- From: Charles McCathieNevile <charles@sidar.org>
- Date: Thu, 30 Jan 2003 15:57:42 +1100
- To: "Jukka K. Korpela" <jkorpela@cs.tut.fi>
- Cc: w3c-wai-ig@w3.org
The Speech Synthesis Markup Language effectively requires support for IPA: http://www.w3.org/TR/speech-synthesis/#S2.1.5 and if you have a full Unicode system there should be no problem (fonts and tools are now reasonably widespread, although getting everything to work harmoniously sometimes produces a few hiccoughs). --end of real message The problem of diacritics is particularly pronounced in vietnamese where there are many possible combinations of them, and they change the meaning of words very significantly Because consonants are not very well distinguished in spoken vietnamese (the difference between Northern and Southern pronunciation of the same letter is clearer than the difference between different consonants in the same accent. Research at MIT showed that speech recognition of Vietnamese-born Australians speaking english depended on vowel recognition. Personal experience suggests that Vietnamese speech output would also need to get this right in order to make sense. As far as I am aware, the "International" Phonetic Alphabet is not actually sufficient for describing Vietnamese pronunciation. I don't know of a vietnamese speech synthesiser or recogniser (I haven't looked, and none has actually landed on my doorstep), but I would be interested to find out how they do deal with these issues. cheers Chaals On Thursday, Jan 30, 2003, at 11:01 Australia/Melbourne, Jukka K. Korpela wrote: > On Wed, 29 Jan 2003, Nick Kew wrote: > >> Does anyone support the International Phonetic Alphabet on the Web? > > Not very well. IPA can be written using Unicode characters, and most of > the modern browsers support Unicode in some sense. However, the actual > support depends on the character repertoire, and I'm afraid fonts with > full IPA support are very rare. Besides, full utilization of IPA means > that one can construct characters that do not appear as such in Unicode > but are formed by combining a base character and one or more diacritic > marks. This is rather challenging, and although some browsers make some > attempts at it, this works for simple cases only, and not very well > typographically. > -- Charles McCathieNevile charles@sidar.org Fundación SIDAR http://www.sidar.org
Received on Wednesday, 29 January 2003 23:58:04 UTC