hello, currently i’m testing convai with metahuman using unreal 5.6 and for convai version is 3.6.7, Then I fill in the extraparams with “MinBufferDuration=0.5, AudioLipSyncRatio=0.5, EnableSync=1”
I have an issue because I’m using two languages English and Bahasa. For both languages, the lip sync is not synchronizing properly.
Could someone please help me understand how to fix this?
Hello,
You can try our Unreal Engine Beta plugin, which includes the latest updates and improvements:
https://github.com/Conv-AI/Convai-UnrealEngine-SDK-V4/releases
Please test your MetaHuman setup with this version and let us know if the lip sync behavior improves.
thank you for the reply, so i’m trying the new beta plugin with version 4.0.0 - beta-8. for lipsync behavior improve and better, but it still not fully synchronize with same setting, also it has delay in the beginning before we can ask to NPC. The animation sometimes it doesn’t play but NPC is replying in widget or the animation and the sound NPC stop before finishing the text.
A few quick clarifications:
Lip sync quality is not identical across all languages. It currently works best in English.
The short delay at the beginning before you can talk is the connection/initialization phase.
For the issue where:
the animation sometimes does not play while the NPC still replies in the widget, or
the animation / audio stops before finishing
this can happen if the character is being interrupted, for example by background noise or an open microphone when using Hands Free conversation.
Please double check your microphone state and noise environment. If the issue continues, could you share a short video and your log output so we can investigate more precisely?