how to replace an 'a' in a font?

liquidmonkey's picture

i'm really liking the ROBOTO font but have never liked the 'a' that many fonts use.
i would like to replace this 'a' with an a that has a single circle and a near verticle line on the right hand side.

1) how can i find fonts with that type of 'a'
2) what program would i need? is there a free version?
3) how can i ensure that this 'new a' will look in-line with the other letters in the ROBOTO font?

OR, maybe there is another way since i'm totally new to messing around with fonts.

thanks for any and all suggestions :)

charles ellertson's picture

Well, for a one-tierd a, you can look at pieces of d and q... A little cutting & tugging...

Karl Stange's picture

In educational publishing (that I am familiar with) that style of lc /a is often referred to as an 'infant' or 'ball and stick' a. It is often used in texts for children as it falls in line with the requirements of educational institutions or bodies, though I am not aware of any specific standards that require it. The best argument I have heard for using it (as a straightforward replacement) is that it looks more familiar to children who have been learning to write, e.g.,

http://www.handwritingforkids.com/handwrite/manuscript/animation/lowerca...

That said, my son has no problem with the more traditional /a (and /g) which is used (with greater frequency) in his favourite story books.

If you want to seek out other examples, searching for infant and ball and stick might help.

Té Rowan's picture

I suppose that Avant Garde and Futura are the most (in)famous fonts with that 'a'. SIL Andika's 'a' has a similar, though not identical, shape; it being closer to handwriting.

As for a font editor on a small-beer-budget; if you are on a Windows computer, you might want to look at Type 3.2 light, a free-beer version of CR8's commercial program.

charles ellertson's picture

Come on guys, most italic fonts use the one-tierd a -- especially the traditional ones.

Nick Shinn's picture

Quite a few typefaces have “infant” /a and /g as a Stylistic Set option.

hrant's picture

Making a mono "a" is one of the simpler things you can do in type design, so it's probably worth doing on your own. The problem is AFAIK free font editing software isn't too reliable.

Are you committed to Roboto?

BTW you should know that a mono "a" isn't good for extended reading.

hhp

Thomas Phinney's picture

"The best argument I have heard for using it (as a straightforward replacement) is that it looks more familiar to children who have been learning to write"

Yes, that's the best argument (such as it is). But as children who have been learning to write generally already know how to read, and have been exposed to both forms of the "a" the argument does not make much sense, except in the limited case that the typeface is intended as a model for writing.

There are plenty of reasons to believe that the two story "a" is usually preferable for legibility, and what little research has been done says that beginning readers prefer it anyway (Sue Walker did a study on the question).

charles ellertson's picture

There are plenty of reasons to believe that the two story "a" is usually preferable for legibility

Has anyone told the Greeks?

hrant's picture

You should see what I tell Americans!
http://typophile.com/files/papazian.pdf (Minus the illustrations though.)

hhp

Bendy's picture

>Has anyone told the Greeks?

I think the alpha is a different story. In Greek there's no danger of confusing it with /d/, /g/ or /q/.

>what little research has been done says that beginning readers prefer it anyway (Sue Walker did a study on the question).

I don't remember Sue Walker's research reaching that conclusion. As I understand, she found that letterspacing and linespacing had a significant impact on children's reading performance, but the single- versus double-storey /a/ and /g/ didn't make a difference. The children understood from an early stage that we use different forms in reading (typefaces) and writing.

Here's a link to Walker's findings.

charles ellertson's picture

In Greek there's no danger of confusing it with /d/, /g/ or /q/.

Nor in Latin based languages.

(BTW, the Greeks still have nu & upsilon and phi & psi to contend with...)

hrant's picture

Nor in Latin based languages.

Balderdash. Immersive reading isn't about taking seconds to consciously decipher individual letters; it's about taking a fraction of a second subconsciously deciphering multiple boumas, more than half of them blurry. This is why for example all-caps is harder to read.

Every writing system has problems (and ways to subtly minimize/avoid them).

hhp

Thomas Phinney's picture

Agreed with Bendy on the alpha.

Bendy, Walker’s research found no difference in reading performance for schoolbook/infant vs traditional forms. It did find a difference in preference, where 10 children preferred the traditional forms, vs 6 preferring schoolbook/infant forms (and 8 having no preference).

The previous linked page just mentions a couple of Walker’s findings; the research is presented in full in: Sue Walker and Linda Reynolds (2002/3) “Serifs, sans serifs and infant characters in children's reading books,” Information Design Journal 11, 2/3, pp. 106-122

Also, it is written up pretty well in her book, which is available as a free PDF here: http://www.kidstype.org/?q=webfm_send/4

Bendy's picture

Thanks for expanding on that, Thomas, I hadn't noted that point.

Match Mitch's picture

The 'a' that looks like a circle and a stick can be found in the standard circle and stick fonts. For instance the ZMethod for Zaner-Bloser worksheets at
http://www.SchoolFonts.com

You can also find it in Normographe at
http://www.Fontmenu.com/site/_Normographe.html

Hope this helps.

Michel Bujardet
Type Designer

ahyangyi's picture

hrant,

I guess it's usually more easily confused with /o/ (and if allowing Turkish, /cı/). Aren't their boumas more similar?

hrant's picture

The way I would put it is that a mono "a" often contributes a good deal of unnecessary ambiguity to boumas it's in largely because it looks close[r] to the "o" (which is notably also a vowel). This sort of logic is also why I like making my "o" slightly "too" wide.

hhp

dberlow's picture

liquidmonkey>1) how can i find fonts with that type of 'a'

a. Search Myfonts.com, e.g. for "geometric sans"
b. type the letter "a" in the sample string
c. scroll until you see as many a's as you need.

liquidmonkey>2) what program would i need? is there a free version?

There are several. Being "new" to this must mean you have plenty of other, simple, great search opportunities.

Michel Boyer's picture

2) what program would i need? is there a free version?

FontForge. It is free. For the license, see http://www.fontforge.org.

3) how can i ensure that this 'new a' will look in-line with the other letters in the ROBOTO font?

Just copy the small letter alpha from Roboto itself and paste it where the small letter a is. Some would argue it is not really a small letter alpha, anyway. You can do that with all the Roboto fonts from thin to black, condensed or not. If you are not too fancy, there is not much else to do (but the aogonek will come out quite wrong).

ahyangyi's picture

For consistency one may desire the single-story /a/ should be without the small curve at the bottom-right corner. However, the curve may actually benefit the readability, because it makes the /a/ looks a lot more different from /d/, /o/ and /q/.

Michel Boyer's picture

The problem is more acute with a fat sans


grabbed from http://www.myfonts.com/fonts/urw/antique-olive/ (There is a single story "a" in Kaiti, a serif, with that little ''thing'' at the right bottom of the glyph: http://www.typophile.com/node/59688#comment-559846).

On the other hand, if an "a" is just a "q" intersected with a "d", then it is almost as easy to make using a fontforge script than with a mere copy paste; you can get your 16 fonts in less than 2 seconds but this time some outlines may come out "crappy" as Hrant would put it; better then to do it manually.

Syndicate content Syndicate content