I can't read it at all. That is probably because it's not in english.

Hmm, does feel distracting. It's a beautiful face and seems like it was made for text, but readability is probably it's downfall. It's so pretty that it's distracting.

There's a very simple reason: x-height too big.

hhp

I agree with Hrant. The x-height is too big.

What slows me down reading the text are the
letters g and f. The descender (lower case g) seems really small.
The same with the lowercase f, the top of the
ascender (sorry, don't know the right term)
is squeezed in

I can't tell how your example was printed, but when I think of too-light fonts from that era my suspicions go to a design intended for letterpress printed otherwise. I read that Frutiger developed Meridien for Lumitype but also had a metal version cut at the same time. Maybe the design is accordingly a compromise?

In those eras of overlapping printing technology, when fonts were released simultaneously in multiple formats, were they adapted accordingly or more often executed from the same drawings blindly?

Thanks for your input guys.
The x-height didn't occur to me, but it really looks quite huge.

Craig, I was suspecting some influence of a printing technology shift too. I presume this is standard offset printing from digital, and the type does feel quite fragile. (Though maybe that's just me preferring sturdier/darker types.)

Jörg: The "f" is slowing me down too. And I stumble at every "a" (does it feel wide? Or maybe it's the top drooping down).

IIRC Frutiger Serif (the new Meridien) has a better "f".

hhp

Oh, that's right. And Frutiger Serif seems to be closer to the original metal Meridien (see page linked to below).

"Linotype’s Type Director, Akira Kobayashi, began by returning to prints made from the original metal type version of Meridien. In his opinion, the phototype and digital versions had been drawn too wide; the original phototype and digital Meridien had been optimized for 10–12 point text, but even these sizes are not really small enough! A typeface named Frutiger Serif would have to be able to function in smaller sizes, too."
Source: Linotype

It looks lovely..but only at the size of the sample..blown up. Interesting about the "new" version. I think the x height wouldn't be an issue if it wasn't for the sharpness and thinness of the fonts..esp that f. Maybe more suited to display the digital version is? (oh God I'm talking like Yoda)

I have never liked Meridien. For a start, it always looked unpleasantly spikey to me, as it does to Nina. Looking at the interesting link to Linotype, I think it was also the width. Even the new Frutiger Serif, while distinctly better, for me still doesn't work. There is still something off about the rhythm of it.

As to the x-height, I also think it is too big for a book face, but for other uses, such as tightly packed narrow columns, it might work.

It is interesting to compare this with Rocky, which is also a very spikey face, but works beautifully, in my eyes. I think it is because Carter varies the weights of the spikes well.

IIRC John Hudson has written that he's an admirer of Meridien, so perhaps he'll come on this thread and tell us of its virtues.

> A typeface named Frutiger Serif would have to be able to function in smaller sizes, too.

?
Sounds like classic post-rationlization. They didn't want to reduce the x-height, but realized it's only useful small when it's loose (display when it's tight), but found a funny way to admit it.

> I think the x height wouldn’t be an issue if it wasn’t for the sharpness and thinness of the fonts.

On the contrary, vertical proportions are at the heart of reading.
http://themicrofoundry.com/ss_read1.html

BTW, another admirer of Meridien is actually Carter.

hhp

Hrant, do you remember what Carter liked about it? The individual characters Meridien are, of course, very well drawn. But it just doesn't work together for me.

Rocky is only a display face, so the demands on it are different, of course.

"They didn’t want to reduce the x-height"
I'm guessing that came with trying to make it "Frutiger Serif"? At least the x-height vs. cap height ratio seems compatible:

http://themicrofoundry.com/ss_read1.html
Thanks for posting this (I read it before, but it keeps making more sense on re-reading). I think I'm going to show my client some blurred samples. :)

For the record: I like Rocky a lot. But I've only used it for display.

William, I don't remember exactly - I think he admired its atmosphere and letterwise craft. I'm pretty sure it was in "Types Best Remembered / Types Best Forgotten" by R Norton that I read it - a must-have book for the true blue typophile.

Nina, in that setting the x-heights are indeed the same, but thanks to a much smaller talus* Frutiger Serif in fact has a smaller relative x-height. However, it needs more leading, so people might end up setting it smaller (a good thing here) so the two become closer in a way. What's very useful though is that they seem to have equalized the apparent sizes of the two fonts (at the same point size). Maybe this is something Adrian Frutiger did originally? It's difficult to give Kobayashi too much credit however.

* Also called internal leading.

BTW, I'm glad you liked that page of mine. Just to be crystal: the blurring I show there isn't of course exactly the sort of thing the eye sees, it's a... dramatization if you will. Nonetheless, I believe it's a useful parallel to how vision works.

hhp

I think he admired its atmosphere and letterwise craft

Maybe those who have toiled carving punches are inclined to have a nostalgic respect for letterforms that retain that metallic sharpness?

"the blurring I show there isn’t of course exactly the sort of thing the eye sees, it’s a... dramatization if you will"
Of course. I'm guessing it would be (near) impossible both technologically and practically to re-create exactly the sort of blurring that's happening in parafoveal vision. What you did is a normal Gaussian blur or something, right?

"a must-have book for the true blue typophile"
There are too many of those. Need to buy a new bookshelf :)

Speaking of reading: any pointers where I could read up on this "talus" measure? That's the first time I've heard the term (and to be honest I didn't get that line of reasoning).

Yes, Gaussian.

Talus is not a common term; but there are no common terms for the empty space above the ascenders and below the descenders, even though it's a very important part of a font. "Internal leading" is clunky and potentially confusing. Sometimes you need a fresh term to make a clean break, and talus (which I didn't make up myself) fits the bill for me. And it doesn't hurt that it sounds classy. :-)

hhp

Hrant, sorry for being so inquisitive (plus derailing my own thread). Just to get this right: Would this "internal leading" equal the "(Typo)LineGap" set in the font metrics? Or where does it come from?

You include some talus in a font to avoid lines touching, especially when you factor in cap accents. And you leave the right amount at top and bottom* simply by ending your ascenders and descenders "early". Of course you plan all your vertical proportions in advance, because you don't want to scale stuff as a result of changing your mind (and deviating from a 1000 unit Em can cause problems, like on older systems).

* I do around 3/4 at top and 1/4 at bottom.

hhp

Ahh, now I get it! Thanks for elaborating.

"Of course you plan all your vertical proportions in advance, because you don’t want to scale stuff as a result of changing your mind"
I sure have a lot to learn.  :->

[End of derailment.]

Getting back to Meridien: the importance of the x-height becomes quite clear
when things are blurred (Patria on the right for comparison). This was a pretty illuminating exercise, if very quick and dirty – and a bit unfair too, as the original sample is scanned (I don't have Meridien).

Looking at this comparison, apart from the short ascenders*, I'd call the designs of Meridien's "e" and "t" somewhat problematic/unclear (and maybe the tittle too?).

* Come to think of it, I should have picked a word with descenders too. Oh well.

—

i personoly think that how it is used is more important that the type use, in readabilaty. i'm not to fond of the indidual shape of the characters, but what i can say is, it is pretty thin and together with the big x-height it is wise not to use too small. (larger sizes no problem)

but it is posible your problem lies with to much characters on a line, maybe to much lines on page, or not enough leading or spacing, but i have to see it one on one to jugde that.

A font's features guide (if not dictate) how it's best used. Meridien is too loose to work well at large sizes (except in all-caps), the x-height is too large to use for long text at typical sizes, and it's a bit light to use for small text. I'm talking about the original digital version here (which was possibly poorly digitized).

hhp

what do you mean by loose? irregular?

The letters too far apart. Depending on the size of a setting a font should be tighter (when set larger) or looser (when set smaller). This is due to how the eye sees things. Scale is mondo relevant. One could even say that at 8 versus 12 point for example we're essentially looking at different fonts.

hhp

Isn't it being loose, though, something that can be fixed/controlled quite easily while typesetting it,* or am I missing something important?

* Which would mean it'd be a lot easier to make it work for display than for text.

Of color, proportions and spacing the last can indeed be tweaked during typesetting, but I do say tweaked because there are "boundary conditions" to worry about, and if you go too far problems will start raining on you. Think about the right side of "r" for instance: the right sidebearing will typically be very close to flush against the glyph, so if you track the font tighter, things -like the left serif of the "v"- will start colliding with the beak; and if you track the font looser you don't actually want the right sidebearing to increase, because the type designer was fighting the whitespace to begin with. One thing that tends to help though is InDesign's optical spacing.

For a properly spaced font, tracking is poison.

hhp

I don’t know the original size of this sample, but it looks like it has been tracked out too much. The word spaces are getting lost, which makes it look like a jumble of letters. Even tracked out, a little more word spacing would add some clarity.

"For a properly spaced font, tracking is poison."
Got your point, Hrant. But you still need to do it sometimes (tracking, that is); say, how about when you need to set a very well spaced font a bit smaller than it was intended to be; or knock it out of a solid color? Would that be Bad, and better to pick a different font instead?

James, the sample I posted is an inch and three quarters wide. Here is a shot of an entire page (no margins, sorry); the whole column is just short of 4 inches wide.
I agree about the tracking; I may be hallucinating, but to me it also seems to vary too much (this being justified text with possibly weird settings). Actually, in some places, the letterspacing looks a bit wonky to me too (like that "llbl" sequence in "Hüllblättern" – the second "l" looks a hair too close to the "b" to me, and that's actually a delimiter between two distinct parts of a compound, so if anything, it should be looser rather than tighter IMHO).

I certainly agree that tracking can be useful. Mostly for certain display settings, and to help out poorly spaced text fonts. The problem is tracking can only hurt a properly spaced text font. To many graphic designers tracking often seems like a quick fix to fill/squeeze a line/paragraph/page into a convenient box, but in fact it comes at a cost.

hhp

"To many graphic designers tracking often seems like a quick fix to fill/squeeze a line/paragraph/page into a convenient box"
Agreed. I was taught this during my first internship, and felt like I held
The Key To The Kingdom of smartly setting type (and saving a couple of lines
in the process) for the following year or so, until I realized it makes things look… bad.

Still: If you say a given typeface is only really optimized for a pretty narrow range of point sizes, what am I (as a typographer, in this instance) supposed to do if I
do need to set it smaller? – Given that I wouldn't just automatically track it out & let it be, but go through it by hand afterwards to make sure no weirdness is happening? Still better to just not do it?

You pay a type designer to modify the font. Which you seem
to have a knack for already :-) so you're ahead of the game.

hhp

:-)

I love the way this thread keeps branching out (though I'm wondering if everybody else hates it).
That said: please don't hesitate to actually say stuff about Meridien, people.
My "Meridien client problem" has apparently been solved, but this is still highly interesting. Like, what about the tracking/letterspacing on that sample?

“For a properly spaced font, tracking is poison.”

nicely put, and i second that, thanks for the 'loose' explanaiton

> (like that “llbl” sequence in “Hüllblättern” – the second “l” looks a hair too close to the “b” to me,

While I also think there might be something a little off about the fitting of Meridien (or perhaps its just this particular setting), I have to say that this discrepancy you point out in "Hüllblättern" is actually something that I suspect you'll find in most fonts (although, perhaps to a lesser degree).

In a typical serif font, the left sidebearing of a b will be less than that of h k l to account for the lack of foot serif -- I would say 80–85% would not be uncommon. In most combinations, this difference is necessary to get good fitting. The only circumstance where you're likely to notice the discrepancy in ascender intervals is in the sequences "llb" or "dlb." I don't know how common "llb" is in German; I would imagine "dlb" doesn't exist in the wild.

Including a "lb" positive kern pair in a font is probably a pretty rare thing.

Just thought I'd point that out.

-- K.

Maybe I should start a separate thread, but reading a book set in Bembo recently I was noticing something that, for whatever reason, I had never considerd, which is the cap “V” followed by a lc “i” and the arm of the “V” crashing into the dot. Have uc/lc ligatures or an optional lc “i” without a dot ever been generated as part of a face? Possibly Eben might know as he is hot on the contextual alternates.

Mea culpa, Kent, thanks for the pointer. Indeed I was only looking at the pattern of the stems. What's funny is, I have actually made the left sidebearing of the "b" tighter than in the "l" in my own font, but apparently forgot about that again. Must go easy on the coffee… :-)

Of course German is a bit evil about possible letter sequences – we can make all sorts of weird ones by way of compounds. It took me a moment to think of one with "dlb", and this may really be reaching out too far, but in Austrian dialect (which would rarely be written though), "Madl" means "girl" (scary image here); so compounds with that and a "b"-Word like "Beine" (legs), or its Austrian equivalent, would be possible.

I would imagine “dlb” doesn’t exist in the wild.

Hah. Alright, thanks for the -dlb- examples.

Back to Meridien: Looking more closely at the full scan you linked to (and spending time I really shouldn't be ;-), I have come to the conclusion that there is something definitely funky going on with the fitting of this setting.

Since the sorts of weirdness I'm seeing are what I would consider amateurish inconsistencies (especially in terms of straights vs. rounds), and since Frutiger and Linotype are not amateurs, I'm going to go out on a limb and suggest that this book was set with Adobe InDesign's Optical metrics applied.

Nina, is this a recent enough publication where this could be a possibility?

Without testing with Meridien itself to replicate the exact same idiosyncrasies and confirm (I don't have Meridien), I can't be absolutely certain, of course. But I have seen Adobe's Optical metrics do this sort of destruction before -- things like pushing out straight-straights and arch-straights, pulling in arch-rounds; it will even exacerbate the "llb" irregularity.

(I've seen Whitman used with Optical metrics, and at first I was suddenly plagued with self-doubt and thought I was going crazy.)

The more I think about this, the more convinced I am about this theory.

Parents: Please, don't let your kids use Optical metrics! Please!

-- Kent.

>suddenly plagued with self-doubt and thought I was going crazy

"You know when you lean back too far in a chair, and it is just about to fall over, but you catch yourself, and in that moment you don't know whether it is going to fall over or not--I feel like that all the time." --Steven Wright

I'm reeling from the sheer number of tangents this thread has taken. It's as if Typophile forgot its ritalin today.

I once tried to set a book with Meridien, seduced by the loveliness of its craft. I worked for days trying to find a rhythm that worked. Now, to a certain degree that's merely betraying my inadequacies, but the finish and affectations of the font (no matter how expertly executed) always seemed to distract. It was standing on its tiptoes instead of sitting solidly on the baseline.

The next tangent I'd like to see this thread take are examples of Meridien set well.

"I’m going to go out on a limb and suggest that this book was set with Adobe InDesign’s Optical metrics applied."

Ha! That might explain a lot. And yes, the book was published in 2008.
I have contact info for the graphic designer who typeset this, but am a bit reluctant to contact him about this and ask directly, as I'm doing the next book in this edition, which could (maybe should) have been his job… no need to make him feel bad about the job he did I guess.

Anybody reading this thread who has a bit too much time, and has Meridien (and InDesign), to do a quick test setting of maybe a couple of lines with metrics vs. optical? I'm intrigued.

"Parents: Please, don’t let your kids use Optical metrics!"
Doesn't that kind of depend on the font used (& how well it was originally spaced), though? (Smelling the next tangent. :-) )

"The next tangent I’d like to see this thread take are examples of Meridien set well."
That would actually be a tangent that's pretty close to the original core of the thread (which I agree is pretty ADD-ish. Much fun).

> .... this book was set with Adobe InDesign’s Optical metrics applied.

Wouldn't that make the whole thing much tighter?* Although I know there are limits to how far InDesign will deviate from the built-in spacing, and Meridien is loose to start with.

* Except if the point size was small. What was it Nina?

> don’t let your kids use Optical metrics!

But Kent, many fonts are so poorly spaced that it can help.
Optical spacing is just another tool that can be used well or poorly.

> I’m reeling from the sheer number of tangents this thread
> has taken. It’s as if Typophile forgot its ritalin today.

Or simply that it has regained its youthful vigor, at least here and now.
Frankly though this degree is divergence is nothing compared to past cases.
Hey, where was that completely bonkers thread - you guys remember?

hhp

Hrant, comparing with the PDF sample, I'd say the point size on this is
just over 10; I'm guessing 10.5.

And: Couldn't optical spacing have been combined with a bit of tracking for
even more wackiness (and to counteract it getting tighter)?

> I’d say the point size on this is just over 10; I’m guessing 10.5.

So not too small.

> Couldn’t optical spacing have been combined with a bit of tracking

Sure, why not! :-/
In fact when I tested out InDesign's optical spacing back when,
I realized that it doesn't take into account the tracking setting,
which can backfire in a big way.

BTW, here it is - the Jackson Pollock of Typophile threads:
http://www.typophile.com/node/32680

hhp

> Doesn’t that kind of depend on the font used (& how well it was originally spaced), though?

> But Kent, many fonts are so poorly spaced that it can help.

I will admit I am not at all a fan of Optical metrics (in Adobe's implementation), except for perhaps a very few, limited circumstances. In my opinion, it should never be used for text settings (certainly not books). Maybe it can improve poor spacing to become mediocre. I've never seen it produce decent spacing overall; and I've never seen it improve spacing to the point where it is acceptable for book text.

But maybe that's because I would never consider trying to use something for the sort of work I do that was poorly spaced to begin with.

> Optical spacing is just another tool that can be used well or poorly.

Mostly poorly. It's a reckless tool. No controls. Which is why I generally advise against it. If you don't have the discernment to evaluate the results, don't use it. (If you do have the discernment, then what are you doing working with a poorly spaced font in the first place.)

Sorry, is my cranky side showing? ;-)

> Wouldn’t that make the whole thing much tighter?

In my explorations, it doesn't necessarily tighten text settings. Display, yes. But not text size. Mostly it just shifts space around. Seems like it puts in as many positive kerns as it does negative ones. (Like straight-straights on the order of 15–35 units).

-- K.

http://www.typophile.com/node/32680
Typophile on acid, man. I'm loving this.
And here we are still talking about type!

> it should never be used for text settings

As you imply yourself, I think you're making a certain assumption:
that people only use fonts appropriate for setting text to set text. :-)

hhp

Kent, I just did a quick test in InDesign (CS) and below about
11 pt the few "mainstream" fonts I tried were all loosened up.

On the other hand I think InDesign must
notice that Meridien is loose to begin with.

hhp