VibeVoice with WAN S2V - trying out 4 independent speakers for cartoon faces

Problems I met; One or two lines picked out a bit. A kind of the previous speaker bleed. Had to generate a few times so that things work. Overall, Sound needed some optimizations in an audio editor to control some volume variations that were somewhat irregular. I used boldness. The lips are not always aligned properly, and for a character it wins and loses lipstick in various clips. The dialogue was just a bit of fun with co-pilot.