I’ve been on a small caps binge and want to make sure I’m using them appropriately and well. What should I know?
Ooh, I have a question on this topic too: Are small caps meant to be used with full-sized caps? I notice that the full size caps often look heavier and does not seem to go all that well with the small caps. Someone also said that when they are not slightly taller than the lowercase characters, they suck. And I tend to agree, although “suck” is a strong word. It’s somewhere on this forum. I just can’t ﬁnd it now.
Small caps are designed heavier than what a scaled-down version of the full caps would be(*) exactly so that the composition has an even color — i.e. the full caps don’t “look heavier” (except to the extent that they naturally should, if it’s the case). Also, about proportions: a sweeping comment like “when they are not slightly taller than the lc characters, they suck” doesn’t make much sense. It can often be the case, but depends of course on the original proportions of UC/lc to begin with. In faces with a large x-height they’ll probably look better if they’re the same height as the lc. (*) This is not the only diﬀerence between a small-cap and a scaled-down full cap.
As I’m starting to understand it… and it would certainly explain a lot… … (if you don’t have small caps for a certain typeface) using the small caps function is akin to using faux bolds and italics? A computer-generated interpretation? d’oh.
> Small caps are designed heavier And wider. I’m not sure why. It could be because: > depends of course on the original proportions Yes, but it also depends on readability/legibility: the primary modern-day use of smallcaps is in fact not with “tallcaps”, it’s as acronyms in running text — as a result strings like “SC” or especially “OS” (since it’s a pseudo-word) cause a handbrake. BTW: “Suck” was me, and I gave the example of the [otherwise superb] Proforma. hhp
> Small caps are designed heavier And wider. I’m not sure why. They look cramped if they’re not, no? Also, as you’d try to open the counters a bit that would mean a wider shape in general. And set more open too. > depends of course on the original proportions Yes, but it also depends on readability/legibility: the primary modern-day use of smallcaps is in fact not with “tallcaps”, it’s as acronyms in running text — as a result strings like “SC” or especially “OS” (since it’s a pseudo-word) cause a handbrake. Good point, but a) is the small diﬀerence in height you propose, especially at small sizes, of any use in this distinction? b) or, better, isn’t the diﬀerence in shape, proportions and spacing in “traditional” small caps more visually meaningful than this small diﬀerence in height? c) doesn’t context take care of most ambiguities anyway? d) is anything lost?
> They look cramped That’s a good point, although I don’t think it’s a huge factor; some lc letters are more cramped than most UC letters. — a) I think larger size is in fact the best way to disambiguate them. And actually, I prefer somewhat large smallcaps. b) The diﬀerence in structure is secondary, especially during immersive reading — in the same way that an upright italic simply doesn’t work. c) Most, yes — all, no. Think Maest. :-) d) Readability. And what’s gained by making the smallcaps x-height, Regularity? A succubus. hhp
a) I think larger size is in fact the best way to disambiguate them. And actually, I prefer somewhat large smallcaps. What I wrote: “In faces with a large x-height they’ll probably look better if they’re the same height as the lc” If you make them even larger, couldn’t you be running the opposite risk: getting too close to the UC? b) The diﬀerence in structure is secondary, especially during immersive reading — in the same way that an upright italic simply doesn’t work. Oh, and +0.2pt in size will work in immersive reading? c) Most, yes — all, no. Think Maest. :-) Not a good example. We’re talking here about letters that are supposed to have 100% decipherability to begin with. ;) d) Readability. And what’s gained by making the smallcaps x-height, Regularity? A succubus. Beter a succubus than an incubus… :/
1) A face that doesn’t have enough room for smallcaps between the x-height and the fullcap-height isn’t a text font, so naturally Regularity becomes/can_become more important — you can freely sacriﬁce readability. 2) There’s virtually zero harm in confusing smallcaps with fullcaps (I mean in their usual usage of acronyms in running text, as opposed to fancy titling), while there is in confusing them with lc. So you could conclude that a font with a huge x-height doesn’t need smallcaps that much. And I personally would go further: a good text face has very short caps (much shorter than current -or past- convention) — it doesn’t need smallcaps, at least not that much — there are [generally] more important things to worry about. Not least because virtually nobody uses smallcaps, even if the font has it right under their noses… > +0.2pt I don’t know how much larger they need to be — it simply has to be a visible diﬀerence, in fact some distance into the parafovea. BTW, it’s interesting that the smaller the point size the larger the absolute diﬀerence necessary. The problem here is that at very smaller point sizes the x-height needs to be larger too — I guess one could conclude that smallcaps have little place in tiny typography. hhp
You wrote 1) A face that doesn’t have enough room for smallcaps between the x-height and the fullcap-height isn’t a text font and a good text face has very short caps (much shorter than current -or past- convention) A bit contradictory, no? How much space is left there then? BTW, your conclusions, either way, seem to be that small-caps aren’t that necessary. I think — and from personal experience: a technical book that I designed with acronyms every other line - that they’re much necessary (not for grotesque faces though ;).
Good catch. I was sloppy. The ﬁrst part assumes conventional proportions — caps slightly smaller than ascenders. Which is why when x-height/cap-height is too large, so is x-height/ascenders (a key factor in readability). In fact my proposal (making caps notably smaller than ascenders) goes hand-in-hand with reducing the necessity of smallcaps. Is that clearer? BTW, I think necessity is simply relative. An ideal face will have smallcaps (even if the x-height is large), but to me it’s much more important to have sized masters, for example — which very few people do. Too many people are making smallcaps (not to mention the dreaded Unicase) instead of spending the time on more important things. And it seems to be a peer-pressure situation. hhp
http://www.typofonderie.com/gazette/ImagesGazette/GazRsc.Gif If this old image can help?! the yellow R is the capital, the small R outline is the small caps version, and the bolder outline is the same R small caps enlarged at the size of R capital. On this example, the lc is missing to achieve a perfect demo. When I design SC, I start generally with a close to perfect balance between lc and cap weight, then, interpolation between the regular and bold original caps, then, redesign of all serifs. I make most of them wider, more open with generally lower contrast between the letter widths. The eyes still the best judge at the end. Note that the OsF ﬁgures match the SC in weight and size.
Nice primer, Rodolfo.