Using OpenType to distinguish tone letters?

Craig Eliason

Anybody know whether it has been ventured to use OpenType features to parse out and then visually distinguish tone letters (as used in Romanization schemes to transcribe non-Latin languages)?

Erin McLaughlin

Hmm... I did a project once that kind of did half of what you're talking about.

An English teacher wanted a typeface that could have phonetic markers appear above/below words. He had been doing it by hand by manipulating worksheets in Photoshop for a long time. (see below)



I came up with a substitution system to do that work for him. For instance, to indicate a short e sound, you'd type "_e" after the character in the word. Like, the word "bet" would be typed "be_et" and the _e would be replaced with a phonetic marker glyph with a negative sidebearing, that would appear above the e.

The client wanted short vowels to be red, long vowels to be blue. I didn't know anything about colorfonts 3 years ago, so I taught him how to do this by using "find/replace" substitutions that apply a character style in InDesign. He could search for instances of the underscore-based code and change the color of those glyphs.



It was especially difficult because I needed to draw horizontal lines of varying lengths to span different combinations of letters that made a particular sound, and also different heights to accommodate capital letters.



This project was crazy! But so satisfying Also, I must note that this is based off of Source Sans, the Open Source typeface by Paul Hunt. Props to him!

So... I know this isn't completely what you mentioned, but I thought I'd share since I've never shown this anywhere

Kent Lew

Very cool, Erin. BTW, one step further: the search-and-replace routine could likely be built into a GREP style so that the color styling happened automatically, rather than manually after the fact.

Dave Crossland

Erin, was this a paid gig?

Erin McLaughlin

Craig - was that actually sort of what you meant? Like, if there was a dictionary built-in that would substitute IPA symbols (or others) if it recognized a combination/word?
Kent - That is a very good idea! I should have followed up with them and done that!! No time like the present  Thanks!
Dave - Yes! So... as you may be hinting at, a fine example of how people can make money off of OpenSource development.  

Craig Eliason

Erin - Almost what I'm talking about. But where you have presentation of tones triggered by a special series of keystrokes, I'm talking about coding that would recognize the tone letters in an existing Romanized text and sub in custom presentations of them. The underlying text would remain intact.

Kent Lew

Craig — What kind of tone letters are you thinking about exactly? Are you talking about those encoded in U+0363–036F and U+1DD4–1DE6?

You want a contextual routine that will somehow interpret and translate a text that doesn’t have the encoded codepoints and provide a visual substitution without converting the underlying text encoding? Doesn’t really seem like the role of a font.

Can you provide a before-and-after example of what you envision happening?

John Hudson

Craig, it might help to have an illustration of what you have in mind. Are you thinking about things like pinyin tone numbering and zhuang tone letters? Are you imagining visually distinguishing the tone indicators stylistically, or replacing them with an alternative tone indicator, e.g. combining tone marks on the syllable vowel? The latter would be very complicated, because you'd need to identify the vowel in the preceding syllable, but not impossible.