Goudy Village

ivan louette

Because it's so close to Venetians (in fact like an Art Deco or Art Nouveau Venetian), I had the idea to do something with it for a very long time.

That's also interesting to know that it was designed by Goudy at the same time Auriol designed its first release of Auriol (1903)

While Goudy changed the /i dot in next releases I loved the nearly triangular of the first one but it was a challenge to draw accents.

Jacob Casal

That is a tough one with the tittles distantly cradled by the rest of /i and /j. It’s a little hard to put it into words, but the /s feels a bit off next to the other letters, particularly the lower stroke until it meets the spine. Could just be me.

George Thomas

...but the /s feels a bit off...

It should. None of the rounds have any overshoot that I can detect. If there really is some, it is so tiny as to be immeasurable.

ivan louette

@Jacob Casal Thanks, you are right : the /s was slightly off. I corrected it.

Here is a test with accents in French.

Claudio Piccinini

What sources did you use?
I see there are just a few digital designs trying to grasp the typeface’s essence (or meant as "reinterpretations"?) – it would be good to have a rendition wtrying to be (as much as possible) accurate.

ivan louette

@Claudio Piccinini My first emotion with this typeface was caused by the last picture on this page of Luc Devroye's site. At this time I was working on Uccello, which is an attempt to draw an accurate digital version of Nicolas Jenson's roman. My second source is this one found on Internet archive. I don't have a source precise enough to draw a perfect outline of the Village characters but here I am interested to capture the dynamics and the texture of the typeface. It's an improvisation around it and it allows me to try some stylisation. At one time I will also create an alternative with slightly soften angles.

Craig Eliason

Overall this appears quite relatively light, which seems to me a bit contrary to the Arts and Crafts "robustness" favored by that era (but maybe suits the different interests of today?). 
Your tittles read as more "linear" to me (like accents aigus) than the triangle shapes I get from Goudy. Your /r seems wider to me than his. 
I love when considering a revival like this gets me to look more closely at historical designs. Those varying descender depths are wild!
I'm trying to diagnose why he can get away with that fender-bender of a /g: it may be that the more intact top bowl is bigger in his version than yours.

ivan louette

@Craig Eliason About lightness I understand your point of view if it referates to William Morris Golden Type, which is robust and even remembers block books much more than fine Renaissance printings. But while my second source (the re-edition of Printing by Goudy) seems to show a relatively strong print my third source (more recent) doesn't look so dark. Of course Steve Matteson's elegant version is lighter again but I think this is his own choice and it would take its place too in display or fine printing. I think Village is a very modern design which allows interesting experiments with weight. For example if I want to be closer to my second source I would increase it slightly like below. (But in this case it would need more attemption to details and spacing).

Nevertheless I verified my /r and it's not too wide, and /g seems correct too. But I don't know your sources.

Jasper de Waard

I don't know about sources, but this darker version looks pretty great to me. I agree about the tittles: they look like accents.

ivan louette

@Jasper de Waard Thanks for your comment. At the moment that's only a trial (an automated increasing of weight) and it needs some cleaning. But I like it too This is the interest of discussions here that they drive to open the eyes.

About tittles and accents my choice isn't closed : Goudy's first version has triangular tittles and no accents (it was probably dedicated to English only and thus no confusion is possible), but more recent prints, like those of my third source show smaller elliptic and slanted tittles… and tiny accents are added to the list of french names of paintings. This second version of tittles is a little bit too small for my taste, and the accents much too tiny, but they could be inspiring for an open type style set.

Nick Shinn

LTC Village.

ivan louette

@Nick Shinn That's another interesting option. In 2010 I developped an SVG Morphology filter effect which is now included in Inkscape. It can be set to create this kind of effects you may vectorize afterwards. It's based on gaussian blur, and you can even increase it independently on x or y axis.

Here are some different uses : for the first one the text was rotated and grouped before applying effet and moved back to the horizontal after vectorizing, which gives a diagonal shift. And for the second try blur is increased more vertically. Of course this is exaggerated for these examples.

John Savard

Since the books typeset by Nicholas Jenson were written in Latin, they had accented letters, so it's not as if there is no model available. But the approach you chose clearly makes the accents more legible in smaller type sizes, compared to distinguishing them from the dots on i and j by making them lower and more horizontal, which I think is the more usual and conventional approach, instead of higher and more vertical as you have done.

But what am I saying? Am I praising you for being bold and unconventional, or am I criticizing your new approach and demanding that you stick to the tried-and-true? More the first than the latter, but also, in a fundamental way, neither: I don't think my subjective opinion is that valuable.

Or, perhaps, although I really don't think you need my advice, it is possible that accents that are lower and flatter instead of higher than more vertical might be desirable, as an alternate, for use when the typeface is being used in larger point sizes. Always sticking to the conventional is no way to make progress, but also conventions are sometimes there for good reasons.

EDIT: Upon further reflection, I do see one possible major potential problem with your approach, although I may be completely mistaken. The letter í - where an accent replaces the dot, if the accent is more vertical, then it's less likely to be seen as an accent than as a tittle, at least it would seem to me.

André G. Isaak

John Savard said:

Since the books typeset by Nicholas Jenson were written in Latin, they had accented letters, so it's not as if there is no model available.

Latin doesn’t normally use accents. Jenson’s works contained a variety of latin abbreviation marks, but no accents.

ivan louette

@John Savard Thanks for your reflexion about the role and the perception of diacritics. That would help my quest in this very difficult area. The balance isn't so easy to reach between the aesthetical point of view which consider a diacritic as integrated into the drawing of a glyph and a more utilitarist point of view which consider it only as an appendix. There are so much weird diacritics around, even in mainstream fonts and that's the reason why I try something else to satisfy my taste, but I may be wrong too.
About the difference between the /i tittle and an /i with an acute accent I will do some testing.

@André G. Isaak Despite I don't undestand their sense I included all glyphs I saw in Jenson prints into my Uccello. And if the two first lines below are probably abbreviations, the shapes above the glyphs on the third line seems to be accents.

Adam Jagosz

That last line contains abbreviations as well: on, en, an, un / om, em, am, um. Accents would more probably (though not necessarily) be centered above the letter, which here they are not.

ivan louette

@Adam Jagosz Thanks, you are probably right. However Griffo used true accents in Latin in De Aetna a few times later. (An this is intricatingly funny to see that his accented /e had always an inclinated transverse while his other /e didn't.)

ivan louette

Here is my attempt to solve the too accent like /i tittles issue. The original Goudy's tittles were far better than my first interpretation. Thus I came back to them and change my accents drawing too, making them more angular. It seems to work better now in french texts.

Nick Shinn

You should abandon these Goudy’s tittles. Thing is, he was designing for English, with scant regard to accents. To maintain a certain amount of Goudyness though, you might consider diamond tittles.

Also, rather than French, test this issue in a language such as Hungarian, which uses the í accent.

ivan louette

@Nick Shinn Diamond : interesting option indeed !

Nick Shinn

Also, you don’t need to make the grave accent the exact flipped image of the acute. There is plenty of historical precedent from French foundries for making it a little less steep. After all, it represents a less “sharp” sound. But I wonder what Francophones today think of doing that?

ivan louette

@Nick Shinn I think you are right about the grave accent. In this case it could be slightly more slanted to the left. I did it already in the past for other fonts. Also the pronounciation of /è in classic French is shorter than /é (of course this is variable regionally : sometimes in Belgium, my country, it's a little bit longer than in France

Jasper de Waard

The diamond works okay, but I think it's a little too tame. Have you tried a triangle (like the goudyeske one you had), where the sides have roughly the same length? This would probably prevent it from looking like an accent. In other words, take what you had and scale it down vertically, then see if it works.

Jacob Casal

And if you do keep the diamonds, then perhaps try placing them at an angle. At least to my eyes, the tittles’ flat vertical and horizontal edges of their points feels disconnected (no pun intended) from the shape of the rest of the /i.

John Savard’s tittle in the following comment are closer to what I saw in my head, but with blunter points, the /i cradles it more.

John Savard

Perhaps it only reveals my lack of taste, but this crude mock-up shows what I personally would prefer:

the accents low and close to the letters, and the tittles modelled after those of Morris' Golden type (or at least its imitators).

Thomas Phinney

I find either of these two extremes of angle (either Ivan’s upright or John’s horizontal) more unconventional than most of the other elements of this typeface. Somehow for me the accent treatment is more disturbing than anything else.

John Savard

Well, I'll admit that I did a very crude drawing - I simply turned his accents through 90 degrees. I'm happy to have the accents at more of an angle. Putting a bit more work into the drawing, here is a revised conception (I also didn't quite do the tittle on the i correctly, it was too close to the body of the letter):

The important thing, to me, is that it's very easy to discriminate the tittle from the accents. But the accents are overly conventional, and the tittle is not faithful to Goudy's original design, so I'll freely admit that this design has flaws - it might, however, be a fallback if doing it right can't be made to work.

ivan louette

Wow ! Thanks a lot to everybody for this bunch of interesting reflexions ! Here are mainly three yesterday trials. @Jasper de Waard Yes I tried a triangle closer to Goudy's original and it works better than the accent like tittle. However if accents are added the proportional sizes of these differents elements is important too to differenciate them. @Jacob Casal I was also aware that a too rigidly drawn diamond isolate it too much from the rest of the letter /i because it breaks the movement of the glyph and also the general movement of the typeface, and thus I have drawn an asymetric diamond and slightly moved it to the right. @John Savard for the same movement reason (movement direction and vigor initiated mainly by the top end of the upper serifs) I have kept my accents more upwards (despite I changed the angle of the grave accent). On the other hand if the closeness of the accent and the main part of the character works in certain fonts I think this isn't the case here. My third picture below shows a text portion from a Catalogue of an exhibition of paintings of the Dutch Masters in 1910 at the Metropolitan Museum of Arts. It shows the most known evolution of the Goudy Village tittles which became elliptic and slanted. But it shows also tiny triangular accents far above the characters (like the tittles and double tittles were). I din't know if the accents were designed by Goudy himself, but it seems they are made very discrete to avoid any noise.

Thus at this stage my idea is optimizing again the third version in the first picture below (finding the right sizes for the ideal impact is my goal and at the moment this is not perfect : perhaps still too big) and use it as default and adding two style sets : one close to the first example in my first picture and another close to what is shown in the 1910 catalogue.

Christian Thalmann

Ivan: That bottom row is nice. I'm a fan of steep accents in general, and they work well with the flamboyant style of your typeface. I'd prefer the tittles a bit lower, but their shape works.

The tittles of the top row are also pretty, but should perhaps be stowed in a stylistic set and/or an alternate cut of the typeface.

Craig Eliason

Although the diamond sounded like a good idea to me, and I do think that “de-accentifying” the original design is necessary, I’m not persuaded of the fit of those diamonds. Their sides read as convex (though zooming in I see they’re not actually so) which feels foreign to the font that has so many concave shapes (including the head serifs of i and j); and they may be yet too symmetrical.  

John Savard

And here is how I would "improve" the one in your top row, to get something that is indeed much preferable to the earlier drawing I provided:

The tittle is moved downwards, and closer to the letter; the accents are slanted a little more, to be more conventional, but not much more, merely 15 degrees.

This, I think, avoids the flaw of my earlier image, which was not at all faithful to your design or Goudy's original, and yet the accents and the tittles are very distinct and very legible.

Thus, I think it shows that high legibility isn't incompatible with your design goals.