Design of exotic Latin letters H/h with stroke, kra, T/t with stroke, Eng and Vietnamese diacritics

and's picture

The following extended Latin letters and diacritic marks are completely unknown to me in the sense that I have never seen them in actual use, and I was hoping to get some advice on preferred designs. Thanks in advance for any help. My immediate goal is to come up with designs that work for Bodoni 72, but most of the questions are rather more general.

***

U+126 Ħ (H with stroke) often has a horizontal line through the letter, but there are also examples of designs with a double crossbar:

Would it be acceptable to centre the crossbars vertically?

Is a particular design preferred for Maltese? Is this letter used in any other languages?

***

U+127 ħ (h with stroke) normally has a horizontal stroke (as opposed to a diagonal one), which probably makes sense for consistency with the uppercase.

***

U+138 ĸ (kra) appears indistinguishable from Cyrillic lowercase k in the Unicode tables, whereas I would have expected it to have the same top-left serif as a Latin k or n, and perhaps a straight rather than curly lower diagonal.

Apparently, this is an obsolete Greenlandic letter. Does anyone know its correct shape?

***

U+166 Ŧ and U+167 ŧ appear with a horizontal stroke in the Unicode tables, but with a diagonal stroke elsewhere (e.g., ECMA-94):

The main Language using this letter appears to be Northern Saami. Would the following designs with strokes borrowed from Polish L/l be acceptable?

***

Uppercase Ŋ often has a very thin descender, to match the right vertical of the N. A design based directly on N+J works less well than the standard lowercase n+j design:

Apart from the oversized lowercase approach, what would the best solution be (for a Bodoni)?

***

The Vietnamese horn interferes with the right serif on uppercase U. Is the preferred solution to have the horn at different heights for U and O (in such a way that it naturally replaces the serif on the U without ending up too close to the centre of the O)?

Are there any designs with the horn below or extending to the right of the serif?

***

Would the Vietnamese hook diacritic in a Bodoni normally have a top-left ball terminal and curve slightly to the right at the bottom (like a dotless question mark in miniature)?

Examples of well-designed Vietnamese fonts in a Bodoni-like style would be greatly appreciated.

Theunis de Jong's picture

(tracking)

and's picture

Just a short follow-up to my original post.

For capital Eng, using the large ball from the J and connecting it with a thin stroke might work:

Should this letter have a full descender (200/1000) like a lowercase letter with descender or just a short one (100/1000) like the capital J has in this font?

Bendy's picture

Good set of questions! I'll be tracking this for sure.

Re Vietnamese letters, I started a thread a while back and several others gave some great links. The thread is here. It may not answer your exact questions but it's good anyway to see some examples of horn and hook there, and doubtless you are drawing other Vietnamese diacritics.

John Hudson's picture

Ħ : I usually design this so that the lower bar is slightly lower than the crossbar of the H, to create more room above it for the second bar. And I usually extend the upper bar through the side stems, because that is the form as I saw it in earlier printed Maltese books. When I was in Malta, I saw both forms in use: with the bar extended and not as in the examples you show. I also saw versions with the two bars optically centered as you suggest, but these were in sans serif signage and display type, not text.
___

ĸ : See http://typophile.com/node/18546

I believe the glyph shown in the Unicode charts is misleading. Unless your Latin lowercase has characteristics that support such a Cyrillic-like form, it will look like an alien visitor.
___

Ŧ : I've always seen this with a horizontal bar, which corresponds with that of the letter Đ/đ, also used in Northern Saami.
___

Ŋ : Your revised form is heading in the right direction. The hook should descend as far as that of the J, or the tail of the Q, i.e. not usually as deep as the lowercase descenders.

Note that this uppercase form is normal for European fonts. For African languages, the n-like shape is preferred.
___

Vietname horn : I've seen (and designed) both arrangements: with the horn coming out of the side of the letter and with it coming off the top right. Your solutions all look reasonable to me; just watch out that the horn has enough room to breathe, both between it and the letter to which it is attached and with regard to a following letter.
___

Vietnamese hook : Always tricky. Your shape looks okay, and is appropriate to the style. But you have to test in print and see how it looks alongside the other marks. It is easy to make this either too dark or too light.

and's picture

Thanks for your replies. Below is a summary of my current understanding with further comments and questions.

***

As far as the length of the upper bar is concerned, Ħ can take either of the following forms, but the second is the more traditional one in Maltese typography.

Vertically symmetric placement of the bars is somewhat of a novelty and probably best avoided, but the lower bar may be at a slightly lower position than that of the single bar in a regular H. This leaves the question as to the preferred distance between the bars; fonts on my computer have distances varying between approximately 1/8 and 3/8 of the letter's full height:

(Many of these are clearly not well designed.)

***

John Hudson did not mention ħ, which probably means that the form with horizontal bar is the only acceptable one on Malta. The following design matches that of đ:

***

Despite the ĸ glyph in the Unicode charts, kra is essentially a k without ascender (1st picture below), but it should have the same top-left serif as a u (2nd picture); a shorter top-right serif akin to a (small-)capital K might be better (3rd picture):

(The u serif is so flat in this font that the only obvious difference between kra and a small-capital k is that the latter has a top-left serif extending to the right.)

Samuel Kleinschmidt's 1851 grammar does show a kra with ball terminal, but all the Greenlandic letters are printed in italics, so this applies to k as well:

***

Thanks for pointing out the parallelism between Ŧ/ŧ and Đ/đ. A ŧ with diagonal bar would have matched an eth-like đ, but 19th-century Lappish grammars already used the modern/current forms with horizontal bars. It would be interesting to see an extract from the late 19th-century Bible.

The bar on ŧ should perhaps be a bit lower:

***

Ŋ: Perhaps the slightly smaller ball from a lowercase j is more appropriate?

As for the depth, Q has full descender depth in this font, whereas J is ‘half-deep’, which probably gives some latitude.

***

Many of the Vietnamese TeX samples from the earlier thread show the U/u horn at an intermediate vertical position:

This works well for the lowercase u, but perhaps less well for the uppercase U.

My original question was whether it would be acceptable to put the horn lower on O than on U lest it end up too far to the left on the O (without making the horn exceedingly wide).

***

Your comments and the samples in the previous thread confirm the basic shape of the hook. I may return with further questions related to that one once I try to put together complete Vietnamese accented letters.

and's picture

The Ħ seems to look better when the bars are fairly close given that the horizontals are rather thin in this font.

Igor Freiberger's picture

In the font I'm developing these glyphs looks like this:


Probably I must make hook in Vietnamese characters slighty bolder and bars in /h/ and /t/ a bit lighter.

and's picture

I agree with your own remarks.

Personally, I think that the barred characters feel less like an afterthought when the additional bar on the upper/lower-case letters matches the (original) horizontal bar on H/t. Many fonts do not follow this, but it is difficult to know whether the choice of a different stroke width was a conscious one.

K with hook is an African letter which has not been mentioned earlier in this thread and of which I know nothing. Your kra (cf. the Palimpsest thread) appears to have the correct shape.

The capital Eng is perhaps a bit heavy in the bottom-right corner of the N (i.e., around the bottom of the diagonal) and a bit light at the bottom, but there is no other obvious solution given the shape of the J/j. Sorry for this not particularly constructive remark.

It has been mentioned in an earlier thread that Vietnamese grave traditionally goes to the left of the circumflex.

Some of the Vietnamese diacritics extend quite far to the right (e.g., lower-case a with circumflex and acute) — do they not interfere with a following ascender (e.g., l or b)? You might want to shorten the acute slightly (cut it at an angle at the bottom to make it fit atop the circumflex).

Syndicate content Syndicate content