when use audio input ，it won't do the correct lip sync
when use audio input ，it won't do the correct lip sync, it seems the LTXV normalizing sampler ignore the audio mask
I haven't tested audio input. I'm not sure it is the normalizing sampler though, I seem to recall someone else thinking the same but then still struggling with other samplers. I'm not sure if anyone figured it out yet. Odd, for sure. That is one of the reasons why I renamed the workflow to "DoAlmostEverything", heh.
If anyone finds this out, please share!
Hey! I got custom audio working with your workflow @Phr00t . Just delete the LTXV Empty Latent Audio Node and add the nodes from pic below.
Connect the LATENT to the audio_latent and edit the audio_normalization_factors.
The audio output from Load Audio goes directly to the Video Combine audio input.
I hope I could help :)
Thanks，it works！the “audio normalization factors ” is the key！