bisse123
02-15-2011, 09:57 AM
Hi. In the coming months I will be creating a project featuring facial animation in UDK. The goal is not to create a complete game, but to create a prototype investigating facial animation (and to a lesser degree body animation) in a NPC interaction scenario.
Currently I'm looking into the different possibilities that UDK offer in this regard and I must say that the documentation and user made tutorials are rather limited at this moment.
On to the questions:
1. FaceFX offers lip synchronization coupled with audio files. Is this the biggest argument for using FaceFX or does the tool offers other advantages as well? I'm comparing this to creating your own bone rig in 3ds max with all the animations, which is then blended together in the "ordinary" animation node editor in UDK.
2. This relates to the above: I plan on using Zbrush to create morph target data, meaning that I use Zbrush to sculpt the various phonemes and expressions.
Is it possible to then export these morph poses to UDK and in UDK create the final animation?
3. Again using Zbrush, I can create "stress-maps" which is normal and occlusion maps rendered when the face are positioned in the most extreme poses. In earlier builds of UDK, the ”TextureSampleParameter2D” has been available but has been removed. To me, this seemed as an obvious component in blending between two texture maps.
How do one blend between the standard maps and the stress maps and time that to the animation?
Currently I'm looking into the different possibilities that UDK offer in this regard and I must say that the documentation and user made tutorials are rather limited at this moment.
On to the questions:
1. FaceFX offers lip synchronization coupled with audio files. Is this the biggest argument for using FaceFX or does the tool offers other advantages as well? I'm comparing this to creating your own bone rig in 3ds max with all the animations, which is then blended together in the "ordinary" animation node editor in UDK.
2. This relates to the above: I plan on using Zbrush to create morph target data, meaning that I use Zbrush to sculpt the various phonemes and expressions.
Is it possible to then export these morph poses to UDK and in UDK create the final animation?
3. Again using Zbrush, I can create "stress-maps" which is normal and occlusion maps rendered when the face are positioned in the most extreme poses. In earlier builds of UDK, the ”TextureSampleParameter2D” has been available but has been removed. To me, this seemed as an obvious component in blending between two texture maps.
How do one blend between the standard maps and the stress maps and time that to the animation?