How to make small caps in FontLab?

Rene Verkaart's picture

What would you say is the best way to make small caps in FontLab? I'm busy with my Insider and Nordic Narrow and I wish to make small caps for them.
Nordic Narrow will be available soon on Fountain. After the first wave of fonts I want to expand the family with small caps and perhaps alternate versions, swash or something like that.

Would you scale the UC's and them make them heavier in weight? And if yes, doesn't it them become out of proportion in relation to the lc's?

Does anyone have a practical tip?

Thanx,
®

dezcom's picture

Leslie Cabarga's book "Learning FontLab Fast" http://www.logofontandlettering.com/ has a section on this. I found it to be a starting point and ended up doing a great deal of visual correction. For my font, I reduced the caps to 96% to get the stroke weight but this will vary with your font. There is no simple quick way to do it right. The stroke to counter ratio is different so it must be optically corrected. Even the slant angles on M and W need work. As is usual for me in all parts of type design, it is all in the eyes.

ChrisL

Nick Shinn's picture

Rene, if you want to advertise your product line on Typophile, buy a banner ad.

raph's picture

For what it's worth, I took a pretty close look at the way Benton did the small caps on ATF Bodoni. For the most part, the small caps are automatically derived from the regular caps, with changes consistent with the Benton optical scaling technology. They used anamorphic scaling to widen the letters slightly with respect to the overall scaling, and used stroke offset to add more weight. Perhaps most importantly, in traditional work anyway, small caps are much more loosely set than either standard lowercase or caps.

Even so, a few of the glyphs appear to have been completely redrawn (as you will see from studying this 200 dpi scan of the Bodoni Book 18. Most obviously, the "W" is simplified a bit, so that it is almost halfway between the lowercase 'w' (no middle serif) and the overlapping VV design of the uppercase. I don't remember exactly which glyphs were tweaked, but the serifs on the "H" look a little longer than on the caps, and I think a couple of letters had their widths adjusted.

Best of luck. Small caps are important for typographic sophistication. Oh, and a query like this doesn't come near pinging my annoyance meter; in general, I'd say spamming involved repetitiveness, and posting without additional content. Congrats on getting distribution!

Nick Shinn's picture

>annoyance

I was refering to the flashing icon.

As for the question: sans is easier, because you don't have to manually extend the serifs.
So really, just scale down to slightly talle rthan x-height, expand the weight a bit to match overall lower case weight, and perhaps widen a few characters.

Expanding weight mechanically will mean that your SC have less contrast than your caps, but how this plays out is, as Chris says, up to your eyes to determine.

How does one assess small caps? -- be sure to try them out in a variety of settings: Caps with small caps title, as an all SC subhead centred above a U&lc text block, as an insert into a LC text block, as title abbreviations in a bus. card, as a side-heading, following a drop cap, etc.

Toby's picture

I´d scale the cap down slightly above x-height, and then use the interpolation tool to get the desired width of character and stems..but that´s just me

Rene Verkaart's picture

Thanx for the tips.
I'll try doing it like that. I was just looking for a good mechanical way to do this. I will have to review the complete small caps font as is, when it's ready. I'm sure some glyphs need to be reconstructed.
Most important now is that I've got a lead on how to start making them.

Nick. Does my icon bother you? It is meant to catch the eye, and judging by your reaction it does. It's not meant to be like a banner, but why not showing a bit more than just one picture? I mean, Typophile offers the option. Why ignore this?
Does it bother anyone else? Then I'll change it.

Regards,
®ené

Also tired of Helvetica? Get an Insider with more personality

*** www.characters.nl { Typography to express yourself } ***

Rene Verkaart's picture

Other then that I think it's good to present yourself on this forum. I think the complete picture is important, not just the font you're currently working on.
Do you know how many 'outsiders', scouts and otherwhise type-interested people are scanning this forum?
I got some direct requests from people if they could use my fonts, over this forum. And I'm talking about some good contacts.

I say; take the chance to get your product out there, but don't spam on Typophile.

Regards,
®

PS: I haven't seen another animated gif on Typophile yet, so it's a good USP. ;-)

Also tired of Helvetica? Get an Insider with more personality

*** www.characters.nl { Typography to express yourself } ***

oldnick's picture

Rene, I vote with the other Nick on your icon: it's distracting when you're trying to read anything in its vicinity.

andreas's picture

Rene,

if you like an animated icon you should made the animation very soft. oldnick is right, its very hard to read since your eye catcher is hammering my eyes.

Some examples from other users. I like Kyles eyes.

http://typophile.com/user/103 Stehpen Coles
http://typophile.com/user/7230 Fredrik Andersson
http://typophile.com/user/604 Kyle Johnston

--astype.de--

Nick Shinn's picture

Rene, sure it works as publicity, but Typophile would be very tacky if everybody was hustling their stuff the way you do, so please don't abuse everyone else's restraint.

hrant's picture

Nick, still feeling lonely I guess?

Rene, the amount of self-promotion you do here is totally within bounds as far as I'm concerned. Some people only post to sell [themselves]. I'm the type of person who hates advertising, but you're all right in my book.

Functionally however your icon's movement is too distracting; all you need is to get attention once, so just limit the number of frames, and don't loop.

hhp

dan_reynolds's picture

Rene, I have never had a problem with your icon. As has been said above, if you are trying to attract attention for your foundry/site, there are better ways to do it (a more soothing icon, maybe pulsating?).

But I don't think that you are under any obligation to change it.
__
www.typeoff.de

oldnick's picture

Dan,

I think that most folks are offering constructive criticism, and not making demands. As of this moment, four out of seven participants in this forum are bothered by the constant blinking, not the content (or the intent, for that matter).

hrant's picture

More Like: Four out of 1000+ are bothered [enough to complain].

hhp

oldnick's picture

30 million Frenchman can't be wrong.

hrant's picture

I saw a great bumper sticker once:
"50,000 wolves can't be wrong. Eat moose."

hhp

oldnick's picture

I saw a great bumper sticker once:

My favorite is "My OTHER car is a piece of junk, too."

Rene Verkaart's picture

Got the message. I can imagine it's distracting. I'll try to find another solution in the next days.

Hrant, I also think it's within the boundries, but if it bothers other people... It's a community and we should all get along as good as possible. And if I can do my part to change my icon - which is nog such a big deal to me - I'll do it.
I don't feel offended by the remarks. I see the challenge in finding a more subtile solution.

Regards,
®ené

Also tired of Helvetica? Get an Insider with more personality

*** www.characters.nl { Typography to express yourself } ***

John Hudson's picture

This is the approach I use to make smallcaps in FontLab.

1. Scale uniformly to the desired height and put into the background mask layer. These will be the guide not only for height, but also for internal space. So, for example, the distance between the uprights of the scaled H will be the distance, edge to edge, of the uprights in the smallcap. This will produce smallcaps that are nicely proportioned relative to each other (presuming that your caps are nicely proportioned).

2. Paste in another copy of the cap outlines, and scale these uniformly for weight. I usually match the vertical stem weight of my smallcaps to my lowercase, even though the smallcaps will be taller.

3. Use the FontLab 'Interpolate' function to reduce the height and width of the weight-adjusted caps to match the height and internal space of the mask. This should be done in two passes: one for the y-direction and one for the x-direction. Be careful to maintain weights.

4. Test and fiddle.

John Hudson's picture

When Typophile finally allows us to post images again, I'll try to remember to illustrate these notes.

dezcom's picture

"When Typophile finally allows us to post images again"

Do any of the moderators know when this feature will be available again?

ChrisL

Rene Verkaart's picture

> When Typophile finally allows us to post images again

I can hardly wait for it, because point 3 is difficult to me. Could you elaborate a bit more on that, John?

Thanks,
®

Also tired of Helvetica? Get an Insider with more personality

*** www.characters.nl { Typography to express yourself } ***

.'s picture

While there are some shortcuts to get you into the neighbourhood, making small caps is primarily Mr Hudson's #4: Test and fiddle. (If there was an equation for making small caps, all fonts would have them, necessary or not.) Enjoy the testing and fiddling!

paul d hunt's picture

Not the same as inline images, but you can post an attachment to the Small Caps-HowTo article that I started over in the wiki. All I know about images is that getting this functionality back is on the top of the list of features to reimplement, so hopefully soon? i'm glad we have search back, woo hoo!

paul d hunt's picture

err... make that SmallCaps How-To. i wish i could edit my own mistakes!

eomine's picture

Hrant wrote:
> More Like: Four out of 1000+ are bothered [enough to complain].

Come on, this is not fair. Most of these 1000-people currently "online" have never posted one single comment to the forums. Nothing has bothered them enough (yet) to make them post something.

The animated icon does bother me, as does the three-line signature (repetition: see Raph's post).

hrant's picture

Hence my "enough to complain".

hhp

John Hudson's picture

Paul, I've prep'd some illustrations to post to the wiki and started to edit the entry, but I hit submit before I was ready, and now I'm getting an 'Access Denied' error when I try to go back to it.

John Hudson's picture

More Like: Four out of 1000+ are bothered [enough to complain].

I'm mad as hell, and I'm not going to take it any more!

:)

paul d hunt's picture

john, for some reason, case matters with wiki entries whenever i get "access denied," i try it in all lower case. at any rate, try this link: http://typophile.com/wiki/smallcaps%20how-to

Rene Verkaart's picture

He, thanx for the Wiki.

®

Also tired of Helvetica? Get an Insider with more personality

*** www.characters.nl { Typography to express yourself } ***

William Berkson's picture

John, thanks for taking the time to do your 'how to'. This and others you have done are really a valuable contribution to Typophile. And thanks to you Paul for setting up the 'How to' wiki. Hopefully, over time this will get into it some of the great 'how to' past posts also--as well as letting John in to post his graphics!

dezcom's picture

"at any rate, try this link: http://typophile.com/wiki/smallcaps%20how-to"
Paul,
The "%20" in the URL indicates a space has been yped there. If you replace "%20" with a / (slash), the link works.

ChrisL

John Hudson's picture

Okay, I have uploaded my illustrations to the wiki. This link works for me: http://typophile.com/wiki/SmallCaps%20How-To

But you can also navigate to the entry from here:
http://typophile.com/wiki/How-To

dezcom's picture

John,
I get "Access denied" from your first link.

ChrisL

John Hudson's picture

So do I today, but it worked yesterday. Well, it does say *beta* :(

dezcom's picture

John,
Are you using a Mac?
I am just trying to figure out why it is so intermittent

ChrisL

TBiddy's picture

Very nice instruction John. This is going to come in handy soon for me as well. BTW, you ever think of doing the same thing for an "italics" tip section?

John Hudson's picture

Are you using a Mac?

No, Windows.

BTW, you ever think of doing the same thing for an “italics” tip section?

You have to make italics from scratch. There really is no other way to do it well, and certainly no shortcut from roman to italic. These are two completely different styles of lettering: it would be like trying to make a nastaliq Arabic type from a naskh type, or an ashkanazic Hebrew type from a sephardic.

hrant's picture

I agree that making italics is a totally different animal than making smallcaps, but John's stance assumes the -mainstream- belief that italics must be qualitatively more cursive than the roman. At the very least, this depends on the design at hand; but I would go further and state that a proper italics is generally quite close to a roman (except -primarily- in slant). The relationship between the roman and italics in Patria is I think much healthier than what the mainstream stance would produce; both the roman and italic were quasi-mechanically derived from a slanted-roman master (Harrier).

So basically: if one decides to make the roman and italics "genetically" close, there are indeed methods, good and bad, just like in the making of smallcaps. It's not always "from scratch". But a separate thread, please! Wait, didn't Terry already start one like that? Yeah, the "true italics" thing. Where did that end up going?

hhp

dezcom's picture

"You have to make italics from scratch"

I wish I had the benefit of John's experience. I found that out the hard way and I totally agree. I don't think it is a much an issue of more calligraphic or cursive but what is clearly different enough to be differentiated easily by the reader from roman. After all, italics function within other text is to make itself separate out enough to clue the reader yet not blast out by being so different as to be jarring.
As for construction--I find all of the techniques for quicker transformation from slanted roman to italic I've seen to be of very limited value (read-not a time-saver). Maybe other people have found otherwise but I can just draw it quicker straight away. I just use the roman in the mask layer as a guide for general stroke and counter weight. The curves just don't translate for me in a way that makes it worthwhile.

ChrisL

hrant's picture

> what is clearly different enough to be differentiated
> easily by the reader from roman.

And slant does that fine. Sometimes with a little help from other things, like lighter color, narrower width, and yes, cursiveness too. But structural differences (not the same as the issues of "finish" you mention) are not required (or even good all of the time).

> italics function within other text is to make itself separate
> out enough to clue the reader yet not blast out by being so
> different as to be jarring.

I agree, but "jarring" doesn't have to be an interruption in reading. It can also be a "skew" in the voice of the text - which is more likely to happen when the structures are very different. See the relevant part of my Daidala interview* for some elaboration on italics.

* http://www.daidala.com/25apr2004.html

Chris, tell me what's wrong with Patria's roman-italic relationship. Please.

--

BTW, didn't somebody (David Berlow?) recently explain how to tweak the curves in mechanically slanted forms to correct for the stroke distribution warping?

hhp

dezcom's picture

"Chris, tell me what’s wrong with Patria’s roman-italic relationship. Please"

I never said that there was a problem with your Patria italics. I am just saying that there is more than one way to skin a cat (is that expression too american?). . . that there is more than one way to accomplish creating a good italic. It can be done as you say (which works with your face) or in other ways that a designer deems fit for their purpose. This includes chirographic treatment, shift in weight axis, whatever works. I hate to slam the door on innovation by someone and would prefer to let type designers find their own way. I am not looking for THE answer. I am only looking for the answer that works for me in a given situation and the answer can change with the situation. I am also allowing for anyone else to find their own answers in their own way. I would like to learn from their experience too.

ChrisL

William Berkson's picture

The thread about making italics is this one:

http://typophile.com/node/13743

David Berlow & Mark Simonson's stuff should be 'wiki'ed. Mark used Briem's method.

John, do you think that the need to start from stratch even applies to italic Caps?

John Hudson's picture

do you think that the need to start from stratch even applies to italic Caps?

The italic caps that can be easily made via slanting roman caps, e.g. those without bowls, can also be easily made in a number of other ways, e.g. from primitives. I think the relatively straight forward relationship between italic caps and roman caps -- compared to the range of relationships possible between the respective lowercases -- can be misleading for designers. Yes, there are methods to quickly arrive at italic caps from roman caps, and there are tricks for adjusting the weight on the bowls. But speaking as someone who has made use of such methods and tricks numerous times, I'm suspicious of them. I think they lead to a kind of dull repetetiveness that tends to make italic caps the least interesting part of any type family. Recently, I've been designing a display face that has a slight slant and some cursive features, and because there is no original roman to start from I was forced to think about the caps in terms of the design as a whole, and not as something derivative. If one has the luxury of time, I think it is worthwhile to approach italic caps in a similar fashion, independent of the roman.

William Berkson's picture

>an also be easily made in a number of other ways, e.g. from primitives

Thanks for the suggestion. I am working on this now, so I'll probably do a mix of approaches and see what comes out.

The simplest ones, which I started by slanting and narrowing the roman caps, I had to extensively redraw, but at least slanting was a start, and I got some consistency. The italic caps should have a fairly close visual relation with the roman caps, I would think--moreso than the lower case where it is a matter of getting two quite different styles to look good together, as well as go with the italic caps. Now that I think of it, we are really trying to harmonize four different alphabets with the roman and companion italic u & lc. Fun, but quite a challenge.

hrant's picture

Ironically, I personally think italic caps are conventionally
too close to the roman! :-/ As John, I've realized the need to
rethink them, and have arrived at what you see in Patria.

hhp

John Hudson's picture

The italic caps should have a fairly close visual relation with the roman caps...

I think that is one legitimate approach, but I think it has become too much the default treatment. The reason for this, of course, is that when the first italic types were cut they had roman capitals, and the development of italic capitals happened a bit later and the forms were typically sloped versions of the roman caps. That delay had a lasting impact, I think, because although many chancery scribes would also use roman capitals with their cursive hands, there are examples of more cursive capitals that predate the slanted capitals in italic types. But these examples had no influence on the latter because by the time people came to make slanted capitals in type they seem to have referenced their own roman types rather than chirographic models.

I think one way to approach italic capital design is to consider that what really distinguishes italic lowercase from roman is speed.

Roman lowercase letters represent the (re)formalisation of the miniscule: the introduction into the lowercase alphabet of the formal qualities of the classical capitals, as rediscovered in the renaissance. The irony, of course, is that the miniscules had developed, via the uncial, away from that formalism, over hundreds of years, only to meet up with it again right at the time when type met up with the renaissance in Italy. The various local descendents of post-carolingian miniscule letters became more upright, stiffer, and slower to write.

When we apply the term 'cursive' to italic letters, we should reckon what this means in terms of speed. There is a tendency -- perhaps because all type goes at the same speed -- to consider the slant of italics and the 'cursive form' of the letters as somehow two separate components of what makes an italic. But the slant and the form of the chancery writing hand on which italic type is based are both the result of the cursivity, or speed, of the writing. The characteristics of fast writing are compression, slant, lack of reverse strokes (entry and exit strokes only, no double serifs), rapid transition from stem to bowl (deep cuts, again with little reversal)... in other words, all the hallmarks of traditional italic lowercase letters.

When I'm designing italic capitals, I sometimes ask myself 'How can I add some speed to these?', even if I'm basically following the usual relationship of italic to roman caps.

dezcom's picture

"When I’m designing italic capitals, I sometimes ask myself ‘How can I add some speed to these?’"
Very good concept to think about. I never looked at it that way. Thanks John!

ChrisL

paul d hunt's picture

John, when you say to use the FontLab ‘Interpolate’ function, what function is that? I'm assuming this is one of the Multiple Master tools and not the interpolate nodes function, correct? Can you explain this a bit more?

Syndicate content Syndicate content