Quote:
Originally Posted by BetterRed
But what is the proper Spanish voice, if the Spanish text is in dialogue spoken by someone from Brooklyn it would sound mightily strange if they enunciated the words like a Castilian newsreader.
|
All the more reason you get as specific as needed with your markup. Here is the basic list of currently accepted languages:
https://www.iana.org/assignments/lan...ubtag-registry
And here is the rule for subtags:
language-extlang-script-region-variant-extension-privateuse
https://www.w3.org/International/art...-tags/index.en
I suspect you COULD go down the rabbit hole, but I doubt many tools currently support anything beyond the first main language. (Although Search Engines would definitely benefit.)
Side Note: Or take LaTeX English Hyphenation as an example, there are the two major ones (US/UK), but there are also very minor hyphenation variations you can specify: Canadian, Australian, New Zealand. The differences between these variants are
extremely minor (maybe a handful of words), but that doesn't mean the gap won't grow in the future.
Quote:
Originally Posted by BetterRed
I would have course have the name of Cousin Esmeralda's restaurant tagged as French, but I wouldn't want Griselda pronouncing like a Parisian, instead I would expect her to say it as a 1960's Sloane Ranger would have done. Time is another factor to be considered...
|
I mean, we don't even have many Text-to-Speeches that can do Male/Female at the same time, or give you access to 2 variants of English (Male)... and you want to jump right into Brooklyn-Spanish <-> Castilian-Spanish, 1960s Parisians...
On Android they currently do have a few variants of English though (and a few have Male/Female):
- English (Australia)
- English (India)
- English (United Kingdom)
- English (United States)
or Spanish:
- Spanish (Spain)
- Spanish (United States)
Maybe the linguists will get around to creating those algorithms some time in the future, and then our 5 layer deep langs can be spoken properly!