[mary-dev] syllabisation

[Ingmar Steiner]

Dear Florent,

it is important to distinguish between the concepts of phonetic 
syllables and "orthographic syllables" that seem to be output by 
LIA_PHON. While it could be argued that the latter are to some degree 
motivated by the former, such insight provides little help in this case.

Unless you can somehow convince your phonemizer to give you the phonetic 
output with syllable boundaries, I expect you will have to write your 
own syllabifier. You could start with a simple rule-based splitter, 
using phoneme sonority and applying the maximum onset principle. That 
should get you past this particular problem, and you could always 
improve the syllabification later.

You might be able to use the Syllabifier in modules.phonemiser for this 
purpose.

Best wishes,

-Ingmar

On 01.08.2011 15:04, fxavier at ircam.fr wrote:
> Dear all,
>
> My ALLOPHONES file are not correct as they take the entire word for
> syllabe.. The output of LIA_phon (the external that phonemises the text)
> have as output
>
> actuellement aktHEl at ma~ [ADV]        ac tuel le ment
>
> But I think there are no ways to split the phonetical transciption into
> transcription according to the syllabes (like it should be in the
> ALLOPHONES xml). Indeed, some of them are 2 phonemes long, 3, it really
> depends on the word.
>
> How does the other language succeed in that? Is there a shortcut I can use
> from this output, beside proting LIA_phon in Java and then modify the
> pogram?
>
> Best regards,
> Florent
> _______________________________________________
> Mary-dev mailing list
> Mary-dev at dfki.de
> http://www.dfki.de/mailman/cgi-bin/listinfo/mary-dev

-- 
Ingmar Steiner
Postdoctoral Researcher

LORIA Speech Group, Nancy, France
National Institute for Research in
Computer Science and Control (INRIA)