Hi guys, I tried to solve the issue I had via ICE, but unfortunately it didn't worked as I wanted, so I'm asking for a custom solution done with scripting.
The idea is to extract phoneme values from FaceRobot ( only English phonemes ) using some SDK commands currently available. Specifically the data that I need can be found using the VBscript "GetPhonemeBlendFromSpeechClip", which output phoneme weight from current frame. What I need: 1. Recognize the actual speech clip lenght, or gather data range from timeline lenght. 2. Get phoneme weights for each frame ( script GetPhonemeBlendFromSpeechClip )...if phoneme is not called in that frame default values should be 0. 3. Create a set of 38 nulls/bones, named accordingly to the phoneme name. 4. Drive the X position of each null/bone from 0 to 1 by the values ( per frame ) which you get from the script "GetPhonemeBlendFromSpeechClip"; if the value of the phoneme is greater than 1 max value should be 1. Idea is to be able to drive blend shapes in Maya/Max by exporting the nulls/bones values based on the X axis, which then will be driven via SDKs ( in Maya ) to the Blend Shapes automatically, hence lipsync from FaceRobot to Maya in a couple of steps. I would like to know if someone is available for this ( I suppose ) relatively quick job and how much it'll cost. Cheers guys Nicolas