NEW WORDS BREAKDOWN MODULE FOR ITALIAN LANGUAGE

Discuss Papagayo issues here

Moderators: Víctor Paredes, slowtiger

Post Reply
AloRom
Posts: 1
Joined: Wed May 10, 2006 7:52 pm

NEW WORDS BREAKDOWN MODULE FOR ITALIAN LANGUAGE

Post by AloRom »

Hi to all,

using the "spanish_breakdown.py" module as template, I have created the new "italian_breakdown.py" Python module to breakdown words in phonemes for italian language.
You can download the source here:
http://provosa.tripod.com/DOCS/italian_breakdown.txt

I hope that in the next version of Papagayo this module will be included, so we will have three breakdown languages: english, spanish and italian.

For now you can use italian breakdown by changing the "LipsyncDoc.py" module in this way:
- adding import italian_breakdown
- comment # pronunciation = spanish_breakdown.breakdownSpanishWord(text)
- adding pronunciation = italian_breakdown.breakdownItalianWord(text)
now you call italian breakdown by pressing the Spanish button.


note: I use Linux, and I have tested my module only with this OS.


have fun!

best regards
NJ
Dany
Posts: 23
Joined: Thu Dec 22, 2005 1:52 pm
Location: Switzerland - Suisse

Post by Dany »

Hi,

It looks like your template can not directly be used in Windows because this version of Papagayo don't use .py files but .pyo and they look like binary files.

Have anybody an idea how to use the Italian template with windows ?

Just another question. Is it possible to add other phonems than the papagayo's defaults ? I don't know how you do it in Italian, but I try to use it in French and in German and there are too less phonems, I think.

If someone could help me, it would be nice !

Thanks ! :-)
DarkCryst
Posts: 24
Joined: Mon Jul 24, 2006 9:36 pm

Post by DarkCryst »

py files are python source files. This would needf to be added into the source and then compiled. pyo files are compiled python byte-code (the pyc files are just the byte-code), and so are the result of the final build.

Basically - its an additional code modification that would be incorporated into the source code.

In answer to your second point - the preston-blair phenomes are actually not the native phonemes for the software, it uses the CMU (Carnegie Mellon University) Phoneme set, which has 39 vs PB's 12. The software could potentially support any Phoneme set, it would just require a code translation for CMU to whatever set.

You might find that PB has too few for you, one of the goals of the set is to provide the fewest possible shapes that are needed for lifelike (note - lifelike, not accurate) lip-sync.

I'm working on a code modification that will allow other phoneme sets to be chosen in the software, as well as easier language pack selection, but that is further off than you might want I'd imagine ;)
pepe
Posts: 3
Joined: Sat Feb 28, 2009 10:37 am

Post by pepe »

Hi DarkCryst,

I would like to know how is now your working on a code modification that will allow other phoneme sets to be chosen in the software, as well as easier language pack selection. I'm intersting in implement french in Papagayo...


Thanks a lot !!!

pepe
Post Reply