There are very few optical fonts, mostly because few designers know or care what they are.

I also think for many type families, once you have three or more weights, you're well equipped to set type across a broad range of sizes. I think this limits designers' need for optical fonts.

Their use was not so obvious to me, growing up with digital type as the norm (Oh, what I've learned!). I agree with James that few designers know what they are. I think I only heard about them 2-3 years ago, and those were the big Adobe families. I'm delighted and surprised by the non-Adobe ones I see from time to time.

I like them a lot too, but I can't imagine how much work they take to create.

You could/should read Tim Ahrens' book on the subject http://markbattypublisher.com/books/typography-monographs-volume-2-size-specific-adjustments-to-type-designs-an-investigation-of-the-principles-guiding-the-design-of-optical-sizes/

Has anybody here read that book?

hhp

http://typophile.com/node/67229

**

But it seems to me that there are comparatively few of these

Your list would be longer if you included the Thin/Extra Light weights of sans and slab faces, which are not intended for text use.

**

I've produced optical variants of the majority of my Serifed type designs, but economically speaking, once the base fonts in a family have been made, I would probably have better invested my time in more popular genres, i.e. Scripts, Sans Serif, and Slab Serif, rather than extending with opticals. It may seem that making display or micro styles isn't much work, merely adapting the base fonts, but it's actually extremely time-consuming to do all the nuancing -- which is very necessary if the fonts are going to be "displayed" at large size with every minute detail exposed for inspection.

Part of the problem is that the different optical sizes of a typeface have to be sold as separate fonts.

One could perhaps use the programmability features of TrueType and other modern digital font standards, normally just used for hinting, to make a typeface wider and thicker as the size is reduced - but then there would be no way to turn that feature off, since word processing programs do not, in general - and by "in general", I mean all the way down to WordPad - allow one to specify an "effective" point size in addition to the point size used for spacing.

Old type catalogues, or data from Monotype on the set widths of different type sizes, allow one to research what was considered a typical increase in width as a typeface was reduced in size, and looking at specimens under suitable magnification would help answer the question of thickness.

> ... the Thin/Extra Light weights of sans and
> slab faces, which are not intended for text use.

But that of course is not enough to make an optical size, not
least because of spacing issues (but many other parameters too).

> Part of the problem is that the different optical sizes
> of a typeface have to be sold as separate fonts.

I propose we turn this into a feature, by charging almost nothing for the "regular" stuff and charging much more for the refined stuff. Segue Nick's "economically speaking": we charge a small fortune for things people who really care need, thus paying for the refined effort.

> One could perhaps use the programmability features of TrueType

This has sort of been done before.
It is/was a great idea - I forget why it didn't work out.

hhp

I also think for many type families, once you have three or more weights, you're well equipped to set type across a broad range of sizes.

Only if you never use serifs. I don’t care how many weights of a serif text face you’ve got, it will still look like shit blown up for decorative pull quotes, and don’t even get me started on people who use text faces for outdoor advertising.

…we charge a small fortune for things people who really care need, thus paying for the refined effort.

Isn’t that how it’s done now? Do the good stuff on commission? ;)

It is/was a great idea - I forget why it didn't work out.

It’s actually a pretty bad idea. Designers who don’t care shouldn’t be encouraged to let the computer do the typesetting for them; designers who do care probably don’t want the computer doing the typesetting for them.

Dead right about proper opticals going way
beyond mere weight variation. Scale does
all kinds of stuff.

Getting commissions is great, but I think
there's room for tiered retail pricing too.

hhp

HTF Didot

Has anybody here read that book?

Yes. The content is great, I'd say the book a must read if you're interested in optical sizes. The prices is steep for POD though, and I have some complaints about the publisher/production/quality that have put a bit of a damper on the book.

Their use seems to me obvious.

There is rarely a particular merit in combining display and text styles from the same family.
In fact, it can be a bit precious.

Misuse of text faces, in display setting, is a different issue.

>There is rarely a particular merit in combining display and text styles from the same family.

That sounds like saying it is only second-rate designers who would do it—which I don't think is right. Optical sizes, such as Adobe's can be used very well together, as intended. Using different families for text and display to give more contrast is also a legitimate way to go. But using optical sizes together can also be beautiful, if the fonts are good, and I don't think is necessarily lazy or second rate. It is just a different look and in the right hands be very good.

Precious? I would call it sober.

(And if it's precious I guess making multiple
optical sizes would itself be that as well...)

hhp

and don’t even get me started on people who use text faces for outdoor advertising.

It has been my experience that a well designed text face works much better in large-scale signage than a well designed display design.

It has been my experience that a well designed text face works much better in large-scale signage than a well designed display design.

In theory I agree, but I’ve seen some really nasty cases of people not following up theory with taste. Text faces with little contrast or gaping ink traps can be really goofy looking when they’re three feet tall, especially at street level.

@Bill: That sounds like saying it is only second-rate designers who would do it...

I said there is rarely merit in combining optical sizes: that observation is based on the general practice of designers, of all rates.

Most of the time designers contrast different typefaces in display and text setting (even when optical variants are available).

The distinction between optical sizes of the same face is usually too refined, i.e. precious, to provide sufficient stylistic contrast, producing too sober an effect for most layouts. A contrast of typefaces is the preferred norm.

It is more likely that an optical display font of a typeface be used with the text version of another face, than keeping it all in the same family. So no, it's not precious to make multiple display opticals, because they may well be combined with text in other faces.

> A contrast of typefaces is the preferred norm.

So is texting while driving.

"Precious" is negative; "sober" is positive.

> they may well be combined with text in other faces.

Which however is not the "preferred norm" there...
People make them to be used together.

hhp

and don’t even get me started on people who use text faces for outdoor advertising.

Kind of a fuss-pot way of looking at it if you ask me... Somethings may be less likely to work than others, but, if in a given situation a text font works on a sign, it works on a sign. The rules are never that rigid and you only limit yourself if you think they are.

If a text font works on a sign, great,
but then it can't really be a text font.

hhp

Not true.

We just have different definitions of "text font".

hhp

Wow.

MR >But it seems to me that there are comparatively few of these [optical sized types]

All of the large typeface families we make with a wider variety of weights, widths, and sometimes styles modified in contrast/x-height for small sizes, serve the needs of composition with optical sizing. Perhaps hiding the fact of massive use of optical sized fonts in publications, is the fact these large families also give optical styling capabilities, e.g. from delicate to emphatic, at a range of sizes. But our users most definitely employ optical-sized font families from agate to meters tall, and millions of others do too!?

HP >But that of course is not enough [a bunch of weights] to make an optical size, not least because of spacing issues

And so the condensed, extra-condensed and condensed-er styles in all weights were born, with the spacing needs of larger sizes in mind!?

Anon >and don’t even get me started on people who use text faces for outdoor advertising.

If the designer of the outdoor advertisement wants the reader to see it from afar, 800 to 1000 meters, e.g., don’t even get me started on people who use display faces for outdoor advertising!?

I don't think there are different definitions of text, only different perceptions of the import of distance on that issue, and on the issue of optical sizing, I'm guessing you all live under rocks somewhere.;)

There are few designers working today that are more accomplished with optical sizes than Robert Slimbach. I would recommend that anyone who's interested in the subject study his work. (Not to suggest most here haven't.) Arno, particularly, is worth checking out because Robert felt it necessary to create two text sizes. Some might consider that overkill, but I would respond that Robert's judgement is very good about these things.

There are very few optical fonts, mostly because few designers know or care what they are.

Maybe that's literally true, but it gives the unfair impression that most professional/practicing designers (the people around here, or who show up at TypeCon and ATypI) don't know or care. I think it's just as likely that people don't have the time or resources to devote to it. Designing a single font is a lot of work. Designing a full family is a lot more. Adding an optical size axis to all that is a pretty daunting task on a practical level (i.e. if one wants to pay the rent and eat).

> And so the condensed, extra-condensed and condensed-er styles in all
> weights were born, with the spacing needs of larger sizes in mind!?

?
What I meant was that weight variation isn't enough. So for example you can't take the Bold of a Regular that was designed for 11 point and use it at 7 (not even if you -mercifully- track it looser).

> I don't think there are different definitions of text

From these years of public discussion I've seen one important dichotomy: people who believe in the difference between immersive and deliberative reading, and those who think it's all the same. For the latter group it's very hard to formulate a cogent definition of "text face".

> Some might consider that overkill

I'd say two is a good minimum!

hhp

There are manifestly several—to say the least—opinions on this issue. I'll say this at the outset: I don't care much for the solution of adopting sans serif with seriffed text (or vice versa), for I think it merely avoids the issue. But provision for sizes ranging from 6 pt to 96 pt (as in HFJ Didot) is too much for me, as I've never set for formats over A4—and I don't like Didot.

The answer that all type designers are too young to have heard of adjustments for scale, because when they were very young all typesetting was done with with scaled types won't wash: one shape fits all sizes didn't work satisfactorily then, and it doesn't work now. Anyway, not many of these youngsters seem prepared to say that they were ignorant of optical sizes.

The answer that different scales for different size ranges are acknowledged to be necessary, but won't sell, or won't sell enough in view of of the work they entail, deserves consideration: as it is, professional type designers have to make a living, and about two hundred glyphs were bad enough, but now customers expect OpenType with not less than twice that number of glyphs, the type designer that makes that number might be less than keen to start work on another few hundred. Let typesetters set Brotschrift!

BTW, the Ahrens book is "Typography Monographs
Vol. 2", which makes me wonder, what was Vol. 1?

hhp

Maybe that's literally true, but it gives the unfair impression that most professional/practicing designers (the people around here, or who show up at TypeCon and ATypI) don't know or care.

I think it’s safe to exclude type specialists from my gross generalizations ;)

@Hrant: People make them [different optical sizes] to be used together.

People make them to work together, if typographers are so inclined to use them.
But typographers don't always oblige.
For instance, in my experience, those who license the Display weight of Scotch Modern, or the Normal, or the Micro, don't necessarily license the other optical sizes, and only a few license the full family, despite the attractive pricing.
Also, there is no guarantee that they won't use the fonts at the "wrong" size!

I for one think optical sizes are potentially a good idea, but should not be used together all the time. Cycles for instance has seven optical sizes. I licenced three of these (11, 5 and 36) so I have one for captions and small text, one for regular text and one for headlines. But I can also use either one of these independently, for their intended purpose.

I think optical sizes are often not used or used incorrectly because designers don't know about them and how to use them. Typefaces are therefore often used in settings that are not fiiting with the character of the font (because of spacing, detailing, x-height etc)

Old URW fonts also have a system of optical sizes, but here the difference lies not in character shapes but in spacing. (The T(ext) version to be used for regular text, D for display, P for Poster etc.)

HP> What I meant was that weight variation isn't enough. So for example you can't take the Bold of a Regular that was designed for 11 point and use it at 7 (not even if you -mercifully- track it looser).

Well, you can use that bold at the small end, but the point of that bold is to use it at the larger end of text sizes when the regular becomes too spindly. That point joins with the point I make above, that optical sizing is everywhere on the large end of the size spectrum. With fewer choices on the small end, (because it's small type), there is less need for so many agate or real text designs. And then! the web crushes the spectrum of possibilities with low resolution joined by dehumanizing rendering to provide pressure for fonts made specifically for small sizes on the web.

> People make them to work together, if typographers are so
> inclined to use them. But typographers don't always oblige.

Yes, but that wasn't my point. You wrote: "it's not precious to make multiple display opticals, because they may well be combined with text in other faces." This implies that making them with the intent of them being used together (which is in fact "the norm") is precious.

hhp

That was in repsonse to your hypothetical backwards causality between use and intent:

... if it's precious [combining opticals in a layout] I guess making multiple
optical sizes would itself be that as well...

To which I could have better said, to avoid the implication you noted:

...similarly, if mixing opticals with other faces isn't precious, then neither is making multiple optical sizes.

It might seem like common sense to infer that type usage dictates the intent of type designers, but I don't believe that to be the case. Obviously, as a type designer, one designs an optical family to work together, it's the norm. And one follows the conventions. How else would one go about it? And that is the default performance quality one builds into an optical family, even though a review of usage might reveal that opticals will mostly be used in combination with other typefaces.

By the way, I haven't done a survey of how opticals are used, so my observations are as anecdotal as anyone else's.

One thing being missed so far is that magazines, newspapers and books have their differences which affect the usage of optical sizes. I pretty sure, for example, that Simbach thinks of many of his faces as first of all book faces. And here what Nick called "sober" about the use of matching display faces might just be "appropriate" to long texts that are made to last.

In fact we need much more sobriety these days.

Also, such opinions could very well be explicitly expressed by a type designer - one doesn't have to stay quiet and hope for the best from users. For example during my TypeCon talk I not only recommended that type designers take the slanted-Roman more seriously, but I also suggested that type users spec them, and even talk to their clients about them.

You can't get upset by mis-use, but that doesn't equate to simply crossing one's fingers.

hhp

This blog has turned up, seemingly in response to the recent discussion:
http://blogs.adobe.com/typblography/2010/09/optical-sizes-on-the-web.htm...

This blog has turned up, seemingly in response to the recent discussion ...

It should not be seen as a response. When I mentioned optical sizes in our Typekit announcement, I realized at that time that it might be worthwhile to come back to the topic with a dedicated post. This thread here turned up coincidentally.

If it isn't obvious, the point of that blog post is to give some general background on optical sizes to those who need it, and to remind users to think about how they use them.

I did say 'seemingly', but your blog was apposite in view of a statement in this forum that young type designers might not know what optical fonts are.

Optical sizes can easily fall victim to economics. With metal punches, you had to make a different set for every size anyway so there was no added cost. Phototype brought the scalable outline in to play. Making optical sizes from that point forward caused a large increase in production cost and foundries weaned there customers away from the optical bosom. There are now decades of type user out there who don't know what the thing is and surly don't want to pay more for it. But not ALL type users--some foundries (with a large enough customer base of savvy type users) still go the extra mile and make them available. The world sells many more Chevys than Porsches but the Porsche buyers are a vocal and demanding bunch. Porsche dealers don't bother to hawk their wares to the WallMart crowd, though.

I do think the majority of young type designers know what optical sizes are. (It's somewhat dependent on one's definition of "type designer.") The problem is that we are now looking at a generation (or generations) of web designers who have been using type in relatively simplistic ways up until now. Now we have web designers who not only have more typefaces to choose from, but we're making optical sizes available to them on top of that. It doesn't hurt to throw in a little education, just to be safe.

>>I do think the majority of young type designers know what optical sizes are.<<

Type designers, Yes, but what of type "users"? Graphic designers, editors, etc., I would bet that many of the younger ones don't know and can hope that design schools will teach them.

Type designers, Yes, but what of type "users"?

Right, type users probably know much less about it. Especially web designers, who, on average (I'm guessing) have less graphic design experience and typographic knowledge than print designers.

I would bet that many of the younger ones don't know and can hope that design schools will teach them.

Teachers don’t always have the time. Even in programs that focus on traditional design fundamentals there isn’t enough time to teach everything, especially now that entry-level designers need to do digital media production work at the level of experienced professionals just to get a job. Finding time to shoehorn advanced typographic concepts like optical sizes into design curriculum isn’t an option for some programs. And for those of us teaching in liberal arts programs at big universities it’s just a lost cause.

Please don't use "apposite".
It looks just like "opposite" but means the opposite. :-)

hhp

>>Teachers don’t always have the time.<<

Just a 10 minute intro and a reference to read would be fine for a 101 intro course.

Just a 10 minute intro and a reference to read would be fine for a 101 intro course.

That’s true of a lot of things. But that doesn’t mean that there’s time to cram them into a course that’s already stretched thin.

Well yeah, opticals is, like, rocket science, and all the brainiacs go into law, medicine and business.
Not.

I don't understand how people can set type properly unless they are aware of optical size.

For example, the compositor should know that type usually benefits from being tracked tighter at larger sizes (unless the font is designed primarily for large size). While you are teaching that—and that, for example, caps usually benefit from being tracked out, you can mention that some font families have increased contrast "display" version that wouldn't work small, but look more elegant in larges sizes. And condensed sans for headlines etc.

It seems to me this kind of stuff is essential to good looking typography.

>I don't understand how people can set type properly unless they are aware of optical size.

They don't. The mass of composers from print, used and knew about optical size. This section of the forum seems deranged, with all due respect, in my opinion. :)

@Dez:
>Optical sizes can easily fall victim to economics.

But they never have, as I explain above and below. At the large end of optical sizing, money for optical sizes made a font boom – the world-wide 100k of display fonts, (free and commercial), that have been created since Fog 1.0. And this says nothing about the text faces funded in the same period!

I think there is simply an error in education, where teaching says “There are 10s of 1000s of fonts out there and no one knows why, now let’s look at these optical sized text thingies.”

>With metal punches, you had to make a different set for every size anyway so there was no added cost.

Well, they didn’t usually make all sizes as you know, they usually had a discrete set of sizes to decide how to make. This added potential cost in the decisions for each size, or range master, i.e. if they didn’t just trace it down to a smaller, or up to a larger size, — the additional skilled choices and testing, were/are the added costs. The drive to do this well and economically has been with us all along, but now it is all sizes that are required.

>Phototype brought the scalable outline into play. Making optical sizes from that point forward caused a large increase in production cost and foundries weaned there customers away from the optical bosom.

Not so fast, everlast... when phototype came in, to replace wood, hot metal headline machines and other large messes, the jobs of setting the variety of optically scaled fonts transitioned to different kinds of composition systems. Larger sizes were hand set phototype, and small and medium set, e.g. with a Linotype VIP, which had different range masters, as did both competing systems, the systems before and after. Optical sizes, and range masters are there all the time, because professional publishers of many kinds demand(ed) it.

>Porsche dealers don't bother to hawk their wares to the Wall-Mart crowd, though.
Have you not been to Webtype.com?

Anyway, the desktop publishing market turned itself from a chaos of optical sizing, to a stratification of optical size use in a few years according to the needs of various user markets. The web... had 16 years of default faces that were really 10-12 pt text designs, and most of the larger optical sizes were produced as graphics with other fonts, because the demand for specifically styled, optical sized fonts, are of huge importance, and pretty obviously worth the additional composition cost and performance penalty all the way up to type composition for web site design.

sorry, for so long a post.

. I don't think there is a lack in typefaces with optical companions.
. Scalable type can be beautiful too, using the same typeface in different sizes on a page can create a desirable spacial depth that would be destroyed with optical sizes.
. There are typefaces that don't require optical companions because they come in a reasonable range of weights and widths. Sometimes just weights. I would find it really weird for example if there was a release of a Caecilia Display or Caecilia Caption, it's such a graciously scalable typeface that it would be almost insulting to modify it for various sizes.
available:

Here's a design problem I faced that to me called for a display size of a text font. The design otherwise did't have a sans, and introducing one would make the design less coherent and change the tone. On the other side, using a second, different serif font looked out of place. So the the best option seemed to be a related optical size of the same font.

So I drew some display italics for my Williams Caslon Text, for titles in my forthcoming book Pirke Avot: Timeless Wisdom for Modern Life.

Here's part of the title page, with the red in "display" style and the rest in "text" style.

.
And here's part of the table of contents, again with the red in display style and the rest in text style.

.
here's a comparison of text and display sizes side by side.

Yep, this is a great example for using optical fonts.
Having a Display typeface in your Caslon Family will give it a technical edge over Adobe Caslon :D