The post
Does anyone have a tip for me on which prompt instructions to use with Venice AI to get the voice-over from an audio reference into my generated video using WAN 3.0? I've noticed it works sometimes and sometimes it doesn't. If I prompt a standard video between two couples in a house or bathroom with the correct instructions, it works. However, if I prompt a cinematic scene in ancient Rome, for example, with a different location or other images, it doesn't work. Has anyone else experienced this, and if so, how can I fix it? Cloning an audio seed file works, but it sounds monotonous and unrealistic. I only want to use WAN; I know it works with other software. Any tips would be appreciated!
View the original thread on Reddit →