From Valve Developer Community
Jump to: navigation, search
Faceposer playing back a complex scene with multiple actors.

Faceposer is the SDK tool used to produce choreographed sequences for the Source engine. It creates facial expressions (including lip-syncing) and manages skeletal animations, the position of actors in the world, and any map triggers that need to be fired during the scene.

  • See Choreography creation for a step-by step guide to creating choreography in the Source engine.
  • See Faceposer reference for a quick-reference guide to the different parts of the Faceposer interface.
Note.png注意:In order to open Faceposer in SDK 2013, use the Faceposer.bat file in located in the bin folder. Using the shortcut .exe will give you an error asking for Vconfig!
Note.png注意:To use Faceposer with your own 2013-based mod, set the VPROJECT path to the game directory of your mod.


Blank image.png待完善: Clean up the grammar here, and verify if everything is correct here.
  • Incorrect model image probably made in Photoshop as with HLMV or with incorrect size will make "Load model" menu freeze and white and you will need to stop poser process.
  • In hlfaposer for SDK 2006/2007/Orangebox, the play/stop/pause buttons are missing their icons by default, and display as white boxes instead. To fix this, extract fix to the proper folder, details here.
  • Models may not load at all. To fix this, go into regedit, and delete HKEY_USERS\..\Software\Valve\faceposer. Another solution is to clean the hlfaceposer.rf file in your SDK's bin folder.
  • To make faceposer Phoneme Editor extraction work (if you are not on Win XP) there are plenty things you need to do. Let's work with source 2007 as the most stable revision. For some reasons there are lot's of info how to make it work with Microsoft Speech 5.1, but it's incorrect approach. (it doesn't work properly anyway and it's quality is far from ideal). Go to here. Take lipsinc_data/ from anywhere, even Hl2Beta, copy it to ..\SourceSDK\bin\source2007\bin, copy ims_helper.dll and phonemeextractor_ims.dll from Source SDK Base 2013 Singleplayer\bin to ..\SourceSDK\bin\source2007\bin\, you can edit GameConfig.txt to add your mod here and launch faceposer through steam or just add .bat file with "-game" to your mod (to make it work you will need many things! like to unpack proper shaders from Source SDK Base 2007\vpks (otherwise lips won't move) and other stuff, also don't forget this, write the correct gameinfo.txt for old source engine games. Change Speech API to Lipsinc Speech API in faceposer, it should work now.
  • If all fine in faceposer, but in game lips not moving and you use sounds from other games or old engine versions just don't forget that faceposer formats also can be changed from one engine version to another, go to Phoneme Editor, then press Redo Extraction, if you want to get less size, try use export word data to .txt, then you can import it and then commit extraction to your sound files, same rule spreads on .vcd scene files, just resave them.
Note.png注意:Always make sure that your audio is encoded to 44khz 16 bit mono ADPCM/PCM format (stereo also work though). Otherwise you won't hear anything in game and lips won't move.