[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
smart.fm and fooling text-to-speech
This isn't especially relevant while the automated upload isn't
there, but by using either the Spanish voice or the Portuguese voice
(or maybe even Italian), you can end up with a really good
approximation of Lojban on smart.fm.
It seems you can get "good enough" for anything that doesn't contain
both ' and j or x (because none of the voices for smart.fm have
both).
' is a problem in general: only the German and English voices have
it, AFAICT. English has crazy vowel issues, and German seems to
have problem with trailing vowels (I can't make le'e in german; it
comes out "le'i"). Even so, all the gismu would save a lot of
recording.
http://smart.fm/goals/208205/content -- a couple of examples; you
can see I couldn't really get je'e to work.
You can use the word entry for the munged voice version, and the
transliteration entry for the actual Lojban spelling.
It should be possible to make a simple script that would munge the
word lists appropriately, tell you which langugae to use, and point
out the ones that aren't any good, but there's not a lot of point
without mass upload.
Then we'd only have to fix the icky/impossible ones.
Helper links, should anyone decide to tackle this in earnest:
http://www.lojban.org/tiki/Pronunciation+guide+in+Spanish
http://www.lojban.org/tiki/Pronunciation+guide+in+Portuguese
-Robin
--
They say: "The first AIs will be built by the military as weapons."
And I'm thinking: "Does it even occur to you to try for something
other than the default outcome?" See http://shrunklink.com/cdiz
http://www.digitalkingdom.org/~rlpowell/ *** http://www.lojban.org/