Expanded character sets on font sales.

blank's picture

Do expanded character sets in retail fonts—anything beyond Latin-1—pay for the time it takes to develop them?

As a newbie font designer I keep throttling my ambitions so that I can actually finish an original project. One limit I am considering imposing on myself is not developing character sets beyond Latin-1. I enjoy working with diacritical marks and learning about how they’re used in other languages but I’m not sure that anybody outside of the Latin-1 world will buy a font to use the expanded character sets anyway. I know that this has a lot to do with the high piracy rates in “developing nations”, and given the state of the economy, I need to worry more about bringing in some money.

Stephen Coles's picture

The answer to this question depends a lot on the kind of fonts you're making. Is it a display typeface? Or a text family that is appropriate for rebranding a multinational corporation or publication?

blank's picture

Anything I’ll be doing is display type. I’m not even planning to work with lowercase letters for most of 2009.

Si_Daniels's picture

Also depends if you're planning to sell these yourself or work with a distributor - who may have character set requirements. If you are selling these yourself you might want to provide smaller char-sets and do extensions on request.

Stephen Coles's picture

A foundry may have requirements but also offer assistance.

Thomas Phinney's picture

From a strictly financial POV, adding non-Latin (say, Greek and/or Cyrillic) is almost certainly not worthwhile. Extended Latin is less certain, but probably doing much beyond, say, Adobe CE, is not going to pay off so well. Heck, even Adobe CE may not be worthwhile financially, but since it's a pretty small amount of additional work, it's not much risk, either.

That being said, doing more language support may be fun, and is certainly a service to the global community. Then again, in terms of "service" one might simply seek out specific *really* under-served languages and try to support them....

BTW, one would think that having much more language support would make one's typeface more likely to be selected for international branding purposes and the like. Although this is probably *somewhat* true, it seems to be much less true than it ought to be. In my experience working with major corporate customers at Adobe, it seemed like 80-90% of the time the typeface was selected first, and then whether it supported all the languages they wanted that was an afterthought, unrelated to the selection process.



Stephen Coles's picture

On the other hand, my experience with FontFonts is that language support is often a major influence in the buying decision. FontFont's European customer base (in contrast to Adobe's) is certainly a factor.

paragraph's picture

In my experience, it does not make enough difference to sales, especially in display typefaces, to justify the extra effort required.

Stephen Coles's picture

I would suspect Paragraph's clients are Euro-centric as well.

acnapyx's picture

Generally speaking, taking a little additional effort always pays up - but not always from a financial point of view. Frankly said, even in Basic Latin there are some glyphs used in quite rare languages like Icelandic and Faroese (eth and thorn come to mind). How many customers from these countries from these countries you expect to have? And still, most typefaces contain these glyphs. Small effort, but valued by the people who license your fonts.

So my 2¢ are: best to stick with Basic Latin, and if you have the time and will, optionally you can expand later the character set with Latin Extended-A and charge extra for these versions. Cyrillics is worthwhile, but requires considerable effort (and is country-specific!), and the market for it is somehow thin. Still some foundries (Storm, ShinnType and CastleType, for example) provide well-designed cyrillics, so this may rise the sales of your fonts. Of course, I'm speaking as a person, who uses Cyrillics regularly, so I'm biased; when deciding which typeface to use for a project, for me the availability of Cyrillics is a major consideration.

Sye's picture

I know as a customer I am often impressed by larger character sets, but since nearly all of my work only needs basic latin it's not really a big buying decision. Don't FontShop have 2 prices, Standard and Pro, with Pro offering the expanded character sets?

Maybe a model like that would work for you?

Dan B.'s picture

I'm Romanian, and I won't buy a font that does not offer language support for Romanian*, simply because I cannot use it without the diacritics. I admire Jos Buivenga (another European) for doing the extra work even in his free fonts! That said, I don't think the market for typefaces is large in my country.


*In fact, most typefaces don't support Romanian entirely, as they use a cedilla for ş (surprisingly not for ţ). I really wish people would start getting that right.

Jos Buivenga's picture

How many hours does it take to extend your font(s) with a CE set? You can generate most characters automatically and you only have to set up your kerning classes once, so I agree with Thomas on this (... it’s a pretty small amount of additional work, it’s not much risk, either ...). Scrolling through my sales reports I do see quite a lot of CE customers, so for me it's lucrative.


Thanks, Dan!

piccic's picture

I had similar doubts, and tried to have a better intelligence of the issue with a different starting question, here.

Difference between character set and codepage
African language support

They helped me a lot, and I think you should just decide which language/cultures you wish to support with your display alphabets…

Nick Shinn's picture

How many hours does it take to extend your font(s) with a CE set?

It can take a lot of time to get right.
For instance, "comma below" is a nasty little thing to squeeeze in if you don't have much descender room and the weight is bold. It has to be designed: does it match the comma or the acute, or is it just a vertical dash? Decisions, decisions. Dan, please start a thread on this, showing samples!
The slovakian d- and l- caron characters require extra kerning. Does one make the character widths normal and add kerning where there is overlap, or wider, and remove it where there are gaps? Should this character be the traditoinal comma shape, or the newer "thin acute"? Decisions, decisions.
The ogonek characters are nicer if a proper composite glyph is made, rather than just superimposing the accent.
Is it really possible to fit an acute (over l) or circumflex (over h) above an ascender? Perhaps if one cheats the height of the ascender down a bit, flatten the angle of the accent a tad...
Will this breve look too much like a caron in Czech?
Will the tilde above "i" collide with adjacent ascenders: to fix by making it narrower, or adding kerning?
And so on, into the small caps.

Dan B.'s picture

@Nick Shinn,
As suggested, I started this thread.

dezcom's picture

I find the kerning time required for the i diacritics and caron-to-right glyphs is great but after doing it several times and making every possible permutation of mistakes in the process, I am finally getting the hang of it. The bolder weights are the toughest. I must admit that I get a kick out of seeing my fonts set in Finnish, Polish, and Icelandic just for the fun of it.


K.'s picture

Goran Soderstrom's picture

I think you definately should do it, it's not that much extra work to do a couple of extra accents and a few more composite glyphs if you compare to the total time of developing the full typeface. There are very good resources on how everything should look for us who are not used seeing these accents :) And it gets easier for every new typeface you work on, and eventually you'll get a nice workflow and can make a lot of your developing time more effective.

eliason's picture

@K. But if you're doing a script typeface, you're already a glutton for punishment by definition! ;-)

piccic's picture

K., are you the author of Kafka?
It's great, I'd love to know more of your rationale behind the design decisions of the OpenType features…

Syndicate content Syndicate content