Serrano: a custom typeface for Bank of New Zealand

[billtroop]

>Bill, your opinion doesn’t prove anything.

Nick, you're not going to learn how to fit until you give up that system (for I must admit that it is a system, in that it is calculated to do everything wrong rather than right). And you aspire to release a Scotch in competition with Matthew Carter?

[Miss Tiffany]

Thanks for reminding me, Hrant.

[Nick Shinn]

you’re not going to learn how to fit until you give up that system

You have no idea what you're talking about.
I used "that system" once, as an experiment--which was quite successful, actually.

...you aspire to release a Scotch...

I already have.

[paul d hunt]

Paul’s Grandia typeface, which shows such deep understanding of the intangibles of serious reading, but I don’t think the fitting is there yet based on the sample referenced here.

thanks, bill! believe me, i'm keenly aware of the efficiencies of the fitting and struggled with it a long time before i had to give up in order to have something to submit for my practical. i fully intend on revisiting that, but have much yet to learn in the realms of inter-letter spacing. anyway, back to the lovely Seranno... :D

The age of the straight line might be coming back, which is a shame, since that would seem to preclude organicity

i'm sure the organic will be in vogue quite a bit longer...

[hrant]

Nick (and/or Bill), how does this "system" work?

Concerning the "in competition with Matthew Carter" bit, if we're talking about commercial competition* then quite sadly, as Mrs Eaves has shown so well, good spacing doesn't necessarily matter much. It's been the best-selling Emigre font (and that's saying a lot) even though Robin Kinross has equated it with a loose bicycle (while I've equated it with a high attractive women with a severe speech impediment).

* As opposed to critical competition, which is not something Nick (or Matthew, who is however too modest to do so anyway) should present an opinion about, leaving it to others to determine.

Paul, I most certainly hope you're right.

hhp

[Nick Shinn]

...how does this “system” work?

A bit off topic for this thread, but it's explained in the Richler PDF.
http://www.shinntype.com/Richler.pdf

[billtroop]

The theory apparently is that squarish rounds should be fitted as if they were straights, not rounds. (This has got nothing to do with DIN, which in Pool's version is fitted by a master.) The result is that rounds have about twice the sidebearings on either side that they need. (Strangely this did not occur to Zapf whose competently fitted Melior is the ultimate parent of this typeface -- not that Melior should be encouraged to breed.) Another global problem is that the caps are incorrectly fitted in relation to the lowercase, leading to the excessive gaps between uppercase letters followed by lowercase letters. Look at commissioned, where mmi is reasonable but co is too loose, om is looser still, and the snobbish e has erected a hedge to shield its unwelcome neighbours from view. Ca and Ra show the problem with the caps.

[billtroop]

To forestall Hrant, the reason Mrs Eaves succeeds is not because of its bad spacing but in spite of it. The typeface has real charm and the looseness is tonic. Fixing the bad relationships but maintaining the looseness of Mrs Eaves, as more than one expert technician of my acquaintance has been paid well to do, only improves the typeface. One hopes that Licko has learnt something in the meantime, but refitting the commercial font would be impossible because of the necessity to preserve line integrity in older documents. On the other hand, there are people who are simply too invested in their sense of authority to learn.

[William Berkson]

>Ultimately, “fit” is not a matter of right or wrong.

I don't agree. Getting good rhythm and color is important to readability, and so readability is a constraint on text faces that does make for "right" and "wrong". This is probably a range rather than a bright line, though.

[Nick Shinn]

Bill, you are too tied to convention. Overly invested in a sense of authority, one might say.
Innovation is not, per se, incorrect. Of course, the sidebearings of Richler don't look right when one examines them closely and looks for divergence from standard theory and practice. In fact, I had to overcome my prejudices and concentrate on how the face reads in text, to fully commit to the concept.

Ultimately, "fit" is not a matter of right or wrong. Like other aspects of type design, it's more a question of style and tone, and the personal discrimination of the designer. There are different styles of fit, and Richler is wide and open, whether you like it or not.

The face hasn't been used much, due to the exclusive conditions of the licence, but if you'd care to see it in its intended context, you might enjoy this--and learn something:
http://www.amazon.ca/Dispatches-Sporting-Life-Mordecai-Richler/dp/067697...

[hrant]

> The theory apparently is that squarish rounds
> should be fitted as if they were straights

It seems to go even further. For example, the left side of the "b" would be spaced the same as the left of the "h", right? Basically it sounds like it's aiming for uniform spacing of the main verticals, ignoring lesser parts of glyphs.

> Melior is the ultimate parent of this typeface

I'm not seeing that.

> To forestall Hrant, the reason Mrs Eaves succeeds
> is not because of its bad spacing but in spite of it.

Well, that's what I already said.

> Fixing the bad relationships but maintaining the looseness of
> Mrs Eaves, as more than one expert technician of my acquaintance
> has been paid well to do, only improves the typeface.

When it comes to text, I can't agree. There's a relationship between a font's vertical proportions, color and spacing, and Mrs Eaves has to either be a lot tighter (to work at around 14 point) or darker, larger on the body & a bit tighter (to work at smaller sizes). For display work though I can see its airiness being aesthetically useful.

BTW, who are these re-designers? Is it legal?

> refitting the commercial font would be impossible

Not impossible, just annoying to a minority of users who for some rarified reason have to go back and reset a large work with the new version. Plus those people could just have both versions on hand.

BTW, do I actually remember correctly that Emigre eventually -and quietly- released an updated version (I mean after the OT one) with the spacing fixed? (Maybe it was a dream I had...)

> there are people who are simply too invested
> in their sense of authority to learn.

This is the real problem.
Artistes.

--

BTW, to those of you who are weary of the chronic Mrs Eaves bashing and/or fearful of yet another flame war sparked by the bashing, understand that the bashing happens because of Mrs Eaves's combination of commercial success and poor craft. This combination unnerves some of us, because it makes us feel like all our attention to detail goes ignored by most graphic designers. And it does. But we have to remember why we really do apportion this attention.

hhp

[Nick Shinn]

I don’t agree.

But Bill, surely if you asked several different type designers to "fit" the same set of glyphs, they would each do it differently, and each would argue that the discrimination of his/her eye was best?

[hrant]

> concentrate on how the face reads in text

This is pretty much impossible for a designer to do
when it comes to his own work. One needs help.

> “fit” is not a matter of right or wrong.

That's art talk, not design talk.

Richler however does have great style. If/when the
exclusivity runs out you might consider respacing it.

> each would argue that the discrimination of his/her eye was best?

That doesn't mean they're using different philosophies of spacing. Plus some designers -like Paul in this very thread- are modest enough to admit they are not Perfect.

hhp

[Nick Shinn]

Back to Serrano:

Kris, your website won't work for me.
Can you please post images of the Serrano fonts in this thread?

[Nick Shinn]

...how the face reads in text...This is pretty much impossible for a designer to do.

What I mean is that the designer should not apportion metrics primarily by examining large images of glyphs in the metrics window of FontLab, but by reading paragraphs of text printed out at book text size.

That’s art talk, not design talk.

Even though they may adhere to the same philosophy, can't type designers have personal style in implementing it?
Doesn't one type designer "fit" differently than another?
Slimbach and Licko, for instance, are at different ends of the tight/loose spectrum.
But why should there be only one acceptable philosophy of fit?

[gohebrew]

> Getting good rhythm and color is important to readability, and so readability is a constraint on text faces....

I agree with William. Good rhythm is consistantcy in color. There should not be strips of white space between the letters or words due to the shapes of the sides of the letters. This is where good kerning comes into play, as certain letter pairs need to be closer together, whiles others need to be further apart.

Generally speaking, larger type is tighter to a fault, and smaller type is looser. Text faces for children are a little looser even at medium sizes, or the typeface design is very wide, like Schoolbook.

In a well-made OpenType font, kerning becomes obsolete, or is only used as an override, like in InDesign diacritics panel (I assume the non-ME regular version has this pallete).

I wonder if the leading OpenTpe font-makers are adding this extra feature. David Berlow of FontBureau, is this the case with your fonts? Thomas Pinney of Adobe, co-makers of OpenType, is this feature in your fonts? Is Linotype still around?

[dezcom]

Fit should have more to do with the actual glyphs in the individual font than with an external philosophy. It is more about visual physics and balancing forces inherent in the font than a personal philosophy of either aesthetics or readability. The problem is that there is no true measure that fits what the human eye and mind does in the act of reading. There are several possible ways to balance an equation when most of the factors are unknowns. Each designer uses his or her own set of visual measurements and ratios to find an answer. There have been both successes and failures in fitting using any methods available. Perhaps it is not the system one uses but the quality of execution one arrives at?

ChrisL

[Nick Shinn]

Right Chris, that's what I was trying to get at.

[dezcom]

"I wonder if the leading OpenTpe font-makers are adding this extra feature."

What feature are you talking about? You can't mean the glyph palette, it is part of the application software.

ChrisL

[William Berkson]

Israel, Open Type doesn't do away with the benefit of kerning, at least for roman fonts.

>Perhaps it is not the system one uses but the quality of execution one arrives at?

I think the problem is if you substitute a 'system' for the eye, which is I fear what Nick did with this one type face. But I would need to see it in text before judging.

Both Kern Master and iKern are reputed to do a pretty good job. That tells me that there is a lot of constraint on how much variation you can do, and still have it work well for small, extended text. In fact the constraint can be done as an algorithm, to a great extent. That tells me that the eye has pretty strict demands on what is successful in text.

>visual physics and balancing forces inherent in the font

But that is exactly what makes a font readable, Chris. If it were just personal expression, you could say to hell with balance. But you won't end up with a successful text face.

[Nick Shinn]

Sorry, re. "Dispatches", the hardcover is a better option:

http://www.amazon.com/Dispatches-Sporting-Life-Mordecai-Richler/dp/B0001...

[Nick Shinn]

if you substitute a ’system’ for the eye, which is I fear what Nick did with this one type face.

As I said, I had a non-standard "system", more of a general principle really, which I fine-tuned by eye, i.e. by according primacy to reading paragraphs of printed text, not the monitor. But yes, please judge the face in a book, not from a PDF on screen.

[hrant]

> Slimbach and Licko, for instance, are at
> different ends of the tight/loose spectrum.

They're at different ends of other spectra as well.

--

BTW Kris, I was wondering but forgot to ask:
How did you (plural?) choose the name?

hhp

[billtroop]

>There are several possible ways to balance an equation when most of the factors are unknowns. Each designer uses his or her own set of visual measurements and ratios to find an answer. There have been both successes and failures in fitting using any methods available. Perhaps it is not the system one uses but the quality of execution one arrives at?

Yes. There are four basic questions to answer: how do you space H, O, n, o ? Everything follows from that. So: you have four dependent parameters: cap straights, cap rounds, lc straights, lc rounds. (Not that these do not have variations. For example, left b will sometimes be substantially less than left l. And it can happen, not seldom in a 'Modern', that what is right for n is not right for m.)

The problem is that we look at these parameters as independent, when they are all dependent. Richler is a textbook example of this situation: the lc straights are a little loose, but the lc rounds are very, very loose. And the caps don't relate to either.

Nick, if you can't see the disharmony of the Richler spacing, it's time to take a fresh look. Don't you want to be remembered for your successes rather than your failures, interesting and educative as they may be?

Needless to say, spacing is only, ever, optimal at one point size. You hope, in a single master type, to have designed that optimal size such that positive tracking can take care of the smaller sizes and negative tracking can take care of the larger sizes. Or you design other sizes with different spacing. Or you tell your single master customers that they should only use the type at once size.

There is no one right way, but just because that is true, it doesn't make a wrong way a right way. For example, o spacing in relation to n spacing is typically looser in Lino metal (and photo/digital); tighter in Mono metal, and tighter still in ATF metal. (Benton liked tight o's.) Matthew Carter has suggested that one reason for this progression may simply have to do with the comparative fragility of the type in the three systems. However that may be, Sumner Stone for example prefers looser o's in relation to n's than Carter does.

Whatever way you solve the relationships, you must do so plausibly, without obvious rivers of space running through a single word. Like 'commission' in the Richler pdf where it will be obvious even to an un-typographically sophisticated reader that something is amiss. Look at 'personality'. lity is nicely spaced. But everything else is too loose, and the p looks as if it weren't even attached to the word.

I suspect there may be more subtle problems, too. Look at full. ll is very nice, but I would guess that u has the same right sidebearing as l. It needs less, because it is only at x-height. It's a pity, because the typeface could really sing with good fitting!

[hrant]

> ... o spacing in relation to n spacing is ...

From my own experience I have a hunch that this might be due -at least in part- to the relative importance one gives to control-string spacing versus real-language spacing. If you rely too much on control-string spacing, the round-straight sequence (and by extension the overall) becomes too loose when you actually set text. One solution is to set round sidebearings tight, then gently kern the round-rounds positively. This is what I did in Harrier, and people seem to think highly of its spacing.

hhp