Hi Michael, just out of curiosity, What is the project that you couldn't set up new H&Js to suit your needs?

Great thread. I'm generally against typefaces with open bowls in the P and especially in the B. I'll have to go find a sample.

You're right, I could have. But we were under a major time pressure and I wanted it to work "right out of the box."

--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

Wow, I didn't realize how small the wordspace is on FreightSans. Definitely too small for extended reading!

See, I'm not crazy. Lazy not to have set up custom H&Js, perhaps, but not crazy... :-)

--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

The Freight word spaces aren't that small. How hard is it for you to set them yourself? It would take about 5 seconds.

—K

Well, I beg to differ, Kris... set a test sheet of a few common humanist faces -- Freight Sans, Syntax, TheSans, Frutiger, Meta, etc. -- at 10/14 with no H&J adjustments. I think you'll agree that Freight's default wordspaces make it harder to read than a good number of the others. I was resetting a huge, multipage document in an old version of InDesign, and it was a "get-it-out-the-door-NOW" scenario. Guilty as charged.

I guess the tight default wordspaces are not really a deal-breaker, but they got me thinking about this pet peeve topic.

Hope that makes more sense,
--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

I wanted it to work “right out of the box.”

Since we are living in the world of scaleable type, one could argue that there is only one instance where a font will work "right out of the box". All other instances will require some user modifications, to word space, letterspace, kerning etc. It seems that your "out of the box instance" and Freight's "out of the box instance" are two different things. I guess that is going to leave you with an awful lot of pet peeves.

Who has the time?

James

Just a suggestion, ever tried Kievit?

Kievit is a great suggestion for out of the box play ... only minimal tweaking is required if you want to find god in the details.

Nice thread idea, Michael. The Thesis Q is something we hear about at FontShop at least once a month. I know more than one company that has created and licensed their own version of the font with a more reasonable tail.

Adding to the pet peeve pile: @-signs that sit too high. It's amazing how many type designers are still plopping the @ circle on the baseline. I assume they haven't broken free of the pre-digital era when the symbol was aligned with lining numerals instead of lowercase. The more thoughtful type users will see this and baseline shift, but it's often overlooked.

FF Max's @ is delicious, but still misplaced:

Adjusting wordspacing is a snap in Quark.

I have beefs about nearly every Emigre font, which results in my not using them that often. The a in Priori, about half the letters in Dalliance, etc.

Today Sans has the ugliest ampersand ever.

My biggest beef however is incomplete character sets. I need fractions and it drives me crazy when they only include 1/4, 1/2 and 3/4.

I love the Q in TheSans.

Actually, Emigre has been pretty good about the fraction matter. Tribute, for instance, has a full numerator/denominator numeral set.

My hope is that with the slew of their new OpenType releases, they do improve the kerning table. Maybe then we'll consider sparking another "Mrs. Eaves for body copy" topic :)

The Q in TheSans I have mixed reaction of. I love it, but it's not as flexible as the regular Q construction -- not that you would use it that much.

Personally, the absence of an ff ligature would be unforgivable. :^)

I have considered Goudy Sans for several book designs, but i think the 'A' is too silly.

@-signs that sit too high.

To me, not just the @ sign, but the copyright symbol ©, parentheses (), brackets [], and braces {}, as well as the slash / and mathematical operators, sit too high in most fonts—especially when used with old-style numerals. (In Adobe Caslon, even the hyphen, the en dash, and the em dash look too high to me.)

Parentheses, brackets, and braces also tend to have too little space “inside” them. Actually, I find even periods and commas too tightly spaced in many fonts—for example, the ubiquitous Adobe Garamond Pro (the same goes for many of the kerning pairs, e.g. Ta To Te Tu).

These are part of the reason why I would want to be allowed (by the EULA) to modify the fonts I use.

In ITC Mendoza, I didn’t like the old-style numerals 1 and 3, so I made my own versions (thought 3 didn't turn out that well), along with “correcting” many of the things that I mentioned above.

EDIT: I do understand that those of you who design type from scratch might not be too happy about the thought of someone modifying the actual outlines of glyphs (esp. when that someone is a beginner like me). But fonts aren't really works of art, are they?

But to go back to the start of this thread, I think the character spacing is too tight in most sans fonts—especially the lighter ones. For example, Frutiger Light is much less readable than the normal weight (Roman). Sans fonts, of course, have the problem that when there are no serifs to tie the characters together, spacing that is loose enough for readability tends to look too loose esthetically (so I don't like to use them for extended texts). Just my two cents' worth.

Olli

My pet peeve: the silly "I" in Bell Gothic. It's a serif "I" in an (otherwise great) sans serif typeface.

Oh, and many common fonts have terrible umlaut characters. Since I'm a Swede and my language has some very common vowels with umlauts, this is a biggie. I spent a long time looking for a Franklin cut with decent umlauts.

My pet peeve: the silly “I” in Bell Gothic

I too kind of dislike this feature (which is part of Verdana as well) at a gut level--but it was done for a good reason, legibility.

Olli

Thermopolium, what are the faults you see in the 'terrible' umlaut characters? What are some good ones?

My native Finnish, too, has ä and ö in its orthography. I can't name any fonts offhand, but for one thing, I remember seeing umlauts that are too small (and perhaps sit too low), thus looking like insignificant flyspecks. The umlauted version represent different phonemes than a and o, so the umlauts are important.

Also, Tä and Tö kerning pairs are often too tight.

Olli

I like the I in Bell Gothic. Otherwise it just looks like any old sans.

I don't like the R in Gill Sans and truly hate how silly it looks at the bolder weights.
Don't like Frutiger's R either. It's a dealbreaker for me in that font.

Caslon's italic is a bit too slanty and skinny for me, not readable.

I wish Seria & Seria Sans had a true italic, the one they have looks indistinguishable from the roman.

Martin Majoor designed a more angled Seria Italic but has yet to release it. I'll get on his case.

"Caslon’s italic is a bit too slanty and skinny for me, not readable."

Patty, talk to Bill Berkson about this. He is working on a new Caslon and the italic is one of his targets.

ChrisL

trenbuchet MS is clearly not one of my prefered Fonts, but given the rather poor choice of fonts one can use for the web you it need now and then and after all its quite okay as a groteque. But for the the Umlaute (i am from Germany) are odd in my eyes:

What makes the umlaute bad and what would make it proper? The only thing I notice, as someone who does not understand German, is that the dots seem a bit too far to the left. The resolution above is too low for me to tell if the shape is off or if it is just pixelation.

ChrisL

Chris - I am aware of William's Caslon redesign - he IM'd me about it.

I just can't pay FontBureau prices, especially not for a font I already own in several versions.

I think that Trebuchet umlaut is awful - dots should be closer together.

You are right, the example is not so good. The Dots are a bit to much on the left and a bit to high. In small size they seem far to big. Maybe it gets clear when you look at it on a website? Just a random example:

http://www.theater-der-gezeiten.de/

I jump to the defense of Lucas! (Belatedly!!) I think Thesis is an amazing, and well thought out, family. I've never had trouble using it, in either display or text settings. It's easy to work with, which is to say it's character spacing and it's kerning tables are good and do not require further fine-tuning. It's a very well designed face. (family of faces??)

Though quirky, I love the Q. Especially in text. In English, the Q occurs infrequently enough to give bodycopy (set in TheSans) more personality than, say, Frutiger. Without overdoing it.

However, now that one of my fave faces has properly been defended from you scurvy lot, I wholeheartedly agree with Michael's observation that an alternate Q would be a good thing.

(Can't comment on Freight Sans. I've never used it.)

The 7 in Adrian Williams' Monkton causes me problems. A lot of my work passes over the desks of accountants. And many seem to dislike Monkton's 7 when set in tables and charts.

So my pet peeve with this face would be *accountants*.

Are you really saying that in 'pünklich' at the bottom of the page you link to the umlaut is too large? Looks quite small to me (and properly centered). Maybe at the top the umlauts on the caps are too big?

With screen fonts so much is hinting. Could you or other readers of German give examples with fonts that are normally outline?

>I just can’t pay FontBureau prices, especially not for a font I already own in several versions.

Whoa, one tough customer! You know sight unseen that it won't be worth buying, even if upon completion it has achieved my goal of being far more readable and inviting than any other digital Caslon?

Just checking, Fontbureau's retail prices seem to be about the same as fonts in the FontFont catalogue. Mono-Linotype and Adobe are less, the smaller European founderies are more.

William - I have every confidence it will be beautiful, and if I can find a good reason to buy it I'll consider it. But when it comes to plunking down my own moolah for a font I'm gonna go with something that fills a gap in my already large collection of fonts. Fontbureau's fonts are great, and also pretty extensive but as a result I'm only gonna buy it if I can swing the whole collection, and when you add it up... big bucks. I already have a ton of text faces so nowadays I'm more inclined to buy one-off display fonts when I need them for a job. I have - and use - Adobe Caslon which has a much more readable ital than 540 (which I also have).

But I can't wait to see your new Caslon.

Another try to show you what I mean:

"I think that Trebuchet umlaut is awful - dots should be closer together."

This may be a screenfont issue. Screenfonts have to find a happy place where a 2-pixel wide and a 1-pixel wide stroke work well. This means small shifts are not possible as they are in fonts made for print. This is a similar problem to small type designed for use in phone books being enlarged for use as display type. Sometimes the specific intent of the font precludes its use outside of that arena.
I would be curious to see if the German speakers here on Typophile can point to some examples of very good and bad umlautes in fonts and tell us why this is so. The same is true for any diacritic glyph. I was quite happy to see Oscar Bjarna, an Icelandic speaker, comment on glyphs peculiar to his language in a previous thread.

ChrisL

IMHO most well known fonts have quite good Umlaute. To name an other typical screenfont: Georgia's Umlaute look fine in any size. In my expirience readability is increased if the dots-part is a bit smaller then the letter below and placed a bit lower then upper-case letters tops. A to great difference in size and placement between the umlaut-dots and the i/j-dot seems odd to me.

Some examples to judge for yourself:

'Hey', in the example you just posted I see the effect you wrote about: the umlauts do look too big on my screen also.

Another issue is whether the dot on the i and the umlauts should be the same height. If I'm not mistaken the dot on the i is often higher. Looking at Georgia in my post, in 'pünklich' they seem the same height. Surprisingly to me, in Trebuchet in your example, the umlauts are bigger and higher than the i dot.

edit: we cross posted 'Hey'. Thanks for the info. I'll look for some examples myself.

An obvious thing, I can't stand faces with not lowercase numerals. Also dumb looking lower case "a" like in Trebuchet (extra big bowls) or Helvética (mannered upcur) make me think twice about the face.

Héctor

The dreaded old style O (Zero, Zed). Especially in typefaces that aren't monoline.

The upside-down question mark in Stempel Schneidler. I had to have it redrawn for me.

http://www.myfonts.com/fonts/linotype/stempel-schneidler/roman/charmap.html

Here are some example of o-umlauts with dotted i's. Do you German speakers think that the difference in Galliard's heights is too extreme or OK?

If I'm not mistaken, you never have an i and an umlaut next to each other in German. Is that right? What about other languages?

edit: I just saw your graphic--did you post it after the text? The ninth in the list has a considerably higher i dot. Is that one Ok to your eyes? Galliard?

I think that Galliard is just at the limit. I might use it for small amouts of text, but certainly not for long texts in german. There are some words with i next to an Umlaut (i.e. böig) but not many.

Thanks, Philipp. The reason I ask is that in the original Caslon a high, heavy dot on the i and j is part of the 'look' of the face. For my Caslon I have to decide what to do. Of course Caslon himself was producing for England and its colonies so the umlauts probably weren't an issue.

One possibility might be an alternate i with a lower dot for setting German language.

As far as I'm concerned, Lucas needs little or no defense... :-)

I just wish I had an alternate "Q" to work with. To my eyes, everything else... and I mean EVERYTHING else... about the entire Thesis family is perfect. I just found it unsuitable for a "Q"-heavy project.

OK, onto more pet peeves...

1) The lack of a true italic for Syntax. (Why pair an oblique with a humanist sans?)
2) The above-cap-height ascenders & below-baseline descenders of Eidetic Neo Omni (Rodrigo might kill me for saying so... I'm one of the people who suggested he create a unicase version of Eidetic way back, but unicases don't usually have ascenders or descenders...)
3) The too-narrow-for-my-tastes italic of Joanna.

Happy we're back on track after my 30 lashes for H&J laziness... :-)

--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

Michael,

Shhhh! Don't remind them :-)

ChrisL

The reason I ask is that in the original Caslon a high, heavy dot on the i and j is part of the ‘look’ of the face. For my Caslon I have to decide what to do. Of course Caslon himself was producing for England and its colonies so the umlauts probably weren’t an issue. One possibility might be an alternate i with a lower dot for setting German language.

I think in your case i would prefer to keep the typical i-dot in german too. There seem to be room to place the Umlaut-dots a bit higher without getting them to high.

edit: I just saw your graphic—did you post it after the text? The ninth in the list has a considerably higher i dot. Is that one Ok to your eyes? Galliard?

Yes I posted it afterwards. The ninth is Day Roman and I think here again that the difference is too big. Still suitable for small texts though but not for long ones.

Thanks, all, for posting images of diereses and i's together and discussing them. This is useful information for a type designer, book designer or graphic designer, since we all are likely to deal with multilingual texts going into the future. Does anyone have comments about the height of accents over capitals?

It's little things such as have been mentioned here, which are deal-breakers for me in purchasing a type family for general use; if I spend a few hundred dollars for a tool, I want it to work all the time, as well as it can. As James points out, no type works perfectly in every setting, at every size, without some modifications. But individual letters shouldn't bother the user.

There were problems with every humanist sans I knew about, when I started drawing Mundo Sans; the obliques in so many (Syntax), irritating spurs and curved terminals (Today, Gill Sans), and other things like the narrow oldstyle 5 and unjoined Q in Thesis. Some were spaced obviously for display, some had proportions that wouldn't work below 24 point. I wanted a family that could work at a lot of sizes; Syntax starts to look weird at display sizes, as does Gill, and others like Formata look regimented and square in text sizes. Gill's weight progression is bizarre; the mismatch of weight and contrast between the caps and lowercase always bothers me, as do some of the wide caps. Mundo is meant to have a wider range of sizes it performs at, without any distracting bits. Unfortunately you can't space a typeface to look good at every size; the only typeface I think does this well is Frutiger, and some weights are spaced tighter than others.

Yes, I'm very very picky. If I hadn't drawn Mundo, I'd probably be using Kievit.

Does anyone have comments about the height of accents over capitals?

In a displayfont anything that looks good is fine to me. In continuous text I would prefer to keep them low and as flat as it is possible while preseving the typical character. If the accent takes too much space it can look realy bad. That is the reason why in some languages accens on capitals are optional (i.e. in french).

Carl, Mundo is downright gorgeous. Nary a peeve to be found.

--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

Great thread Michael. I agree with what you say about Syntax and Joanna (although the Perpetua italic is pretty darn narrow too – it was just the Gill style).

The weight progressions in Gill Sans are ludicrous, and in a (yet unpublished) piece called 'Eric Gill got it wrong' I've written a critique of how self-contradictory Gill was in the context of the sans' weights and the 'Essay on Typography'.

To respond to your initial point, I can still remember being hauled out during a class critique many, many years ago, for using overgenerous word spacing; therefore I completely support the word spacing arrangement in Freight Sans and see no problem with it. I also sympathise with some of what is said in the chapter called "The school of close spacing" in Ari Rafaeli's 'Book Typography'.

Other pet peeves? Most of them have already been listed, like the wonky italics in Caslon, but most of all, my peeve would be about the complete lack of balance or grace in the lowercase of Rotis. It makes no difference if one is talking about the sans or the serif or the semi of either – the fault is in the construction. The 'e' and 'c' look like they are falling over backwards.

At the Spiekerblog they are talking about it looking OK as signage – but here in Auckland it is the corporate face of the town council and used for a lot of the municipal signage. It looks plain awful in every application.

Pet peeves?
--I love old style figures, so I get enraged when Adobe converts a family to Open Type and the new version STILL has no old style figures! What's the point?
--in some fonts with old style figures, some numbers, like the 3 in FF Acanthus, come out like liner figures instead of dipping down demurely as they should.
--Hoefler Text used to display so very badly on Mac, but that's been corrected in Quark 7, thankfully.
--I know it's churlish to say so, but for freelancers who pay for their own fonts, the prices add up to fancy money rather quickly! How about a deep discount for lone cowboys, the way books are discounted for library purchase?

A bit of a peeve, though more like an unfilled niche:

I'd love to see a beautiful sans serif version of Symbol, or really any good solution for typesetting mathematics with a sans.

Typically, I use Frutiger with MathType, but it just isn't that pretty. And with Greek characters, etc., I end up having to import from MyriadPro Italic.

Any suggestions?

--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

Michael, your pet peeve is a daily aggravation for me. I do a lot of intensive math and engineering typesetting, and we're forced to pull characters from about a dozen different fonts to set everything we need. Rarely do any of the special characters mix well with the text fonts (in terms of style, size, or spacing), which have almost none of the math characters we regularly use.

Technical stuff presents specific problems that have caused us to switch typefaces over exactly the kind of details everyone is talking about here, since characters so often appear in formulas or by themselves, rather than within normal words. If the lowercase l looks too much like the uppercase I or the numeral 1, a typeface will never work for most of what we publish. Or if the lowercase x looks too much like the times sign, etc.

So many headaches...

My biggest pet peeve with common fonts is:

It is nearly impossible to know if the design you're working on will become a common font.

Randy

Daniel (ultrasparky)

I see your live in Brooklyn, so do I. If you want to put together a brief on what you desire in a math font, I'd be happy to discuss it with you. Always looking for a problem that needs to be solved.

James

James, that's an awesomely nice offer, but I'm a few days away from leaving Brooklyn to go to Reading and start working on that very problem (and many others) for the next year. Can I seek your advice when I'm a little more deeply immersed in my thesis?

Quote: Adjusting wordspacing is a snap in Quark.

PATTYFAB, could you take a moment to walk us through just how to adjust wordspacing in Mac Quark 7? I've been looking in vain for the right place to click for that. Thanks...

Any time Daniel.

James

A quick suggestion, Daniel & James.

I almost think that an effective software solution for mathematical typesetting has to go hand-in-hand with new OpenType fonts that cover tons of bases. The MathType-esque software is useless without good, homogeneous font families and the fonts are useless without the right software.

Whatever you do, please hit the ball out of the park. :-)

--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

William Berson:

My main gripes are the following. I prefer umlauts that harmonize with the dot above the "i", they may be a bit smaller but not tiny. Others may think differently.

I dislike:
* fonts where the umlauts don't have the same basic shape the dots over the "i". For instance, in ITC Frankling, with round umlauts and a square "i" dot.
* also fonts where the umlaut dots/squares are tiny and not of the same proportion as the "i" dot. Example: Bitstream Franklin.
* fonts where the umlaut dots are placed too close together. Like Bitstream Franklin, again.
* Typefaces where the umlauts on the capital O sit like little ears almost on the sides. Like Typeshop Franklin (simply impossible to use).
* as an example of great umlauts in a Franklin, check out the Eisner+Flake cut. Now that's good umlauts and my favorite Franklin.

That's just off the top of my head. In fact many great cuts are ruined by poor umlauts.

I think it's pretty obvious that Swiss or German designers (languages with umlauts) pay more attention to this whereas for instance English- or French-speakers often do not.

Thermopolium: So the i dot being higher than the umlauts is not a problem in your eyes, just the umlauts being too small?

The strange "baskerville"-g in Postscript-FF Meta, especially if set in display-sizes. I was not the only one... The new Meta has an alternative g to offer.

William Berson wrote: "If I’m not mistaken, you never have an i and an umlaut next to each other in German. Is that right? What about other languages?"

Well, yes, in Finnish you can have i and umlaut next to each other, eg. öitä, äiti, äimänkäki, mäihä...

To the people with problems concerning mathematic placement in technical documents;

wouldn't LaTeX (or any other TeX parser) be a solution? Though it has a steep learning curve, I find it great for typesetting anything with a lot of mathematical equations. The only big problem would be customising the look of the final document, as they usually look like university dictates.

Hey Thomas – nothing wrong with that FF Meta lowercase 'g' IMHO. It's just part of the charm, especially at display sizes. But then I think Michael knew, with the original post, that one man's peeve is another man's perk (or indeed, that one man's perve is another man's pique!) – which is why he framed the question that way.

But then if you're picking holes in Meta you'd probably also claim that its punctuation is much too big...

As for could you take a moment to walk us through just how to adjust wordspacing?

I was trying to figure out how old a version of InDesign Michael must be using in order to avoid his 30 lashes for H&J laziness: the following screenshots are from InDesign CS and QX6 respectively (I haven't had the chance to look at QX7 yet but I would assume you'll find this in the same place)

In either case the mechanism is essentially the same and the resulting H&J can be incorporated into a specific style sheet within minutes – as Patty said 'it's a snap'.

ID2 has the same H&J settings... it's not the version that was the problem, it was my quick-and-dirty typesetting that day: set size & leading and move on.

Of course, when I have more time, I am in the habit of adjusting H&Js...indeed, it's a snap. But very often, I must confess that I simply grab a face and go.

Does that make me a hack? Perhaps. But I feel that many faces work well in text settings without custom H&Js. What does everyone else think? Should faces "require" H&J adjustments? Do most faces require them anyway, regardless of whether they should?

--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

Hey Michael

Does that make me a hack? well you said it, not me. I think you just answered your own question about pet peeves. Mine is laziness. Make that 90 lashes ye swab!

Michael:
You asked: `... really any good solution for typesetting mathematics with a sans.'

Math typesetting with a sans is quite unusual (at least in print --- about the only time I see it is in slide presentations). Some reasons for this were discussed in Johannes Küster's presentation on `Fonts for Mathematics' at ATypI 2004. A pdf of that (excellent) presentation is available at:

http://www.typoma.com/publ/20041002-atypi.pdf

The only two (complete) math sans fonts that I know of are
* HV math (a sans `inspired by Helvetica'), and
* Computer Modern Bright (designed by Walter Schmidt),
both marketed by MicroPress:

http://www.micropress-inc.com/fonts.htm

Both of these come with LaTeX support; I have no idea how difficult it would be for you to use them in MathType (but at least they have a complete AMS symbol set, which should be adequate for almost all math typesetting).

Grant

Thanks so much for teh great links, Grant. I'll check 'em out.

--Michael.

------------------------------------------------------
// love what you do or do something else. //
Michael Ebert -- graphic designer, jazz saxophonist, horror movie devotee
------------------------------------------------------

Sorry to keep raggin' on the "Q" in Thesis, but I completely agree. There should be an alternate, I also am not a big fan of it. But, this leads me to my typeface pet peeves.

1) Too few sans serif typefaces have "small caps." When setting copy or used to display many levels of information, this is a necessity for me.

I use Thesis a lot of the time for this reason. Zwo and Meta are the only others with small caps I can think of off the top of my head.

2) Too many sans serif typefaces don't have true italics. The sloped roman is simply not enough when setting a large amount of body copy.

3) The number 7 in Meno, too weird, almost a deal breaker.

4) I'd like to see more experimentation in the design of serif typefaces. While there are always exceptions, there doesn't seem to be enough experimentation in the design of these faces. Classics are beautiful, don't get me wrong...but I'd like to see some more modern approaches to serifs. Stormtype, Emigre and Font Font seem to be pushing the envelope to me along with a handful of smaller foundries.

I have one comment how many times do you use certain charaters like a cap Q. Give me a break, just look at the most used characters and make decisions based on what you really see not imagined. I bet if you took all the characters on this thread you would have about a 1% cap Q's and thats because it was mentioned. That character might be used less than one tenth of one percent of the time in a document.

I think Dan's right about this (and it crossed my mind to mention this at the initial rag on the Thesis Q). If it was such a problem one might have come up with a custom Q (either drawn or set as a compound character) and done a search and replace within the text to get rid of the (one or two?) dodgy Thesis Qs actually in the text.

The only time an idiosyncratic Q caused me a problem was at the beginning of my career. The job was done in Queensland, Australia so I should have thought about the incidences of Qs in the text beforehand, but in my naïvety I went ahead and set the whole thing in Bookman. Now there's an ugly Q...

But it turns out I was the only person who had a problem with it at the time. It still makes me shudder to see it.

The lowercase "g" in both Meta and Meta Serif always bothered me. I'm not a big fan of it in Scala neither.

And I'd love to see a slimmed down bold weight of Dolly!
Proxima Nova's "a" makes me think twice.

@frode frank
The is an alternate a available for Proxima Nova … "futura-a style".

poms:

The way it slims down towards the stem and how the top of the bowl curves down just bugs me. It doesn't look right. Having an alternate doesn't really fix my issue, cause I'd wanna use a two story "a" in such a beautiful face.

Ok Frode, then Marc Simmonson will loose a potential client. I don't like the two storey a either. The a is such an dominant letter that makes it sometimes nearly impossible to go with a typeface. I like the a in vista sans, or malaga – others don't.

I like a capital "R" that sticks its leg out and tries to trip the next letter, like this (Albertina) one:

This (Utopia) one is too polite:

Brian

For those still keeping score, the Q's of Luc(as) de Groot's TheSans:

a pet peeve of mine: The almost perfectly geometrically round oldstyle zero in Sabon, with no hint of weight distribution.

Designers sometimes (alright, all to often!) try to make them different and quirky for the sake of doing so. It's better to make something good than to make something original. Just browse around Dutch Type Library's and Enschedé's offerings. It's occurred to me more than once that DTL Documenta and Documenta Sans is the only font family I'll ever need. A bit daft, I know, but no one ever got fired for using Documenta.

Take Dederon Serif. It's *actually* advertised as a Book font. Are they serious?

Ludwig Übele's Mokka is another example. Lately it's become fashionable for some reason to shape the top-most serif of the letter "a" into a sharp, downward-pointing beak. The reasons for this escape me, although it seems like an artful attempt at style. It doesn't aid reading, nor does it make text look good. Skolar seems to suffer from the same non-psittacine beak disease. Another example of what appears to be artfulness for its own sake is the bowl of the letter "a" in Mokka.

I agree with pattyfab 6.Sep.2006 8.30pm

Often Emigre doesn't style link:

We don't do style link by default because many users prefer to select it themselves or the fonts don't fall into a traditional style. We can do that as a *custom job* if you like. —Tim Starback, January 26, 2010

I think that's crap, especially for mere mortals who use word processing programs.

Cool thread.

Patricia mentioned ampersands - let me elaborate on my pet peeve
about them: the "E"+"t" style often renders an otherwise great font
unusable. Why do some designers insist on being anally historical?

hhp

"Å"s with the ring glued to the apex. Like this (Nobel):

Or this (Benton Sans):

... and before someone finds out that I just designed such an "Å" myself: I've changed it.

But Sindre, quite often the saving of
vertical space there becomes critical.

hhp

Oh yes, I know. And while this is almost acceptable for a display face, it looks very clumsy in text settings. I'd say that Benton sans doesn't need this. Many Scandinavian newspapers use altered "Å"s for their headlines, with a half circle cut into the top of the "A" and the ring then lowered, without touching. Like this Swedish crap tabloid:

(Yes, this is awful, but that's the first example I found.)

Michael:

About the Q in TheSans.

Style for the sake of style. In most cases it just doesn't work because what often ends up happening is that the offending letter has nothing to do with any of the others.

But you can't ignore style. And the "Q" is a great
place to express it because it can't harm much.

hhp

"...Like this Swedish crap tabloid"

I think you have defined the problem with your description. All tabloids which do the celebrity gottcha thing are universally crap. It is just there nature. If their content is from the toilet, so is there typography.

@Satyagraha

Looking at that tabloid makes my eyes hurt very much.

Er ... of course. My point was this and only this: An "Å" with the ring stuck to the body is so alien looking for us who actually use that letter (remember, this is not an "A" with a diacritic, but a letter in its own right) that we rather cut a half-moon out of the "A" part to make room for the ring. And yes, quality newspapers too use this solution.

Those ultra bold headline faces, set with negative leading, could have their cap accents knocked out of the letters.
Are there any typefaces that have that feature?

In Polish, the cap acute accent is sometimes run across the letter.

Nick, there are a few of those in the Fancy Diacritics Flickr Pool, e.g. this Charlie Brown Umlaut (name credit: Stephen).

URW did this in their ‘Poster’ versions of heavy display typefaces, like City, Gill Sans Kayo, Antique Olive Nord or Odin:

"Often Emigre doesn't style link … I think that's crap, especially for mere mortals who use word processing programs."

But they do have a point. What’s good for some is not good for all. Style-linking isn't great for Adobe CS users, or those who use a Regular, a Medium, and a Bold for instance. The solution — inelegant as it may be — is to offer separate versions to please both camps. This is FSI’s goal with Office FontFonts.

Sindre (satyagraha) -- Question: Is that opinion about the Å being separated widespread among readers in the various languages that use this character?

For instance, I had heard once from a Dane that the preferred form of Å is merged.

When designing Whitman Display, I was directed by FB to have the ring merge with the A. (Whitman text, on the other hand, has the two separated.) I didn't argue because I could find no definitive answer.

I'm most familiar with Swedish, and I've only heard that Å should be separated. But I suspect most people don't really care either way.

One of my favorite logos is a Swedish company with a combined Å.

With a combined É as well, apparently. ;-)

My suspicion is that Swedes and others have become so used to seeing such stylized combined Å in logos and imported typefaces that it wouldn't bother them that much, even in text faces.

That kind of "Å" is perfectly all right for logos and stuff. (I too like the Åhléns logo). But for text use, I'd say it's no good. I asked several people at work today (two journalists, three editors and one typographer). Five of them thought the merged versions looked odd or ugly (I showed them Benton Sans and Nobel), while one of the editors thought the Nobel "Å" looked cool, but the Benton one looked weird. Neither of them had thought about this before.

Perhaps the Danes are more open to weird "Å"s because they started using it much later than we did (1948 vs. 1917, but some Norwegian writers, among them Henrik Ibsen, used it from the 1890s on).

That scarlet A looks like it's ready to hang around Hester Prynne's neck!

I definately prefer Å unmerged, but would accept embedded one in poster/headline use or merged in logos.

This is a sample I found some years ago, it's a font (I tried to get it ID'd here with no luck). I'd prefer embedding to this treatment.

Mili, that looks like small caps. Is that what they did to all the cap diacritic glyphs?

Chris, I believe the font is like that. It repeats in all such headlines in several issues of the magazine, and there are only Äs Ös and perhaps Ås, as it's a Finnish magazine. It would be too much trouble to do it by hand, I guess.

Thanks, Mili. Those short letters just look so apologetic compared to their full-sized neighbors. Kind of like Ben Stein standing next to Shaq in all those commercials :-)

"Evolution" of Swedish tabloids!

-

-

-

I'm still not seeing any with accents reversed out of the glyph.
Surely someone has given it a try?

I'd guess /Ä/s (and /Ö/s?) would look too much like faces that way.

As opposed to Mickey Mouse?

I don't have a huge problem with the merged Å, and I honestly don't understand how merging the ring and the A makes it less a letter in its own right than taking an A and placing a ring accent above it. But maybe that's just me being Danish. For the record, I know a few people who despise the å and insist on the old spelling, aa.

To me, it's a lot more repelling when lowercase æ looks like a clash between a and e (example: Bembo), instead of being drawn as one, unified letter.

Nick, Florian has a couple of «inverted» diereses in his Flickr stream:
http://www.flickr.com/photos/hardwig/3682168491/
http://www.flickr.com/photos/hardwig/3575554309/
http://www.flickr.com/photos/hardwig/3576358870/
(There might be more. Also see: http://www.flickr.com/groups/fancy-diacritics/)

Well, I have a handful of pet peeves:

01. Kerning too loose for digit 1 in Minion.
02. Oldstyle 0 is almost a circle in many Galarde and Venetian typefaces.
03. Oldstyle numbers too much out of x-height in some serif fonts.
04. Small caps with the same height than lowercases in many serif fonts.
05. Absence of small caps in Utopia italic and Cartier Book italic.
06. Absence of small caps in the whole Myriad family.
07. Absence of small caps and alternate numbers in Iowan OldStyle.
08. Position and size of acute and grave accents over lc in Meta Serif.
09. Lowercase m is much condensed in Meta Serif.
10. Lack of a light style for Tisa.
11. Terminals very big in a, 2 and 3 in Greta Serif.
12. Lining 5 in Arno and Sabon.
13. Uppercase U in Bembo.
14. Dollar in ITC Caslon.
15. Question mark in Futura.

> 07. Absence of small caps and alternate numbers in Iowan OldStyle.

http://new.myfonts.com/fonts/bitstream/iowan-old-style/small-cap-bt/
http://new.myfonts.com/fonts/bitstream/iowan-old-style/roman-bt/characte...
http://new.myfonts.com/fonts/bitstream/iowan-old-style/roman-osf-bt/char...

Sorry, I did mean the absence of small caps as an OT feature.

Now I see there is an Iowan Pro version which must have proper small caps inside the fonts. But strangely its information is incomplete both in MyFonts and in FontShop. Tries to show smcp/c2sp preview always return errors and the character set does not shows small caps (http://www.fontshop.com/fonts/singles/bitstream/iowan_old_style_bt_pro_r...).

Does anybody know if Iowan Pro has small caps and old style figures as an OT feature? Is this available for the whole family?

Thanks Nina, that's what I suspected.
Perhaps some of the present wave of ultra-blacks also have such a feature.

Regarding Freight in it's Sans and Text variants: I, inter alios, have to say that they're pretty much unusable for German. The esteemed Mr. Darden decided to not design an ß for them, but to prove instead that he knows about its ancestry, the Antiqua's ſs ligature (an assumedly younger, but still century-old variant of the Textura's and the Fraktur's ſʒ ligature which gave the ß one of its names, the Eszett). No modern ß is provided as an alternate glyph.

The other Freight family members (Big, Display, Micro) do have an ß, but for some curious reason, Micro's ß changes into ſs in its italic, so Micro falls into the unusable-for-German-unless-you-don't-plan-on-using-the-italic category.

This inconsistency alone would make the Freight family an iffy choice for German, even if the ſs ligature wasn't completely out of use. While the elderly or typographically educated may recognize the ſs as ß (and yet trip over each instance), the rest will just wonder Whatthefont this is (and/or read it as fs).

I've heard the ß being referred to as an ſs ligature, but it's more precisely a single letter that evolved from the ſs ligature. Freight illustrates the point.

A good example of the danger in historical pedantism.

hhp

For the record: The Pro versions of Freight Sans and Freight Text contain what you would probably consider to be a decent ‘ß’ glyph.

Hey R., thanks for the setting the record straight. I never got aware of the Pro versions after I had to axe Freight from my list years ago because of the Eszett issue. Before I wrote the above post, I did visit dardenstudio.com, but no mention of a Pro version of Freight catched my eye (there's is Omnes Pro), so I didn't go over the character set on the site. It seems that Freight Pro is now the only Freight there is (Phil's Fonts, though, who sell the font licenses, list the fonts expressly as Freight Pro). Anyway, good to know there now is a perfectly fine Eszett in Freight Text and Sans (and also in Micro Italic).

PS: Hmm. When I set a sample on dardenstudio.com, the Eszett situation is the same as I described in my yesterday's post, except that Freight Display shows an fl (sic) ligature instead of an Eszett. Wait a second ... most of the accents or umlauts that I'm trying with any of the fonts are more or less messed up ... looks like the font sampler thingy is to blame. Oh well, as long as the actual fonts work.

"an fl (sic) ligature instead of an Eszett"
That seems to be a not-entirely-uncommon encoding (?) issue (see this find).

Speaking of eszetts, to many Germans Dolly has a serious German-usability issue too, because its eszett is too short (it should really reach the height of the other ascenders). I myself have only ever been able to use it in Swiss contexts, where no eszett was needed.

I can see a good reason to make it short (it's heavier
than anything else up top) but that's way too short.

hhp

Yes, it's tempting – I've been there too. But that there just looks wrong.

It does appear to be intentional.
At first I thought it was a mistake -- matching the eszett to cap height, rather than ascenders.
But the descender is also a lot shorter than other descenders.
Perhaps if it had been designed by a German it would be more acceptable :-)

You don't have to be German, you just have to listen.

hhp

It looks odd within the context of the typeface, not the language.
The Souvenir "g", for instance, is odd, but not in an English-usability manner, because of glyphs like "y" which are similarly strange.

So in the case of Dolly, it is not a German-usability issue, other than the problem character only appearing in German text, because the glyph is clearly recognizable as eszett (unlike the Freight glyph mentioned earlier).
And the eszett character is one that adopts many different forms in the antiqua genre, so German readers shouldn't be fazed by a non-standard appearance.

**

I always found the f and t crossbar of Palatino and Laudatio to be unusually low.
Is that related to any German calligraphic traditions, or did I just make that assumption because I know the designers are skilled calligraphers who incidentally hail from Germany?

"it is not a German-usability issue, other than the problem character only appearing in German text"
?
What I meant is: I know a number of Germans (it's not just me) who say Dolly is essentially not usable for German-language text, because the eszett glyph does not work. And when I say does not work, I mean makes readers stumble. Sure it's recognizable when you stare at it, but it sticks out when reading, it seems «wrong».
There is such a thing as dysfunctional glyph design.

"And the eszett character is one that adopts many different forms in the antiqua genre, so German readers shouldn't be fazed by a non-standard appearance"
Well that's really a little too easy.
Lowercase "g"s can adopt a number of different forms too, but that doesn't give you a wild card to do any odd thing you like, especially not if you expect people to actually read it.

> the glyph is clearly recognizable as eszett

That's not a display face designer talking.

> German readers shouldn't be fazed by a non-standard appearance.

That's not a text face designer talking.

hhp

What I meant is: I know a number of Germans (it's not just me) who say Dolly is essentially not usable for German-language text, because the eszett glyph does not work.

Count me in.

I think most people started mixing questions of actual problems in usability and simple taste preferences or dislikes. This does not make sense.
The [Q] of Bookman is perfectly fine. The [Q] in Scala may present problems of immediate recognizability, since it‘s too detached.
But as Hrant said, its style should be preserved.

The serifed capital I in Verdana.
I can understand its use in small pixel text, to disambiguate I from l and 1, but in all other uses, it looks like a "wrong font" letter.
Especially in all-cap settings, e.g.

How about the cap "I" in Bell Gothic?

And Griffith Gothic?
I don't mind that so much, because there are other strangenesses.
(That's why I didn't mind including a serifed "I" in a couple of my sans designs.)
But everything else in Verdana is so orthodox, notwithstanding the occasionally J/j, that the "I" really stands out.

Yes! The open bowl in the P of Stone Sans is why I hate using it.
Openings in 2-story lowercase g's bug me too.

Many designers have become used to the conventional closed-botton binocular "g", and they assume that what jumps out at a designer at display sizes is the same thing as what harms readability among layman at text sides. I don't think those are the same at all, and I believe the open-bottom bino form is more natural to reading. It's just harder to make well because good precedents are so rare.

hhp

Nick: "...everything else in Verdana is so orthodox..."

There is a stylistic alt I and small cap alt I in Verdana Pro, though not labeled "orthodox."

>>though not labeled "orthodox."

perhaps "reform", then?

> I don't think those are the same at all, and I believe the open-bottom bino form is more natural to reading.

I'm not sure it's more natural or not. But I have seen it done very well with a very natural feeling. It's certainly more natural in certain cases.

As with all things, the proof is in the pudding.

I believe it's more natural because the closed form is topologically too alien. The open form harmonizes with other forms better (like the "s").

hhp

Proof is never in the pudding. The proof of the pudding is in the tasting.

You know, I've never-ever-not-even-once had any difficulty reading a letter g with a closed lower loop, nor found any open looped g easier to read. Some of the latter I've certainly found distracting because, ironically, they are topologically alien to most of what I've spent my life reading.

Why not, for that matter, declare the s to be the topologically alien one? It's terminals seldom correspond to those of the lowercase c, with which one would expect it to share such features. Or maybe its the c that it topologically alien. Or maybe its that none of these things have any measurable impact on reading.

"I believe the open-bottom bino form is more natural to reading."

Perhaps this is simply your own particular take on the issue. I have never heard of this from any other source. Do you have any data?

Measurable? Data? Sorry guys, I forgot I was on soilmechanicsophile.com...

hhp

n, o, and x I'd keep. The rest are just trouble. I mean, all the rest of Unicode.

No... Wait, 0 and 1 are okay too. I'd keep o, n, x, 0 and 1.
And caps, i used to think H is okay too, except that crossbar is totally ridiculous and has less than nothing to do with the O,o,n,x,0 or 1. But I guess if I think 0 and 0 are okay, O is too. Dotless i too.

And cap I in, but just in sans. The rest just don't make any sense. Then in serif fonts, just the o and x. The rest are too knarly to design and we should just stack and mix o and x to make the other letters.

No one would hate any fonts then and it'd probably not have any measurable effect on reading for most people as long as the pictures are still there?

I think that's his way of saying that he likes the open-bottomed |g|.

Actually David has previously said he much prefers the closed one, and I doubt he would change his mind. But I hope I'm wrong.

John: I will make time to properly address your points. It's complicated.

hhp

My primary recollection of what I've said previously on the "some-stencil" issue is that I prefer the bowl to be closed when I make a P. I love a good Baskerville as much as the next good shepherd and no blanket should be laid over the issue of topology until one is dead, I think.

Little white spaces in metal type "scaled" exactly the way the type designer wanted them too, (at a different rate than big white spaces), and thus are resolved for the user. Little white spaces in digital type (for most users), scale out of the type designer's control, making some-stencil a non-trivial issue to resolve. If a type is only for display, it matters less, stencil away, drive a Ford over it, use XP.

But I don't think some-stencil g's are rare because they're harder to make, so much as they are harder to use well. That little inlet at the bottom should NOT sparkle. Stencils broaden at small sizes to appear as a white counterpoint to, e.g. a serif... narrow as the size increases to remain well integrated to the eye.

The harder that is made by technology, the fewer of these, Didots, grass scripts and other extremely humanistic typographic elements we'll see in a culture going fast forward, I think. But rejoice, we have a full alternate explanation of the g-ness of the species coming soon.

More soon (gotta finish writing my Granshan talk description first), but for now: http://typophile.com/node/19013?page=1
David's second and third posts there are what I was remembering.

hhp

David bowled me over, double, he is a g-ness when it comes to turning a phrase.

David bowled me over, double, he is a g-ness

Though don't you think some of these recent posts get a little loopy?

"...recent posts get a little loopy?

Just our way of joining in ;-P

FYI: Underware’s recent library upgrade brought a new eszett for Dolly, among other things.
http://www.underware.nl/blog/2013/06/behind-the-font-upgrade/

It's great when a foundry listens to criticism.

hhp

Ah, yes! The new Dolly is a lot better in other respects too. I always thought it was a tiny bit too soft and cuddly, now it looks subtly tightened up and more grown-up all over. Very good job!

I'm looking forward to having similar feedback. It would be great if someone speaking Livonian or Northern Sami or Azerbaijani (or German, for that matter) had opinions on my attempts. A type designer that ignores (valid) criticism from users has no future.

Yay, that’s very cool. Thanks Underware!

My pet peeve: the silly "I" in Bell Gothic. It's a serif "I" in an (otherwise great) sans serif typeface.

Keep in mind that Bell Gothic was originally designed for telephone directories. Also keep in mind the existence of CamelCase names. Once I came across the name "DiIorio": that's dee-eye-eye-oh-ar-eye-oh.

Personally, I hate that "I" and "l" collide in too many sans-serif typefaces to list.

As for confusing characters: I hate the "0" (zero) as it appears in the modern take on oldstyle figures. When I saw "1950s" (or something similar) in a book once, I parsed it as "195 os" before inferring the meaning from the context.

Note to typographers: The outline for a proper oldstyle zero as used in old books is a perfect circle (circle, not oval or anything else) inside another perfect circle. That's it, that's all.

I have a feeling that no matter what you do as a typeface designer is wrong. Kind of like this:
http://en.wikipedia.org/wiki/The_miller,_his_son_and_the_donkey

You just have to be wrong to the right people.

hhp