Paul, Ben and all,

I'm glad this discussion is taking place in this forum so that we can learn from this process and have some of the Hebrew designers weigh in as well, as Liron and Omer did. I see Paul's points in general about several of the design details, but I think that the base forms for the Hebrew script can't be based (at least not entirely) on a "traditional" Sepharadi script, as referenced by Paul. Hebrew typography has changed dramatically in the past ~60 years, mainly thanks to the work of pioneer type designers like Friedlaender, David and Narkiss, who looked past the "exiled" Sepharadi model of the script to more ancient forms (such as the Aleppo codex and Judean stone enscriptions) to invigorate Hebrew type and find its original elegance.
In my opinion, Ben's work is largely influenced by Narkiss and Oron's work with Hebrew type that lives in tandem with the Latin, and while there may be several details worth looking into in terms of some of the details (the top and bottom arms of the א alef, for instance), I think that in whole the script is overall in the right direction. I suggest Paul/Adobe work with Ben on refining specific details and working with the right references (Narkiss/David/Oron perhaps) to better adapt the Hebrew to the overall design direction of SSP.

Meir

Dave Crossland

unread,
Dec 8, 2015, 5:37:27 AM12/8/15
to googlefonts-discuss

Hi Paul 

Sorry for my delay in responding

I'd love to chat with you and Ben before or after your chat with Robert; before might be better, as I've spoken with Ben and he would like to explain the deeper thinking behind the existing design direction (which he's validated by speaking with other type designers in Israel the last few days) and it could be cool for you to hear that and then discuss with Robert. 

I'm in India and Ben is in Israel and you are in California, so early today would be ideal - I'll be online and should have good bandwidth for a call :) 


--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

For more options, visit https://groups.google.com/d/optout.



--
Cheers
Dave

Adam Twardoch (List)

unread,
Dec 8, 2015, 8:01:15 AM12/8/15
to googlefon...@googlegroups.com
Meir, 

I echo your points. 

Source Sans Pro is a close cousin to Morris Fuller Benton's News Gothic, especially evident when comparing to the Bitstream version: 

Source Sans has a slightly smaller x-height and somewhat more regularized strokes, but it very well retains the American neo-grotesque curve tensions. To my eye, the American neo-grotesque sits in the sweet spot between geometric and humanistic sanserifs. If you think of letterform skeletons, geometric sanserifs are drawn with the use of tools, while humanistic sanserif skeletons are "written". The American neo-grotesque is actually "hand drawn", it's the most architectural of all forms, where technology and the human hand contribute equally. 

This is what I always see in News Gothic or The Font Bureau's Benton Sans, and it's what I see in Source Sans Pro, with the addition that Source Sans goes just so slightly more towards the technical. Two things that are most recognizable about Source Sans is that the letterform elements terminate or meet at a very diagonal angle, 45 degree (ExtraLight) to 33 degree (Black), and that towards the black, the curves become rather rhomboid, pointy at the ends. 

One could say that Source Sans, as much as it can, avoids the horizontal emphasis (through the numerous diagonal connections), and avoids squareness at all cost (by making the ovals elliptic almost to the point of pointyness). 

Also, towards the black, the elements terminate a bit early, leaving wide open gaps. 

To me, these are the essential flavor elements of Source Sans. 

I do agree with Paul's view that Ben's Hebrew design is more "ball and stick". Out of all models that Meir mentioned, I would recommend looking at Ismar David's most. In fact, intuitively, I can say that there is a *lot* in David Hadash Formal which could be taken as inspiration for a Hebrew companion to Source Sans, or more broadly even, a Hebrew take on an "American neo-grotesque" style. 


Of course I don't mean the stroke modulation in David Hadash Formal, but the overall structure, and certain structural decisions. In David Hadash Formal, Ismar David had sought to "avoid the horizontal emphasis", and "avoid squareness at all cost". The underlying skeleton of David Hadash Formal has a lot of pointyness in the ovals, and a lot of diagonals. If that model were rationalized and you’d try to redraw its skeleton via an "architect’s hand", I think you could arrive at a very decent Hebrew that could work alongside Source Sans very well. 

I'm not saying that you should go the David model all the way. David Hadash Formal is a bit too angular and too fluid. And of course it would need to be "straightened" and probably made a bit more square (but not nearly as square as Ben's). There are many structural solutions in David Hadash Formal which I personally feel could be made into a good match with a neo-grotesque than in Ben's design, which is, I agree with Paul, far too simplistic (though it may work as a standalone design, but not as a Hebrew companion to Source Sans). 

But — well, I don't know nearly enough about Hebrew type design to claim any kind of real expertise. 

Many thanks,
Adam


 





On 08 Dec 2015, at 11:23, Meir Sadan <meir...@gmail.com> wrote:

Hi Paul, Ben and all,

I'm glad this discussion is taking place in this forum so that we can learn from this process and have some of the Hebrew designers weigh in as well, as Liron and Omer did. I see Paul's points in general about several of the design details, but the think that the base forms for the Hebrew script can't be based (at least not entirely) on a "traditional" Sepharadi script, as referenced by Paul. Hebrew typography has changed dramatically in the past ~60 years, mainly thanks to the work of pioneer type designers like Friedlaender, David and Narkiss, who looked past the "exiled" Sepharadi model of the script to more ancient forms (such as the Aleppo codex and Judean stone enscriptions) to invigorate Hebrew type and find its original elegance.

In my opinion, Ben's work is largely influenced by Narkiss and Oron's work with Hebrew type that lives in tandem with the Latin, and while there may be several details worth looking into in terms of some of the details (the top and bottom arms of the א alef, for instance), I think that in whole the script is overall in the right direction. I suggest Paul/Adobe work with Ben on refining specific details and working with the right references (Narkiss/David/Oron perhaps) to better adapt the Hebrew to the overall design direction of SSP.

Meir


On Saturday, December 5, 2015 at 12:47:07 AM UTC+2, Paul Hunt wrote:
David,

I’m open to having a discussion on this next week. Robert Slimbach (designer of Myriad Hebrew and my lead designer) is wanting to have some discussion with me about design direction, but that will likely not happen until Wednesday next week, so I do not know how much I will have to say specifically regarding direction on a call earlier in the week, but can try to talk about things generally. If you think this will be helpful, let me know and I shall try to make time for this. Otherwise, I am currently in the process of relocating to Austsralia and will be taking some time off work until my work visa comes through (hopefully in early January). Depending on if we want to continue with developing a Hebrew counterpart for SSP, we may need to put that work on hold until after my visa is approved and I can get back to work.

Regards,

Paul

-- 
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-discuss+u...@googlegroups.com.

To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

Dave Crossland

unread,
Dec 8, 2015, 12:41:33 PM12/8/15
to googlefonts-discuss, Lasse Fister
Hi

Adam, thanks again sharing your thoughts here :) 

On 8 December 2015 at 18:31, Adam Twardoch (List) <list...@twardoch.com> wrote:
In fact, intuitively, I can say that there is a *lot* in David Hadash Formal which could be taken as inspiration for a Hebrew companion to Source Sans, or more broadly even, a Hebrew take on an "American neo-grotesque" style. 

And since David Hadash has been OFL'd (and integrated with Gentium) at https://github.com/meirsadan/david-hofshi then perhaps another attempt at a Source Hebrew could do this more directly than just inspirationally. 

This kind of transformation could be an interesting pilot for Metapolator... :) 

--
Cheers
Dave

Paul Hunt

unread,
Dec 8, 2015, 1:49:10 PM12/8/15
to Google Fonts Discussions
I think that Adam nailed a lot of things that I was trying to allude to in my posts from last week. Namely that there is a lot of evidence of the human hand in SSP which I don’t see as readily in the shapes of this Hebrew design. The tricky thing on this project (as with any script matching project) is trying to match the voice and design language insofar as is appropriate. Let’s talk about this more if we can set up a time that works for the involved parties. Let me know when you think we can all have a chat.

Dave Crossland

unread,
Dec 8, 2015, 2:33:47 PM12/8/15
to googlefonts-discuss, Paul Hunt, Ben Nathan

Hi

Would anytime for say 30 minutes between 9am and 11am tomorrow (wednesday 12/9) for you both? :)
 
--
Cheers
Dave

Paul Hunt

unread,
Dec 8, 2015, 5:08:18 PM12/8/15
to Dave Crossland, googlefonts-discuss, Ben Nathan
PST?

Dave Crossland

unread,
Dec 9, 2015, 1:22:42 AM12/9/15
to Paul Hunt, googlefonts-discuss, Ben Nathan


On 9 December 2015 at 03:38, Paul Hunt <ph...@adobe.com> wrote:
PST?

Yes :)

Paul Hunt

unread,
Dec 9, 2015, 3:31:56 AM12/9/15
to Dave Crossland, googlefonts-discuss, Ben Nathan
OK. I shall make myself available then. 

P

Sent from my iPhone

Paul Hunt

unread,
Dec 9, 2015, 3:32:31 AM12/9/15
to Dave Crossland, googlefonts-discuss, Ben Nathan
At 9:00 AM PST, that is.

P

Sent from my iPhone

Paul Hunt

unread,
Dec 9, 2015, 11:47:04 AM12/9/15
to Google Fonts Discussions, da...@lab6.com
I’m online. I just need to know how you want to sync up.

Dave Crossland

unread,
Dec 9, 2015, 11:54:39 AM12/9/15
to Paul Hunt, Google Fonts Discussions

Dave Crossland

unread,
Dec 9, 2015, 2:09:14 PM12/9/15
to Paul Hunt, Google Fonts Discussions
Hi

We moved the call to Skype due to technical issues, and I think it went very well :) 

I noted the following action items to move the project forwards:

1. Ben to post the test char strings he'll use to sketch concepts

2. Paul and team to confirm the strings

3. Ben to sketch ideas and post PDFs here, and/or glyphs files on his repo

4. Paul will review via email on this thread

5. Once a sketch PDF is confirmed as a concrete design direction, then I'll work with Ben to get a smoother direct collaboration on font files themselves going between Ben and Paul via Github 

Cheers
Dave

Ben Nathan

unread,
Dec 9, 2015, 2:41:05 PM12/9/15
to Google Fonts Discussions, da...@lab6.com
Thanks Dave!

We had a great talk today, here's the Hebrew test word: אוסטרליה (Australia) or/and כורדיסטן (Kurdistan)

Cheers,
Ben

Paul Hunt

unread,
Dec 9, 2015, 6:05:32 PM12/9/15
to googlefon...@googlegroups.com
Thanks Ben, I just had a chance to meet with Robert Slimbach now and review the things that we talked about this morning and also have a bit of our own discussion. The key letters that we would like to see in the first phase should include אגהוספש (whatever you do beyond that is up to you, but let’s try to keep it to ~10-12 glyphs max to begin with). Robert and I looked at your ימים ולילות typeface and in a way what we almost want is a sanserif version, but following the model of Hatzvi (attached below) more closely in terms of construction of א and צ. We are looking forward to your interpretation of the guidance I gave you this morning and excited to see where this will now go. Let me know if you have any questions.

Thanks,

Paul
IMG_4345.JPG

Dave Crossland

unread,
Dec 10, 2015, 2:52:15 AM12/10/15
to googlefonts-discuss

On 10 December 2015 at 01:11, Ben Nathan <hafo...@gmail.com> wrote:
here's the Hebrew test word: אוסטרליה (Australia) or/and כורדיסטן (Kurdistan)

On 10 December 2015 at 04:35, Paul Hunt <phunt...@gmail.com> wrote:
the first phase should include אגהוספש 

I'm curious what the reasonings are behind these strings :) 

Paul's seems to have more variety in forms which seems like an advantage to me

Meir Sadan

unread,
Dec 10, 2015, 4:56:03 AM12/10/15
to Google Fonts Discussions, da...@lab6.com
Hey,

Re: test words: normally what we try in terms of control characters is include one letter of each "letter group", this includes -
Yod group - י, ו, ה, ח, ן, ך, ץ, ת, ף
Flat top group - ד, ז
Backstroke group - ב, כ, פ, מ, ם
Gimel Nun group - ג, נ
Diagonal group - א, צ
Round base group - ס, ש, ט
Qof Lamed group - ל, ק

Since it's almost impossible to find proper words that contain letters from each group it's usually just a random bunch or a word that contains letters from most groups.

Re: Hatzvi as a model for the Hebrew – I'm strongly against using it as a base model, as its base shapes are considered very "displayish" in Hebrew. David or Narkiss are a much better fit (and not very far from Hatzvi in any case).

Meir

Dave Crossland

unread,
Dec 10, 2015, 5:01:44 AM12/10/15
to Google Fonts Discussions

On 10 December 2015 at 15:26, Meir Sadan <meir...@gmail.com> wrote:
Re: Hatzvi as a model for the Hebrew – I'm strongly against using it as a base model, as its base shapes are considered very "displayish" in Hebrew. David or Narkiss are a much better fit (and not very far from Hatzvi in any case).

In our call, Paul called out the design so far as having forms that strike him as universally towards display, and I took away 3 key goals for the brief: 1. a good UI typeface for web apps, 2. a good long-form reading typeface, and 3. something warm and familiar, 'mid century modern', and not contemporary (in contrast to Ben's claim that the design so far is contemporary.) 

While Hatzvi may also be displayish, I wonder that since we worked with Monotype to release an OFL version of David Hadash, starting with Meir's David Hofshi and making a "sans" version could be good. Paul, what do you think?

Paul Hunt

unread,
Dec 11, 2015, 5:54:48 PM12/11/15
to Google Fonts Discussions, da...@lab6.com
Dave:

Regarding your point 3) I would characterize this differently: of course we want a typeface that is contemporary and seeks to push the envelope in certain ways. But in order to harmonize with the styling of the Latin, ‘a whif’ of retro styling may be appropriate.

Meir:

Can you say what it is about the forms of Hatzvi that you find ‘displayish’? Is it the specific forms, or the styling, or something else?

Meir Sadan

unread,
Dec 12, 2015, 6:00:01 PM12/12/15
to Google Fonts Discussions
Hey Paul,

Hatzvi is a Hebrew display typeface. Since its release it was used mainly for advertisements and signage, and is hardly used in anything other than in retro 1950s style ads nowadays.
Context aside, Hatzvi's letteforms share several attributes with David Hebrew's, but its monolinearity, sharp angles and wide construction make it unsuitable for body text. In essence, it's considered monumental and "dramatic".
As I mentioned, David Hebrew shares Hatzvi's rationale in terms of construction, but does a better job in use range and versatility.

Meir

Dave Crossland

unread,
Dec 13, 2015, 6:31:43 AM12/13/15
to googlefonts-discuss

Thanks for the clarification Paul :)

Ben, I suggest to take Meirs idea on board and see what Paul and team think of what you sketch up :)

Paul Hunt

unread,
Dec 13, 2015, 1:52:31 PM12/13/15
to googlefon...@googlegroups.com
Dear Meir,

Thanks for sharing your observations. I’ve asked Ben to look at Hatzvi only in terms of letter construction and simplification and then not too, too closely. we are not trying to revive Hatzvi here, so we are just using it as a reference point, but the final design will maybe hopefully have only a bit of the spirit of Hatzvi without looking like it in terms of fine details. I feel like we have a solid design direction, let's wait and see what Ben comes up with before making any more judgements. 

Thanks,

Paul

Sent from my iPhone
--
You received this message because you are subscribed to a topic in the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/googlefonts-discuss/ZD7K2JiCIdw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to googlefonts-dis...@googlegroups.com.

To post to this group, send email to googlefon...@googlegroups.com.

Ben Nathan

unread,
Dec 14, 2015, 2:48:01 PM12/14/15
to Google Fonts Discussions
Hi All,

Attached is a sketch in a new direction.
- wider Hebrew proportions
- rounder shapes
- added straight roofs in some places (to counter the round shapes and make the font "more Hebrew").

I also stared drawing a Hatzvi style direction, but let's see if this one is in the right direction.

Thanks!
Ben
he.pdf

Paul Hunt

unread,
Dec 20, 2015, 9:14:45 PM12/20/15
to Google Fonts Discussions
Dear Ben,

Sorry for the delayed response. I’ve been trying to decide what to say about this. But instead it occurs to me that I would like to see your version in the Hatzvi direction in comparison and then tell you what I like/dislike about both.

Best,

Paul

Ben Nathan

unread,
Dec 22, 2015, 12:00:49 PM12/22/15
to Google Fonts Discussions
Hi Paul !

Attached is my second sketch, it's more "rough" than the first one.

Cheers,
Ben
he2.pdf

Yotam Hadar

unread,
Dec 23, 2015, 1:19:43 AM12/23/15
to Google Fonts Discussions
Hi Ben
This direction looks great in my opinion.

It does, however, appear to be slightly too big when compared to the Latin; I agree that Hebrew Mem height should usually fall between Latin cap- and x-height, but with the specific width, openness and squareness of the Hebrew design, I think you should get closer to the Latin's x-height.

I also think you should post s slightly more extensive proof, most importantly one that shows a more typical bilingual usage scenario (Latin words interspersed in a Hebrew paragraph).

I feel this proposal is far more promising than the next one (the Hatzvi-inspired one), which is far too wiggly for no apparent reason – the Latin version has a simpler skeleton.

Yotam

Ben Nathan

unread,
Dec 23, 2015, 3:33:16 AM12/23/15
to Google Fonts Discussions
Hi Yotam !

I agree with you, I think Hatzvi (or even his skeleton) is not a good fit for a type that should be used in UI design (like source sans). it has too much information.
I'm exploring different direction with Adobe to see if there's another approach to the Hebrew design, so we'll see at the end of this process which is the best fit. In my opinion my original design is the best fit (not the last 2 sketches).

I will implement your ideas - the X height and showing longer usage scenario.

Thanks for your input Yotam! would love to hear more from you,
Cheers,
Ben

Paul Hunt

unread,
Jan 10, 2016, 8:37:05 PM1/10/16
to Google Fonts Discussions
Hello Ben!

Okay, I promised to give some feedback on both designs, but this is somewhat hard to do with such a limited sample of both designs (a large alphabet in the thinnest of weights, no text setting). So I do not think that I can comment on if the apparent size or overall weight is a good match for SSP. At this point, I feel I can only comment on the letter shapes themselves.

There are things that I like about both designs. From the first design, I like the very simple styling of details. I prefer the simple termination of strokes as opposed to the slight curve into the top line that can be found in the second design. Some things that I don’t like in design 1 is that there seem to be square corners in some places where I would not have expected them: in the top right of gimel, yod, vav, lamed, qof, &c. Also I do not prefer the rounded forms of tet, samekh, shin. Is there a reason why you do prefer these shapes? I would like to understand why, if so. Overall, I find that the combination of the square forms and rounded shapes create a very static texture that I find to be at odds with the design of SSP. I feel that there are more dynamic, gestural models of Hebrew that would work better both as a match with SSP and create a more cohesive texture in text (I am guessing a bit on this second point as I have not seen this design in text).

Regarding the second design, I feel that in general it is closer to what we would eventually want, but there are some weird styling quirks that make it feel a bit wonky at this stage. There is not a consistent tension in the diagonals: I think that the forms of lamed and qof are close to what I would like to see them be, but i would like to see forms for tet, samekh, and shin that work in better harmony with lamed/qof. The alef is also pretty close to what I would expect to see. Personally, I like the form of mem with the break in the right leg, however this was to stiff and upright in the first design. Likewise, I am missing the hard corner in the top left of pe and prefer to see the break there as well. I feel that with a bit more work on the second design, it would actually be quite close to what I was hoping and expecting to see.

What are your thoughts regarding this feedback? Does it make sense to you? Would you like to defend some of your design decisions? I hope so, because I would like to learn from this process as well. If you like, I can do a sketch of what I think I would like to see if I can get outlines for your second design. Let me know what you think.

Best,

Paul

Dave Crossland

unread,
Jan 10, 2016, 11:06:41 PM1/10/16
to googlefonts-discuss
Hi 

Paul, thanks for taking time to post this thoughtful feedback. 

Ben, please could you put both glyphs files into your github repo so we can see them directly? :) 

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.

To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at https://groups.google.com/group/googlefonts-discuss.

For more options, visit https://groups.google.com/d/optout.



--
Cheers
Dave

Ben Nathan

unread,
Jan 12, 2016, 1:00:49 PM1/12/16
to Google Fonts Discussions, da...@lab6.com
Hi Paul !

Thanks for the great feedback, here are my answers:

About the first design - The basic reason why I prefer having squarish shapes and straight roof is because Hebrew as a langue is constructed from the top to bottom (imagine the Hebrew letters hanging like laundry on a rope, while latin sits on the base), having a straight roof gives the type a better Hebrew form. I don't prefer the round shapes, I only designed it so we could see if it works :) For SSP I prefer we'll use a long curve (like the second design).

About the second design - I'm happy you like the second design better, it was done quickly and you pointed out all the rough places, so i'll keep on working on it a little to harmonise everything so we could test it.

Great feedback Paul, thanks for doing this! I feel the answer is somewhere in the middle, the second option has some nice ideas and it's a good starting point. I now have a better understanding of your vision of the Hebrew version. I'll work on the second option and upload a new design soon, Thanks again!


Cheers!
Ben

Meir Sadan

unread,
Jan 13, 2016, 4:23:16 AM1/13/16
to Google Fonts Discussions
Hi,

I also feel the second design makes a lot of sense and works well with the latin. Ben's comment on the straight roofs is true, but making the roofs of some of the letters round/pointed is a distinct Hebrew attribute that was downplayed by the Sephardi script tradition in the 12-13-14th centuries, and was remedied in the 1950s by Spitzer, Koren and David (not that there weren't exceptions before them).
My point is there's nothing inherently wrong with adding more variation to the top row, while keeping the type's consistencies, etc. :>

Meir

Paul Hunt

unread,
Jan 13, 2016, 6:37:56 PM1/13/16
to Google Fonts Discussions
I’m glad we seem to be on the same page now. I’m looking forward to seeing what you come up with for the next version. I wanted to also pass on some comments from Robert Slimbach (regarding the latest design) in case they might be useful to you. He says:
I feel the font is lacking in character.  It seems to me more modernist than transitional.I would encourage Ben to back off from the minimalist approach and reintroduce some of the complexity of traditional forms. I bit of quirky, but consistent, angularity mixed with softer elements would be welcome.
Message has been deleted

Ben Nathan

unread,
Jan 16, 2016, 9:17:22 AM1/16/16
to Google Fonts Discussions
Thank you Paul (and thank Robert as well)

I'm working on the new version but it will take me some time to post the update because I'm flying to LA for a week. Sorry for the delay, on the bright side - I'll have a lot of time on airplanes and airports, and that's the best place to design in my opinion :)

Updates soon!
Ben

Ben Nathan

unread,
Jan 27, 2016, 12:25:38 PM1/27/16
to Google Fonts Discussions
Hello Paul!

Attached is a new version:


test PDF attached

Cheers!
Ben
test.pdf

Paul Hunt

unread,
Feb 8, 2016, 12:46:14 AM2/8/16
to Google Fonts Discussions
Dear Ben,

So sorry to have missed this! I have been traveling and somehow missed this update. Let me review this with my team and get back to you. Again, I’m sorry for the delay.

Regards,

Paul

Paul Hunt

unread,
Feb 17, 2016, 8:48:46 PM2/17/16
to Google Fonts Discussions
All,

The feedback from Robert Slimbach was as follows:

I still feel there isn’t enough of either of the transitional style represented the design, nor does the it feel believable to me as the Hebrew counterpart to the Latin. I feel the designer is trying too hard to match the color, proportions, and activity level of the Latin, rather than responding to the unique functional and historical aspects inherent to each script. The current Hebrew design looks very minimalist to my eye. I took a similar approach with Adobe CleanHebrew, however, Adobe Clean Latin has overt minimalist tendencies. Again, I stress that the activity level is inherently greater in traditional Hebrew than traditional Latin, so I would welcome the introduction of more defined angular entry strokes (flags?), slightly wider proportions, and a greater visual tension overall.

I have been spending the last day doing a visual sketch of my own to attempt to address some of these issues and will then show that to Robert to see if we can agree on a direction and then I will hand this sketch back over to Ben for his input and see if we can’t all come to some agreement in direction from there. I expect that it will likely take another week before this handoff. Please let me know if there are any questions or concerns in the meantime.

Thanks,

Paul

Omer Ziv

unread,
Feb 18, 2016, 1:46:05 PM2/18/16
to Google Fonts Discussions
Thanks for the update Paul!

Looking forward to see your own explorations.
Once you've had some time to look into this, would you mind sharing your thoughts / resolutions with the group?

- Omer

Paul Hunt

unread,
Feb 18, 2016, 6:27:31 PM2/18/16
to Google Fonts Discussions
Omer,

Yes, of course I plan to share my thoughts and sketches here as soon as possible.

Paul

Paul Hunt

unread,
Feb 24, 2016, 4:38:02 PM2/24/16
to Google Fonts Discussions
All,

Here  are the results of my first pass on the design, building on Ben’s good work so far. I did this pass very quickly to try to rough out some ideas of where I thought the direction of this design should be going, so some things will appear to be wonky (fitting, specific letter proportions). What I have attempted to do was:
  1. Adjust the proportions to align them more closely with traditional Hebrew text proportions
  2. Adjust the contrast to align more closely with traditional Hebrew text contrast
  3. Harmonize glyph shape classes and details (I may have gone too far in making the triangular/circular shapes all too similar)
  4. Add some details (flags) that I found attractive in the Hatzvi design.
Overall, I like the result but in my conversations with Robert Slimbach this is still not there as it has smoothed out some of the angularity and quirkiness that is present in the Latin design. Also, I inadvertently brought the design too close to that of other existing Hebrew sanserifs, namely Lucida Grande and Myriad Hebrew. I would like to do another quick pass to try to bring my work more in line with the Source Sans design in terms of tension of letterforms and to further differentiate the styling from other designs. This will take another couple days. If Ben is still interested in taking up the project after this point, I will make the handoff back to him. I am open to hearing any criticisms at this point. Thanks.

Paul
SourceHebrewTestV1002.pdf
SourceHebrewComparison.pdf

Ben Nathan

unread,
Feb 25, 2016, 7:23:26 AM2/25/16
to Google Fonts Discussions
Hi Paul,
Thank you so much for taking the time to draw your version, it's puts us in a great direction. I understand your vision clearly now.
Thank Robert as well for his comments
I appreciate the effort you're making to make SSP Hebrew.
I will take some days to design a new version based on your ideas and critic.

Have a great weekend,
Cheers!
Ben

Adam Twardoch (List)

unread,
Feb 25, 2016, 1:46:00 PM2/25/16
to googlefon...@googlegroups.com
Paul, 

one of the distinctive features of SSP Latin that has stuck with me is its non-squarenrss. Source Sans has a very low tension i.e. the arcs are relatively flat (shallow) rather than being very convex. This is especially evident in the bolds. 

I’d love to see a bit of this treatment in Hebrew. Hebrew letters are very square, but by shortening the straights a bit and making the arcs longer, I think, more correspondence between the Latin and the Hebrew could be achieved. 

I do like the skeletal structure you've employed in yor prototype, I think the "triangular" forms are most appropriate for Source Sans Hebrew. Yet I'm a bit concerened whether your Hebrew treatment isn't a tad too "calligraphic" in the flow of the skeleton. 

Since your Latin echoes American Gothics, there is a certain technical rigidity to it — not “constructed”, but more like “drawn by an architect”. Source Sans has this particular architectural quality, it feels like it’s been “built from steel”. 

For some reason, when I see Source Sans, I think of the Berlin Hauptbahnhof train station: 



I don't know, just a clue :) 

A.

Sent from my mobile phone.
--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at https://groups.google.com/group/googlefonts-discuss.

For more options, visit https://groups.google.com/d/optout.
<SourceHebrewTestV1002.pdf>
<SourceHebrewComparison.pdf>

Adam Twardoch (List)

unread,
Feb 25, 2016, 3:07:29 PM2/25/16
to googlefon...@googlegroups.com
Ps. IMO, your take on the mem is pretty good in the context of what I’ve written before. 


Sent from my mobile phone.

On 24.02.2016, at 22:38, Paul Hunt <phunt...@gmail.com> wrote:

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at https://groups.google.com/group/googlefonts-discuss.

For more options, visit https://groups.google.com/d/optout.
<SourceHebrewTestV1002.pdf>
<SourceHebrewComparison.pdf>

Paul Hunt

unread,
Feb 25, 2016, 6:52:31 PM2/25/16
to Google Fonts Discussions
Ben,

I’m not sure you caught the last part of my post but there I stated that I want to do another pass at trying to get the design closer to what will actually be wanted before I hand the design back to you. Please hold off on updating your work until you can take the UFO sources that I will prepare. I want to avoid duplicating of work as much as possible.

Thanks,

Paul

Paul Hunt

unread,
Feb 25, 2016, 7:11:14 PM2/25/16
to Google Fonts Discussions
Adam,

I’m having a hard time envisioning completely what you are describing. Is it possible for you to make some kind of sketches and share your feedback that way?

P

Ben Nathan

unread,
Mar 2, 2016, 6:45:02 AM3/2/16
to Google Fonts Discussions
Sure Paul, take your time, i'm waiting for your next sketch.

Thanks!
Ben
Reply all
Reply to author
Forward
0 new messages