Hmm. Can't think of any with nine or more, at least not off the top of my head.

As for the secondary question: Benton Sans has Book, Light, Extra Light, and Thin, and Interstate has Light, Extra Light, Thin, and Hairline.

Maybe some of the Super Families have what you need?

Nevermind. I misread "more than" :-\

Would H&FJ-style grades count? Chronicle Text, for example, has three weights that each comes in four grades...

So basically, something bigger than Linotype Univers.

Good question.

[edit] Sorry, that only had nine, not more than 9.

@ Phinney: Why?

TheSans has 8 weights plus 8 hairline weights. That’s 16 weights altogether, with 11 of them lighter than Regular (‘Plain’).

Also from Luc(as) de Groot: Taz (10/4).

Flama by Mário Feliciano also scores a 10/4, topped by Jürgen Weltin’s Agilita: 10/5.

TEFF Ruse by Gerrit Noordzij has 11 degrees of contrast.

-Paul van der Laan
www.type-invaders.com

> Would H&FJ-style grades count? Chronicle Text, for example, has three weights that each comes in four grades...

I'm going to assume they treated the grades in Chronicle Text the same as they did in Mercury Text; in which case, the Bold is the same weight for all grades. So the range would count as 9 "weights," not 12.

Not criticizing -- this is standard practice in grading. FB takes the same approach.

Just pointing out, is all.

Interesting! Thanks for pointing that out. I never realised that about grades. Hmm, all the big multi-weight families I can think of seem to have 9 weights, like some magic number...

Oh, I oversaw that Taz has Hairline weights aswell. That sums up to 15/9, as I see no reason why they shouldn’t count as one family.

This relates to a discussion on the W3C "style" list (CSS discussions, including web fonts). I'm watching quite a few things, trying to make sure that specifying fonts in CSS works as well as possible, with an eye to the day when one might be able to use a much wider variety of fonts (web fonts).

Basically CSS only allows a font family to have nine distinct weights (like the original Apple TrueType spec, I think). Folks at the W3C are questioning the need for allowing more than nine weights in a font family.

But Microsoft changed that scale (pre-OpenType, in their TrueType spec, in the OS/2 table) by multiplying it by 100, and the only plausible reason for that would seem to be to allow finer gradations of weight.

I think I'm coming to the realization that this is just another of the many compromises made in CSS font handling. Only a few families would be affected by this one. I suspect it would have to be much more badly broken for the W3C to consider changing it... I'll probably go find some other CSS fonts windmill to tilt at.

Cheers,

T

I am working on a family with 10 weights but it is not ready for release. My reasoning is not so much that you need 10 variations as giving the option to pick which 4 are best suited to particular needs (including reversed type, signage. text, display etc.

ChrisL

http://www.flickr.com/photos/dezcom/3157744858/

Is it possible that the 100-900 system dates back to when MM/GX fonts promised to allow arbitrarily small gradation on the weight axis?

More than 3 weights lighter than regular:

Mitigate
Silicone
Madawaska
Uniwars
Movatif
Tussilago
Carbon
Forgotten Futurist
Pricedown
Ethnocentric
Good Times
Bullpen
Sofachrome
Strenuous
Gyparody
Coolvetica
Pirulen

. . . and a few more.

Just a question ... do you need more than 9 weights on the web? On paper the difference in graduation can be noticed, but on the web, except at the largest sizes, you might not even be able to detect the difference between one weight and another. I can see five weights as being enough, seven or nine being gravy for that media.

As screen resolution improves, you may see more difference, Don.

ChrisL

Depending on how you count, Magma has 4 weights, 8 (adding in the halo versions), 10 (adding the thin weights), or 11 (adding the titling version).

...do you need more than 9 weights on the web?

At any size where the boldest weight vertical stem width is 10 pixels or more, yes.

Hello Tom,

at first I know that "Agilita" from Jürgen Weltin it contains 10 different weights

The Hairline and Ultrathin are not usable in low resolution os print sizes below 20 point. You can see this already at the screen samples if you look on the website.

http://www.linotype.com/en/3188/einleitung.html.

Yours Otmar

> At any size where the boldest weight vertical stem width is 10 pixels or more, yes.

Ah, so we need a weight for each possible pixel width. This could be problematic if we get to the point where a stem might be 24 or 30 pixels wide.

Don

:)

Neue Frutiger has 10 weights:
http://www.neuefrutiger.com/weights.html

Aveny-T has 11 weights.

My new Insouciant has 10 weights with more than a few lighter than regular.

Gingar has a whopping 15 weights.

Hmmm. I still make up some fonts from Adobe multiple masters. That gives me a lot more than 9 weights, kinda, sorta.

Sure miss it, esp. with the expanded character sets available with 16-bit encoding!

We were going to do such great things with MM,
way beyond pedestrian things like weight. Sigh.

hhp

We were going to do such great things with MM, way beyond pedestrian things like weight. Sigh

And what happened, if I may ask?

Some people say Adobe killed MM in some dark deal with MS.
But the reality is probably much more boring than that.

hhp

I heard it was just too clever for graphic designers to understand ;)

Yeah, it took a comp.

Florian, the link needs some http… ;-)

We were going to do such great things with MM,
way beyond pedestrian things like weight.

Nothing’s stopping you from getting right back on the interpolation train. Unless you’re too busy creating manually hinted Armenian fonts or something.

Fonts are for users, and for users it's not the same.
And yes, my Evil Plan takes up way too much of my time.

hhp

Do graphic designers need more than 9?

To be honest, I like to have a lot to choose from, so I can get just the right weight. But it's going to be more than a wee bit frustrating if you can only use so many weight gradations. I might need that forth, fifth, or sixth step from the regular.

Hrant is right. The reality is that MS wasn't interested in supporting MM tech within OpenType, and that Adobe had done a lousy job of evangelizing MM both internally and externally. Adobe was already needing to evangelize OpenType, and avoid making the same mistakes again. It woud have been possible to keep multiple master tech in OpenType, but it would essentially have doubled the evangelism required, and made success that much less sure.

I don't know that it multiple master tech was too complicated for graphic designers to understand, but it was certainly complicated, and it would have required a major effort to educate and evangelize. The relevant people didn't realize that, or chose not to do it. Instead, MM fonts just had their prices reduced in an effort to boost sales. All that happened was that lots of people bought MMs but didn't even know they had them, or what was special about them.

On the plus side, folks at Adobe learned from what happened and worked pretty hard to evangelize OpenType, both internally and externally. Arguably some of the problems solved by OpenType were more important, but I'm pretty sure that the difference in evangelism efforts made a huge difference in the relative success of the technology.

Cheers,

T

Right.
I made some MM fonts in the early 1990s, with Fontographer.
They worked OK in my studio, but not at my service bureau, and we couldn't figure out why.
In those days there was no Typophile Build forum.
MM would have stood a better chance if more foundries had been producing fonts in the format.
That would have helped kickstart the phenomenon.

Right from the beginning there where printing issues associated with MM fonts.

Once the word gets out, it's too hard to turn the herd.

"do you need more than 9 weights on the web? "

Not for text.

For text, I don't think you need more than 9 weights in print, either.

As long as they are the right ones ;-)

We made up a custom version of Minion, by going back to the old Multiple Masters fonts and making up an OpenType instance, using 405 weight, 575 width, and 10-point optical. Heavier, and condensed just enough to recover the cpp setting of the original "regular."

Just got in the first book printed DPT with the type. Man, it is a joy to see Minion with the proper weight again.

The book used Thomas' Hypatia Sans for display (thanks, Thomas!).

I don't know if Amazon's "look inside" can ever accurately show weight relationships, but here is the link:

http://www.amazon.com/Creativity-Its-Discontents-Industries-Intellectual...

It could use a little more leading, and a little shorter measure, but the characters per page requirement didn't allow for that. The type is large enough that even these 65+ year old eyes can read it.

These are my notes for anyone wishing to give something like this a try, things to consider...

405-575-10 settings from Minion Multiple Masters. For characters from newer Minion roman, increase weight by 1, scale horizontal 100.2 vertical 100, raise 1 unit -- (recover from scale only). Not perfect, but will do for non-Latin characters. For superscripts, MM 467 595 6 scale 60H 57.6V. Small caps scaled 102 percent.

Well, that's a fair point. You might not need more than nine weights on a print project, but with subtle variations needed for different projects or different outputs of the same project, that can multiply your needs considerably.

Oh, and thanks for a nice use of Hypatia Sans. :)

IMO it is a matter of range. For a certain use a light might be the proper choice for the lowest weight used (eg the other ones being regular or semibold and their italics — or when more contrast is needed semibold and extrabold etc.). For another purpose the lightest might be a semibold and one would need up to a superbold to get a range, depending on the amount of contrast and the median colour the designer wants.

Nine-weight-families are necessary/needed.

> We were going to do such great things with MM

http://typophile.com/node/88858#comment-490638

> Adobe had done a lousy job of evangelizing MM both internally and externally.

Actually I think Adobe deserves more blame in a more big-picture way: they
made MM too cheap in the unrealistic hope that everybody would pick it up,
because in the US corporate world you either need to make a lot of money
quickly or they'll find somebody else who can promise that, and the actual
product -and its quality- is irrelevant; the stockholders want their yachts,
and they want them NOW. In reality MM could only ever be a niche product,
which means the evangelism should have been done in that light, and most of
all it should have cost a lot more, and expected only modest income. It could
not be a cash cow - it was a thoroughbred. And Adobe cannot be Coalinga.

hhp

Beyond weight, there were some interesting MM concepts of style motility emerging, along design axes such as serifization (Penumbra) and distress (Reliq).

Beaufort originated as an MM concept (with a weight axis)
I can imagine that in a parallel universe where MM endured, I might have produced Sense and Sensibility as such, along the humanist/modernist axis.

Nowadays, variability seems to be conceptualized (by the type designer) and implemented (by the typographer) as discrete alternates grouped in Stylistic Sets, not as infinitely adjustable iterations.