FaceFX Support

FaceFX Documentation and support

Languages

FaceFX supports American English, UK English, French, German, Italian, Spanish, Korean, Japanese, Mandarin Chinese, Czech, Russian, Polish, Brazilian Portuguese, and European Portuguese.  Using the FaceFx Studio python interface, new languages can be supported.

Getting results without Text

If text is not being used, you can get good results analyzing any language using the default US English selection. A slight improvement in quality can be achieved using the closest “approximate” language in the no-text case. For optimal results however, text must be used and the appropriate language must be selected when analyzing the audio.

Language Notes

• Japanese - Katakana and Hiragana must be used. Kanji are not supported.
• Mandarin Chinese - Simplified and traditional characters are both supported.  Implemented using the additional language framework with only a dictionary.
• Czech, Polish- Implemented completely in python with the additional language framework.  Good examples for how to implement other languages.
• Russian - Open the Russian_Beta.py file for additional information on improving the Russian dictionary by downloading a more extensive GPL’d Russian dictionary.

Choosing a language

To analyze an audio file with a particular language, simply select the language from the Language listbox in the Create Animation Wizard. If you are using the analyze command, you can specify the language with the -language flag.  Languages implemented with the additional language framework must be specified using the exact filename of the .py or .dict file in the “Analysis Languages” folder.