Indesign ALL CAPS button: what OT-Features does it apply? What is "desired behavior"?

Sebastian Nagel's picture

Dear readers,

i'm wondering how the "ALL CAPS" formatting in Indesign (not the "Change Case > Uppercase" function) should behave.

What i can see from testing in CS5, it applies the CASE feature (with positional changes) if found in the font, and it exchanges lowercase to upper case letters in a hard-coded way (independent from the font). Is this all it can do, or is there possibility for more if done in a smart way?

In specifics, i'm wondering if a font with default oldstyle figures should/could change them to lining figures when the formatting is applied, and if it's possible to let this happen with otf logic.
To compare: for the small caps button, the SMCP feature is applied, and if the otf font is set to exchange the figures in this feature along with the letters, they are changed too. So is it possible to do something similar with the All Caps button, or is this just and only old-school hard coded indesign?

thanks for your ideas

charles ellertson's picture

It depends on what's in the case feature. I agree with you, figures should be "lining." When going into a font (with permission, of course) I usually add this to the case feature.

This also means adding a bit of spacing to the numbers in the cpsp feature, but for me, not as much as I add to the letters.

Since the InDesign "all caps" routine calls two OT features, one really needs to do work in both places.

Other things to consider are the parens, brackets, quote marks, etc. Dowding had a list o candidates in Finer points. I also set the Spanish inverted question mark & exclam to work with caps in the font, then make up a lowercase version to use when they don't start a sentence. This too can be automated.

I think the reason I wind up doing a bunch of this work is it only makes sense for a compositor, a user of type, not the designer/publisher. There are too many different workflows possible, too many text editors, layout programs, and varying style requirements. So yes, it can all be done, but I'm not really sure the font publisher should.

Sebastian Nagel's picture

Thanks for your reply.

At the moment, i have some pos commands in the CASE feature, that shift spanish punctuation, mathematical operators, brackets etc. up so they line up with the upper case letters. they are no special versions of the glyphs, but the normal ones repositioned upwards accordingly.

Back in the 200x, i read somewhere (prove me wrong and i'm happy) that one can't combine pos and sub commands in the same feature, so this would mean if i want to switch the figures too with the CASE feature, i'd have to make duplicates and manually reposition the glyphs like {}() ... and then substitute them along with the figures?

If you say "all caps routine calls two OT features", what would be the second feature besides CASE?

Thanks a lot

Nick Shinn's picture

For marks, I make alternate glyphs for the case feature.
As I design lining figures to work in a mixed case setting, they are generally too much like lower case glyphs to work in an all-cap setting, so I do a separate set of “.case” figure glyphs.
Sometimes it’s necessary to do two complete sets, one proportional and one tabular, but often I try and make do with the default set as tabular, and only /0, /1, and /7 proportional and kerned.

Sebastian Nagel's picture

:) as i have 10 figure sets already, and the mixed case lining figures work with the capitals, i am reluctant to add another set for case – but the general idea is a good one for some cases.

so i'll have to see what to do ... maybe create duplicates of the puctuation and brackets so i can use SUB in the CASE feature.

charles ellertson's picture

The reason to add glyphs & use SUB for parens, brackets, etc. is so they can have different kern values with the caps when their position is different.

Nick Shinn's picture

Is raising marks like hyphen and colon a desired behavior?
One is so used to seeing them at the normal distance above the baseline that centering them vertically on the cap height (for all-cap settings) doesn’t look right.

charles ellertson's picture

I sort of agree with Nick on this one. Personally, I find parentheses and the apostrophe (& therefore double quotes), a different matter. It does depend somewhat on how the font is drawn -- another case of "if it looks right, it is right."

You also run into editors who have requirements about style. As their money is green, we usually accommodate them, however bizarre their taste...

All part of why I think this is probably a composition matter, not a type design matter.

John Hudson's picture

Raising the colon? Who does that?

With regard to raising the hyphen or dashes, I think it depends on the context of the all-caps setting. If, for whatever reason, one has all-caps within a body of text, then raising the hyphen/dashes seems unnecessary and may look weird. But in titling or other display situations where the all-caps stand by themselves, raising the hyphen/dash seems to me not only desirable but pretty much essential, since it will otherwise look too low in most types. I generally raise it to align with the crossbar of the H.

Jens Kutilek's picture

Sebastian, the other feature that is activated by the Caps button in InDesign is Capital Spacing (cpsp).

Sebastian Nagel's picture

:) i was raising the colon, saw that it looked wrong, and moved it back.

Thank you – so is my thought right that i could use CPSP for the repositioning (like additional/different spacing of capitals and vertical shift of brackets), and use CASE for substitutions (like exchanging figures)?

charles ellertson's picture

Mr Nagel -- and others:

I think you have the wrong notion about what OT features accomplish. What I see here is the notion that type will look as you, the type designer, want it to look.

Well, the glyphs, yes, sort of. Nothing else. What features do is to save handwork for the person who actually uses type. We say "thank you" for everything that can be automated through features, right up to the point where you add a routine that does something we really don't want.

At that point, we have to stop using the feature at all, meaning all those "thank yous" disappear. Instead, we curse you for making handwork out of what use to be automatic.

Nick Shinn's picture

Is that the royal we, Charles? :-)
Your taste in what is appropriate in a feature is not shared by all typographers.
Necessarily so, given the diversity of jobs that require typography.
The trick for foundries is to match feature implementation to the typeface and the kind of work it will most likely be used for.

charles ellertson's picture

No Nick, that's the commoner -- the tradesman who actually does the work.

My point was to put something in a feature that a large enough segment finds wrong negates all the good in the feature.

If there's any royal ego going on, it would seem to be the type designers, who what, want their work hung on the museum wall?

Thomas Phinney's picture

FYI, Adobe routinely adjusts the height of dashes with the 'case' feature.

I am definitely in favor of these adjustments.

The colon is a tricky case, mostly because of the semicolon. I think the latter really seems like it has to stay on the baseline. I might instead make the top dot of the colon and semicolon higher in the 'case' variants.


Sebastian Nagel's picture

charles_e, i'm with you – as a type designer i can't control everything what i would do as type user (what I am to a large extend).

What i can't deliver as a type designer today is a set of lead with no further settings – there is always a "default" behavior – my spacing, my kerning, my decision for default figures, etc. also, the software we are using, has default behaviours, for example, Indesign by default exchanges "abc" to "ABC" when i hit the all caps button.

What i was initally asking was if i technically could, and from usability view should exchange the old style figures if i activate all caps in indesign.
The type user wants capital setting, but indesign only delivers this for letters in a hard-coded way (which is good for some fonts) – except i tell indesign that my font is capable of something more, that is likely more desirable from the users point of view. (or do you prefer to manually set the figures to lining right *after* you have already hit the all caps button?)
Its the same with capital spacing ... i have one default behaviour for capitals – to work with minuscles well. If the user hits the all caps setting, what he wanted is to have a line of capital letters and he'll get that the one or the other way. As a type designer i can give him an arbitrary one, resulting from the spacing with minuskels, or a better one with adjusted spacing. He can still decide to override this manually, but what i try to do is make it – by probability – a little bit easier, doing less overrides.

charles ellertson's picture

I think your goals when wearing the type designer hat are about right. Don't want to quibble over specifics.

Indesign by default exchanges "abc" to "ABC" when i hit the all caps button.

BTW, have you tested this? Set a,b,c, use the InDesign All Caps feature to get A,B,C, written off an .idml file, and taken a look inside that idml with a text editor?

From how you described things, it sounds like you've done this -- I'm just lazy & just checking :-)

gargoyle's picture

What i was initally asking was if i technically could, and from usability view should exchange the old style figures if i activate all caps in indesign.

Technically, it's possible to add whatever rules you want; functionally, changing the figures from oldstyle to lining is one of the recommended effects of the case feature (though, of the few fonts that I tested, only Garamond Premiere Pro seemed to support it).

Re: case changing— InDesign's All Caps function doesn't change the case of the underlying text (in IDML, it simply adds the parameter Capitalization="AllCaps" to the enclosing element). I think Sebastian's point was that the effective change in case is handled by the application, rather than the OpenType features that also get triggered by activating All Caps.

hrant's picture

Well, the glyphs, yes, sort of. Nothing else.

Of course this makes no sense. A font is not merely a collection of letterforms.

Raising the dash(es) for all-caps is frankly a no-brainer (and something that used to be done even back in the metal days, by flipping the sort). And the colon should not be out of contention.

I can even see shifting things based on the language being set.

Simplicity? Strictly for simpletons.


Nick Shinn's picture

… of the few fonts that I tested, only Garamond Premiere Pro seemed to support it …

Then you haven’t tested any of mine :-)

Nick Shinn's picture

All the time, one sees all cap settings with oldstyle figures.
If it were a pointedly quaint oldschool vibe, referencing when almost all fonts with oldstyle figures didn’t have the option of lining figures for all cap settings, that would be understandable. But I doubt that’s the ever case.
I suspect the designers just choose a typestyle and go with the default figures and aren't even aware of the issue.
Here is something I received via email today, from a type company that should know better:

What makes this an exceptional faux pas is the fact that the figure styles in the headline and display are bass-ackwards.

Thomas Phinney's picture

Ouch! Of course, that was in an email newsletter, where we have to rely on system fonts still (unless one goes to all images, which is silly). I think we got bit by the default figures in Georgia there. Unfortunately, email does not reliably support real web fonts yet. If you look at the accompanying web article ( it has no such issues. At the moment we are using lining figures throughout, but we may yet go to OSF in body text.

Nick Shinn's picture

Sorry for picking on that email from Extensis, Thomas, I just happened to get it the same day as this thread. No font distributor is immune to slip-ups, not to mention foundries.

However, the issue here does not have a technical solution.

In the first place, it’s careless copywriting to begin a sentence with a number (unless spelled out)—it creates too many typographic problems. It could have been written “THE YEAR IN REVIEW”, after all, what year other than 2012 is likely to have been the topic?

Secondly, if the font chosen has OSF as default—even if it were a “real” web font with a supported “case” (All Caps) feature—there is nothing other than typographer savvy to guarantee lining, cap-height figures. This is because (a) all-cap text may be typed with the shift key rather than the caps lock key, and (b) the “all caps” feature might only have been applied to the letters, not the numbers—they not being caps!

Thirdly, many fonts with lining figures as default have only three-quarter height and weight, and are not a true match for capitals.

Of course, there is always this fail-safe option:

Perhaps the only technical salve to this issue would be a prompt in the layout application that called attention to mismatched heights in text with adjacent all-cap/figure characters.

Typographers/Compositors—can’t do without ’em, eh Charles?!

charles ellertson's picture

Typographers/Compositors—can’t do without ’em, eh Charles?!

Well, even though Nick's tongue is likely a bit bloody and sore -- especially if he had a laughing fit -- that's my position. It should probably be noted there are a few people like Kent Lew who are type designers, book designers, and *real* compositors, all rolled into one person. I only design and set books, I have absolutely no interest designing type. Now fixing type's another matter. Fixing type's kind of like Sudoku. Nothing is as challenging as laying out a book though, with compromises on every page.

* * *

Nick brings up a good point. Full-height lining figures should be available. More than one use/need for them.

Sometimes, as with Matthew Carter's Miller, they already exist, but for some reason were not included in the basic font. (Same is true for os figs in Miller, too.) Well, that's probably because Carter & Cone only sold Type 1 fonts, but you could get them as a supplemental font(s) for $50 or so if you knew to ask. Don't know if FB has the same policy, esp. as they've made up OT versions of Miller...

For those designing fonts for multiple uses, make 'em available. Charge more if you want. I'd gladly pay, it takes me over a hundred dollars in my time to make up both sets of os figs, especially the 2, 3, and 7. More if you really do justice to the tabular os figs.

Finally with respect to figures, I'm coming around to the point of view that small cap figures should be available too, if you include small caps in a font. Keep running into acronyms with numbers in them. Nothing else looks right in that case. I've yet to find a need for tabular small cap figs.

hrant's picture

Nothing is as challenging as laying out a book though

We each have our affinities, and what one person finds easy another might find effectively impossible. However if you were ever to make a font, especially a text font, especially one that other people would pay good money for, you'd probably change your mind. I think making a tool as opposed to a finished product (especially a static one like a book) is on a higher level of difficulty, at least for most people.

small cap figures should be available too, if you include small caps in a font.



Nick Shinn's picture

I've yet to find a need for tabular small cap figs.

This is the thread that convinced me of their necessity:

charles ellertson's picture

However if you were ever to make a font, especially a text font, especially one that other people would pay good money for, you'd probably change your mind.

Mirror thinking, hhp. I have fixed fonts that people offer to pay good money for, but that's not permitted, usually. I have given the fixes to the designers -- some of the names might surprise you -- and seen them incorporated in new releases.

As for making a tool, the biggest challenge in type design today -- my point of view -- would be making up a font family that isn't a rip-off of exiting fonts. Just look at the comments on these forums.

Most of your tools haven't been tested before being rushed into the marketplace. The practice goes all the way back to early photocomp & continues today in the digital era. How can anyone design a (purported) text font without seeing how it prints on several kinds of papers? Yet that's the exception, rather then the norm. More testing seems to be done with web fonts, which seems passing strange, since most web systems don't yet support typography.

We have more than enough "raw material" tools. Just very few finished ones that actually work well.

Finally, I look at money as a battery or a capacitor. It's a way of storing energy. If you have enough, getting more "money" becomes a religious matter, and that's not my religion.

John Hudson's picture

Hrant, one of the things that makes laying out a book challenging is that one is generally forced to work with tools that one did not make oneself, and which may be limited in various regards. A complex text is indeed a challenging thing to work with, without the relative freedom of creating one's own typeface to fit. I spent much of the past four years working on the Brill fonts, but I never forgot that my challenges derived from the challenges that Brill's typesetters had struggled with for 325 years of scholarly publishing.

Whether this means that 'nothing is as challenging as laying out a book', I'm not so sure. My wife is a firefighter.

hrant's picture

I have "fixed" Garamond Premier, for good money, but I don't pretend I could have made it from scratch (although I can make certain things that Slimbach would probably have trouble making). And I believe that project required more experience and expertise than trying to fix something (not everything) in a given book design.

I think much more things can go wrong when you tweak a tool than when you tweak a product.

the biggest challenge in type design today -- my point of view -- would be making up a font family that isn't a rip-off of exiting fonts.

I might agree, but it depends on your definition of "rip-off". Deriving inspiration, even from "living" fonts I find unavoidable, and desirable. Carter and Lew certainly do that. And sometimes we do see high invention, like with Legato and Fenland.

Most of your tools haven't been tested before being rushed into the marketplace.

If that "you" is plural, I agree (especially when it comes to free and/or libre fonts). But I myself do enough testing that my clients (which are mostly custom, not retail) end up long-term happy. And when it comes to retail, "rush" is not part of my vocabulary; Nour&Patria was started in 2000, and I'm finally ready to publish it in 2013 (although partly because I only found the right partner this year).

How can anyone design a (purported) text font without seeing how it prints on several kinds of papers?

Experience helps. And -as in any field- realizing the impossibility (which I equate to undesirability) of Perfection certainly helps too.

As for the religion of money, it's not for me either; if it were I would not be in type design. But it does help make my family happier, and that makes me happy.


charles ellertson's picture

John, laying out a book is inherently a compromise. Spreads should align. Widows and orphans should be banished. So do you run a page short? Even with running feet? Do you make/save a line & affect the word spacing? How does that affect the next spread?


Editors add their own limits to available compromises, usually with the mistaken notion it is "better" typography. Don't hyphenate the last word on a page, or the last word in a paragraph. Not hard, but where do they think the space comes from to achieve that? From the word spacing, of course.

Or the designer specifies "all chapters start verso" with the understanding by the editor there will be no blank rectos...

And editors have this silly notion that the comp shouldn't rewrite the text...

Then there are editors who let an author call 3-4-5 tables or figures right before a new subhead. Basic thinking is tables & figures should be placed after their callout (or on the same spread at least), but before a new section (subhead).

All this before the (interior) designer mucks things up -- like you said -- by choosing a typeface inappropriate to the work, (& just here, I'll reserve "inappropriate" for the character compliment of the font itself). Last week, we had a book where there were a lot of italic superscripted numbers. That was the author's requirement. The font chosen by the designer? Quadraat... Good luck finding Quadraat serif italic lining figures for those superiors...

But yes, firefighting's probably harder. Certainly more dangerous.

R.'s picture

I've yet to find a need for tabular small cap figs.

This is the thread that convinced me of their necessity:

Great to hear that! I am happy about each and every type designer who includes (proportional and tabular) small-cap figures in his or her fonts. Looking forward to a new year full of small-cap figures—happy 2013 to all of you!

Syndicate content Syndicate content