FaceRobot for "foreign" languages?

[Hirazi Blue]

This came up at the German Softimage Forum, but IMHO bears repeating here: how would you go about setting up the lip-syncing tools to accept other languages than English and Japanese, for instance, as the case may be, German? You'd have to build some shapes yourself obviously, but is there at least any way to get to the "Fonix VoiceIn" language files that aren't included with Softimage, while Fonix does seem to have them on offer. Sounds like an expensive addon, as Fonix doesn't show any actual pricing on their website, which seldom means "cheap" .
And if that's not possible: any good suggestions on how to convincingly "fake" a "foreign" language in FaceRobot lip-synching would be welcome too...

[3DMastermind]

The beauty of the lipsync system in FR is that you have visual text feedback associated with the visemes.
You manually type in this text file.

The real reason why other languages are not supported may simply be because of the German "text" that is not currently comprehended but both languages most likely use the same IPA (International Phonetic Alphabet).

Although at first, I only thought the text file was simply used as a visual representation in the mixer but it seems it's used along with the audio file in the calculation for the proper viseme.
Therefore, it won't function without the text file.

Basically, if you type in the German speech in English text (yes this may be weird) but it will probably work since the visemes are based on the same IPA that the viseme is derived from.

[Hirazi Blue]

Thanks for your reply...
For practical purposes your method probably will have to do, obviously.
But your IPA remark is only partly correct: AFAIK, both languages use the IPA, which basically "covers" all languages, but individual languages use only a subset of this larger system to represent them. Thus there is room for sounds that are not used in English to show up in German and vice versa... (But if there are, there aren't many, I hasten to add...)
But then again, as this is animation and not linguistics, your method probably will have to do, even if it involves a little more work...
;)