Playing around with Venetians

ivan louette

Thanks ! My friend Nicolas J. had found this middle path before and indicated me the way On the other hand did you experience the same kind of display problem than Samuil above ? and do you have an idea to fix it ?

Craig Eliason

I didn’t notice that. Sounds like a hinting issue. Is there any hinting in the font (including autohinting done by the editing software)?

ivan louette

Yes perhaps an autohinting problem. I am not very experienced in Hinting. I will see what I can do.
Otherwise here is another evolution. I noticed that /h,/m,/n and /u curves were more squarish and I understand now that it play an important role into Jenson type ribbon effect (or photographic film like effect).

Samuil Simonov

That's what it looks like on my screen. On paper, as expected, it looks normal. Must be some interference somewhere between your font, Acrobat and Windows.

ivan louette

Thanks ! Strangely that's a little bit like how LibreOffice displays some fonts (and not only mines). And My old copy of Illustrator 10 on Wine shows something similar (see attachment1) while inkscape keeps it fairly clean (att. 2) and Okular, one my PDF viewers just a bit fat but not out of proportions (att. 3).
I will investigate on how to obtain better hinting.

John Savard

On the issue of making the typeface more regular: that, in my opinion, depends on the use to which it is to be put. If it is intended for display or novelty use, or otherwise for simulating the appearance of old documents, imitating the behavior of type in the period in question is appropriate.

But the Venetian design of a typeface has such merits as to be suitable for the composition of text matter for general-purpose use in the present day. Imitating the limitations of printing processes at the time of its origin in a typeface for that kind of use would only be annoying and distracting, and limit the use of the type.

So I'm not sure there is one right answer. Possibly the typeface should exist in fonts of both versions.

ivan louette

One more step. Many changes on lower case drawings and slopes, draft capitals (still no numerals at the moment), and some attempts to weights.
No hints at the moment because automatic hinting didn't give convincing results. But it's a printing font.

André G. Isaak

The bottom serifs on R and P don't really match the font. Otherwise, very nice.

ivan louette

Thanks ! Yes of course you are right, that's why I said capitals are only drafts at the moment. I use them for testing, their proportions are right but their details are still approximative. Some of them are also misinclinated and the stems of some others are too wide.

ivan louette

A better view to the range of possibilities with an Extra Bold attempt.

PabloImpallari

ivan louette said:

A better view to the range of possibilities with an Extra Bold attempt.

You can go a lot more bolder... your current ExtraBold looks to me like a semibold.. so there is room to go further. Go explore the limits in a few letter and use adhesion text to test in some paragraphs.

Also, test the Bold and Xbold in context... I mean, mixed in the middle of a "book" paragraph.. and not only in a separate "Xbold" block of text... this will reveal if the Bold is a god match for the Book or not.

ivan louette

Thanks a lot ! I will try it. It's true I am a little bit shy in my attempts ;-)

ivan louette

Here is my attempt at numerals. I show two differents drawings for "4"

John Savard

Isn't a 1 in oldstyle numerals that looks like a 1 instead of like an I a rather recent innovation?

ivan louette

In fact I hesitated some time for the 1. You are right, at Jenson time it seems that 1 was more like a short I with a symmetrical above serif, and I could try it too. In Jenson printings I only found rare 2, 3, 4, 5, 6 (7, 8, 9 are not too complicated to extrapolate) at the bottom of the pages of the most tardive Jenson's works. I am even not sure if they were cutted from him because inside the text he seemingly always use roman numerals. Thus I was less historically accurate in my drawings than for the letters because my main goal is matching (if possible) the texture of Jenson printed pages. Thus I only tried to draw "elzevirian" numerals which keeps the drawing style, the x height and the weight of the lower case letters.

Craig Eliason

Yes, I would at least have a go at more historical forms for both one and zero. (You can always furnish the font with both as stylistic alternates.)
I think the "extenders" of six and nine get too thin too fast. (Perhaps bottom of seven gets too thick too fast.)
Some of the rounding might get a bit mushy. Top left of five and top of four both seem quite soft to my eye.
I like the top four better but either has potential. 
Be sure to test in mixed settings to ascertain if the scale and color are working with the letters.

ivan louette

Thanks for your advice. I will see what could be done for the best and come back soon with results.

ivan louette

I tried what you told and that seems to work very well. For the lower part of the 4 I tested an intermediate between my two solutions.
After some investigations I found that Manuce uses a zero (from Griffo) with a much wider counterpunch.

But Manuce uses also a higher 1 and it's not clear that his zero keeps any tracks of calligraphic pressure.

ivan louette

Sorry for this splitted post… The picture refers to the last paragraph in my previous one.

John Savard

ivan louette said:

In fact I hesitated some time for the 1. You are right, at Jenson time it seems that 1 was more like a short I with a symmetrical above serif, and I could try it too.

Not just at the time of Nicholas Jenson. Before 1990 or thereabouts too. For example, it's like a capital I in Garamond.

This is a very recent innovation. At least the first time I noticed it was in Georgia by Microsoft.

ivan louette

Interesting ! I didn't notice it was so recent. Here is a new attempt with some changes to regularise oblique extensions in  2, 4, 5, 7 and some changes in weight and color. I will keep the lower 1.

Kent Lew

Before 1990 or thereabouts too. [...] This is a very recent innovation. At least the first time I noticed it was in Georgia by Microsoft.

Uh . . . I’m not sure I would agree with that “very.”

For example, Electra 1935:

I’m sure I could put my hands on a few other examples from the same mid-20th century period, if I took the time.

Nick Shinn

Goudy, Cooper and Zapf often both angled the top stroke, and had it protrude to the right. And they made it wide. That seems to me like a good approach, as it provides breathing space within the character—necessary in today’s manner of fitting oldstyle figures proportionally, and to the same amount of tightness as the lower case. You don’t want the number eleven to be a couple of tiny marks squished up against one another.

ivan louette

Here are some progresses with my Uccello font. I added some ligatures too, particularly those seen in original Jenson.

ivan louette

A lot of work again to attenuate contrast, giving more weight to the too slender parts of some characters (m,n,h…) and giving more curvature to the top of the same ones. More weight too to serifs…

Thomas Phinney

Feels like you changed the optical size from display to text. Both of those optical sizes might be useful, you know?

Your cap S leans forward a bit, but your lowercase s leans back. Either is ok (though forward much more common/typical), but the juxtaposition seems pretty strange to me.

ivan louette

Thanks ! My first goal is going to text. But you are right I should keep two versions togheter. However the display version is a pain for the eyes when printed because of its higher contrast. About S and s I have respected their respective and opposite slanting seen in Jenson's prints. On the whole pages this is not so strange
I just replaced the JensonLatin.pdf file because the short transversal in long s was placed too high while it must be placed just below the x height for the reading flow.

Craig Eliason

Ascender serifs /b/d/h/k/l/ might get a touch too dark. Maybe also some of the serifs on diagonal caps like /K/V/W/. Bowl of /P/ looks a little pointy to me. Width of /N/ serifs surprises me (compare /A/M/U/). Serifs on /T/ and especially /Z/ stand out too, but maybe they work in running text (your specimens don't have too many caps set with words). 

ivan louette

Thanks a lot, I will take a look at these serif weights tomorrow. About the N width Jenson one is very wide, like is the H too. I will provide some text in capitals. These ones were at a draft stage up to recently and thus they remain a work in progress.

ivan louette

@Craig Eliason I did some changes about what you told yesterday. I rework some capitals too and here is a dummy text composed with them.