Elemaints - A Serif Family with Optical Sizes

AbrahamLee

[Comment retracted]

Christian Thalmann

The display style is very pretty indeed!

Some of the Greek lowercase look very light, such as /zeta/ and /pi/. The Latin question mark possibly too? And the /at/.

The capital tonos looks strange to me; I would expect it to be more vertical and solidly connect to the cap line rather than just hanging about there. It's also a bit small. And you shouldn't use it at all in all-caps settings, right?

The alternative /kappa/ looks extremely cursive. You'd need a more stately form for the upright (or ditch it...).

Linus Romer

@Christian Thalmann I tried to follow your advice. I don't think, that the /at was too light, but rather too low in contrast.

/zeta:

/kappa and /pi:

/Alphatonos und /Epsilontonos:

/question:

/at:

Linus Romer

The display faces and the interpolated optical sizes between display and tiny are now (more or less) finished. I would like to share a funny thing that happend during interpolating the ℚ: Due to wrong point numbers, the interpolated shape looked like a person inside a medaillon:

I struggled quite a bit with the math part. It did not come to my mind immediately, that in a single LaTeX font at most 15 different nonzero heights, 15 different nonzero depths may appear. Therefore, I spent much too much time on error detecting. But I also changed some shapes, especially the bracket pairs. E.g. the curly braces have become curlier:

Linus Romer

At the moment, I am working on the bold tiny master. Comparison with the other faces: (display at the top, tiny at the bottom)

As alway, I am glad to read suggestions and tips from you (in this case especially for the "a", "g" and the "f", which are not yet satisfying). Spacing and kerning of the bold faces will be done later.

Claudio Piccinini

I think the tiny text series, in the Bold version, would benefit by being a bit wider. Also, maybe a bit less overall spacing?

Linus Romer

@Claudio Piccinini I have extended the bold tiny face slightly and made less overall spacing. The glyphs are more pleasing to my eye, thank you for your suggestions:

Claudio Piccinini

This large it looks better… hopefully so in the smaller text sizes.

Daniel Benjamin Miller

I find myself preferring the upright /kappa. Is it possible that the working fonts be posted on GitHub or similar so that we could use, examine the whole font and/or send PRs?

Further, are you planning on releasing a version using OpenType MATH (that could be used with unicode-math)? This has the benefit of not being so limited by classic LaTeX's restrictions on math alphabets …

I think that the Hebrew characters (I know, it's only aleph, bet, gimel, dalet …) need some work. They are a bit spindly and look odd to my eye … I'd want something slightly closer to the modern expectation of Hebrew fonts (i.e., a little more like Frank-Rühl). But you are not intending these glyphs to be used with Hebrew text, so it's not as big of an issue as is could be…

Also: overall I really like the design and I'm looking forward to its completion.

Thomas Phinney

The "tiny" version seems more like a regular text size, for 10 or 12 pt.

Christian Thalmann

The apex of Bold /a/ is shifted to the right, which changes its character quite a bit compared to the Regular (especially in display). The /a/ is pretty characteristic of your Regular, so this hurts family cohesion.

Linus Romer

@Thomas Phinney What does make you feel the "tiny" version like a regular text size? If I compare Elemaints to Minion Pro at comparable size, I would say that the (from Display and Tiny interpolated) Caption face is comparable to Minion Pro Caption in terms of pen widths, glyph widths, contrast, spacing:

If I overlay Elemaints Regular (blue) with URW Garamond No8 (red), which I would classify as text face, it seems to me like they lie in the same optical range:

Could it be that you like heavier designs like Arno Pro? Compared to Arno Pro, Elemaints is clearly lighter in small optical sizes: (left column: Arno Pro Display, Subhead, Text, Small Text, Caption; right column: Elemaints Display, Subhead, Regular, Caption, Tiny)

Linus Romer

@Christian Thalmann Good catch: Now the "a" of the bold faces should be more coherent to the other faces:

Daniel Benjamin Miller

I would say of URW Garamond No. 8: it is a reasonable text face, but slightly spindlier to me than some Garamonds, though less than Monotype "Garamond". I think either optical size design is perfectly reasonable for a text face depending on the particular preference of the person.

The comparison with Minion I think is quite apt.

Thomas Phinney

The comparison with Minion is apt, and... that was some 20+ years ago.

I think Robert later realized he could have gone further in making the small sizes more robust, and did so in his later designs. Which is why you see Arno, 10 years later, with more beefed-up small optical sizes.

Linus Romer

@Thomas Phinney Fair point. I now have made the smaller optical sizes darker:

They are still not as dark as Arno or Garamond Premier (Elemaints Tiny is similary dark as Garamond Premier Caption), but clearly darker than before.

Linus Romer

@Daniel Benjamin Miller At this stage of the project I fear that the prototype font files could spread all over the internet. They hardly cannot be removed later and people will use the prototype fonts even if the typeface has become "final". Therefore I will not put the font files on the internet now.

I do not intend to implement OpenType Math. This will be incredibly much more work which others may pick up later happily.

The Hebrew glyphs are actually based on Frank Rühl's Hebrew letters. Indeed, they are math symbols and not part of the "lowercase" Hebrew alphabet and therefore not intended for Hebrew text. I understand your point about being spindly but could you point out why they look odd to your eye? (Image left: Elemaints Regular, image right: FrankRuhlLibre-Light scaled up to fit capital height)

Daniel Benjamin Miller

Well, I think that the basic principle as to why they're spindly shows in your using a light Frank-Rühl face as a basis. But looking at some other fonts which only include Hebrew for mathematics, it seems in line with what they are doing. So you should be fine unless you were trying to make this a Hebrew text face.

N.b.: it's Frank-Rühl because it was designed by Rafael Frank for C. F. Rühl's firm.

Linus Romer

@Daniel Benjamin Miller I have made the Hebrew letters in Elemaints Caption darker now:

However, to my eye, the Hebrew letters are now slightly darker than the Latin letters. What do you think?

Daniel Benjamin Miller

I like the new weight better now.

Daniel Benjamin Miller

Would it be possible to see some running sample texts in the sizes closer to completion to see how they are coming together on the page now (after many adjustments since the last such sample)?

Linus Romer

Actually, there are not that many adjustments that I have made since August 31 (when I posted the last specimen for optical sizes) - except that the lower optical sizes have become darker. Here is a sample image, but the attached PDF contains more examples. The spacing and the kerning will be overhauled when the glyphs of the whole family are designed.

Daniel Benjamin Miller

I feel the top-left portion of the /a is too bent over, making it seem as if it were forced to bend down. Here is the character is in all sizes put together: I think it's more prevalent the smaller the size but doesn't fully go away even at display size.

For comparison, here is the same glyph in Minion Pro:

and in Crimson:

Linus Romer

@Daniel Benjamin Miller  Thank you for your hint. I tried to make the top-left portion of the /a less bent over (from left to right: Tiny, Regular, Display, bold faces in the next line):

Linus Romer

A first version of the bold faces for Latin glyphs:

Craig Eliason

Looking good. I love the optical sizes. 

Teardrops at top of bold /c/ could be larger and bold /w/ may be a little light. Descending teardrop of bold /y/ could also be more meaty (but good as is in the tiny cut). 

Should proportions of white space in /g/ (both closed counters and the space between the bowls) be more consistent between the weights?
/bdpq/ in bold display are too narrow. 

Linus Romer

Thanks for your valuable hints, Craig. I implemented your adviced changes. The thing with the proportions of the /g is a bit tricky, but I think it got more consistent:

Christian Thalmann

The Display Bold /g/ looks a bit unbalanced in that text line (eye too narrow, top darker than bottom). Maybe the counter of the eye could be a bit more hyperelliptical?

That apostrophe in the Display Bold needs a lot more weight. I'd make it larger, too. The period is also rather light.

Linus Romer

/g: I have made the top counter wider and increased the stroke widths at the bottom:

The /quoteright had mistakenly still normal weight (the unicode point 2019 is so far away...). I made it bold and also gave more weight to the period (and the dot accent as well):

Daniel Benjamin Miller

Linus, are you still working on this project? I'd be glad to see how it's progressed!