Renwick is a humanist sans serif designed for editorial use. It started as a college project consisting of single text weight. I attempted to create a less unruly font akin to Gill Sans, which I've found difficult to use in longer settings. Since then I've learned so much about type design and now I'd like to expand the family—and eventually redo the text version. This is the display cut intended for sizes of 36pt or greater.
Thank you to all the suggestions already offered in my introduction post. I'm open to all feedback regarding any part of it!
——
PS: InDesign changes a lowercase eszett into a double "SS" when using the case feature (TT button) instead of a capital eszett. This happens automatically for small caps too. Is this behaviour correct?
0
Comments
I wouldn’t call it correct, but the problem is Indesign and not your font.
I gather this is intentional on Reese’s part, but perhaps it should become less drastic in heavier weights? I think it can be rather charming in the lighter weights
It is perhaps a bit too narrow, but I do like the general “smallness” to it. Maybe just make it a tad wider?
Given the tight spacing, the lowercase “l” seems to become a little problematic at times. It seems you could afford to make it narrower especially in the heavier weights.
The weight of the a’s bowl in the regular style feels a little incongruent with the other two weights. As in the bottom feels too heavy, and the center too thin. It seems you could get away with a more evenly weighted bowl, especially given the lack of contrast in the “e.”
I also wouldn’t be afraid to make the top bowl of the “g” larger (across all weights) too given the tight spacing, and the large counters in the “a” and “e.”
What is the “design size”? That is, if you had to name one specific size at which everything should seem perfect, what would it be?
What is the smallest size at which you expect it to work well?
Basically, the current spacing makes this seem to me like a “poster” size optimized for something like 288 points (or more), not recommended below 144 point.
The case mapping of ß to SS has been the official standard in Germany for many decades, and it remains the default case mapping in Unicode’s special casing rules. Fairly recently, German orthography rules have allowed for the optional mapping of ß to the uppercase form of that letter, which is a 20th Century innovation. For some users, this is a stylistic preference, but in some data processing situations, e.g. where personal names are stored in all-caps, it may be critical to preserving spelling distinctions. In such situations, tailored casing rules may be applied.
If you include Eszett in your font, but your “all small caps” (c2sc) feature does not include a small cap version of that character, then names that were typed in all caps, including Eszett, will not fully convert to small caps, with cap Eszett standing out like a sore thumb.
Do not, however, put uni1E9E.sc in your smcp feature, for that:
—where SS.smcp is a single glyph comprising two small cap ‘s’s.
I guess the options are to loosen the spacing or increase the intended size range to something more like 72pt? I'm leaning towards options two because I like the tightness.
@Linus Romer Great to hear! After making the changes I could see the difference.
@Christian Thalmann The underbite is intentional. I like the character it gives but am not opposed to toning it down slightly. The /P has bothered me for a while but I don't know why. As for /Germandbls I haven't tried a "sharp" top right corner yet, but I will now. Is the width of the character similar to the /H? When you mention "extrapolation errors" what do you mean? All three masters are manually drawn, although they likely include amateur errors.
@Matthew Smith You're correct; the obvious /C underbite was influenced by Gimlet Sans:
I'm curious why you suggest it could "become less drastic in heavier weights" and "think it can be rather charming in the lighter weights". Is the gap it creates strange in the bold weight because it's supposed to be a darker overall colour? And then the light is fine because the spacing is looser?
Looking at /P closer I noticed Gill Sans Nova already addressed this issue and widened their letter. I'll try something similar and see how it changes.
When you mention making the lowercase /l narrower I assume you mean the bottom terminal, and not the width of the stem? Now that you mention the regular /a being an oddity I can see it. I'll try fixing that and play around with the bowl on the /g too.
@K Pease Is there a reason it might "look strange in all caps"? Instead of alternates I could use possibly add a C.case but that doesn't sit right with me. I think it'd be better to design a /C that works for both upper- and lowercase.
@Thomas Phinney The spacing is likely be too tight for 36pt, but I feel like it could still work at a size as small as 72pt. I'll try printing it out at increasing intervals and see how it performs. Is there a specific reason you suggest no less than 144pt?
@John Hudson This is my first time being blessed with a John Hudson knowledge lesson! Is the "optional mapping of ß to the uppercase form" something handled on the program or font level?
@Nick Shinn This font doesn't have small caps, but the original Renwick Book did. I only used the .smcp feature when making it though and avoided .sc completely. Is this bad practice?
Fundamentally, it is a character encoding matter: if a user wants the uppercase eszett character ẞ, rather than SS, then he or she has to use U+1E9E in text, whether via character input or tailored case transformation. It definitely isn’t something for the font to do, because fonts deal in glyphs, not characters, and there is no way to handle the case mapping of ß in glyph processing.
Writing a /b with such an outstroke at the bottom of the stem interferes with the bottom of the bowl, and writing a /q with such an instroke at the top of the stem likewise interferes with the bowl. Easier in those cases to just merge stem and bowl together. On the other hand the outstroke at the bottom of /d and the instroke at the top of /p are free and clear of the bowls because they're on the other side of the stem.
Obviously in sans serif designs this "interference" issue doesn't apply, but the expectations of letterforms can still be influenced by familiarity with seriffed type. And that's especially true for a sans intended to read as "humanist" like yours.
Note that even if the spurs aren't omitted in a sans, they can differ in /b and /q compared to /d and /p: see the pared down spurs in Franklin Gothic as a good example. I think it's fair to say we expect the "stemminess" of /d and /p to be more than, or equal to, that of /b and /q, but less than seems a little weird to me.
I presume it's because of pen logic: The spurs of /d and /p occur in the direction of the strong diagonal, but those of /b and /q in the weak diagonal. See for instance EB Garamond:
@Christian Thalmann Seeing a visual guide like that is very helpful. I haven't found time to return to this project yet but I'll make sure to post the changes once they're done. Regarding the underbite on the /C, I've been looking a few other typefaces and none have the very obvious underbite like Renwick does. Perhaps I've overdone it in pursuit of "coolness".
PS: That's a lovely /Germandbls in Ysabeau.
Revisions are explained in the PDF along with visuals, but to summarize it here:
Comments that aren't in the PDF:
This makes me wonder if font designers committed to super-tight sidebearings/kerning ever darken their wider letters not by narrowing them but by thickening their stroke weights for the sake of balance. Or is the idea of color balance simply sacrificed for the desired spacing effects?
In both weight extremes, the space is still awfully narrow. I might try increasing it by 1/3 or so and see how that looks.
In B P R the transitions from curve to straight in the tops and bottoms of the bowls are overly abrupt (all weights but maybe especially the bold).
In most pro typefaces, the bowl of the P is bigger than that of the R, at least vertically and sometimes horizontally as well. That seems true in your ExtraLight but not the Bold—Bold seems like it is maybe the reverse, the R is bigger?
The bold S seems pinched at bottom left compared to top right. That may contribute to the feeling that it is leaning backwards (to the left).
The weight progression makes the O/o in particular seem much more condensed in the bold, more than I expect given weight treatment in much of the typeface. I look at how you add weight to MN. And bdpq are given more reasonable expansion on the outside. It is OK to make O/o outer contour get a little wider as weight gets added, it does not all have to go to the interior, maybe 2/3, or 3/4 at most.
[Gill Sans Nova light and bold weights]
[Gill Sans Nova light and bold weights]
But either way, I think the space needs to be a bit wider.
So yes, some things would overlap. But at least the TYPICAL spacing would be improved.
In many cases, those potential slight collisions are not a problem. KA or AA are fine to overlap a tiny bit. VV might start to look like W so you could kern those two apart. But VV is a rare enough combination (savvy and revving notwithstanding) that that seems livable to me.
It seems to me that this is the “right” spacing for those letters, given how tight your overall spacing is. But I won’t claim that is the One True Answer. Plenty of skilled type designers almost never do negative sidebearings. (But even they would make exceptions, I think: for example for an overhanging lowercase f.)