Do you have a workflow you use to add audio to Wan 2.2 videos?

@Zogratis seconded, do you have a workflow for this?

@d0000d Very simple to achieve: Take an I2V workflow (say sulphur's example). Add a load video as images node (I prefer vhs), set the number of frames to the video's length (or set the load cap to however many frames you've set as the output), and pass the loaded images to ltxvimgtovideoinplace node instead of the still wherever it goes (and dim calculations, preprocess etc). Done.

The weird thing about ltx is, strength doesn't seem to matter normally when you pass a whole video - it uses exactly the original frames. Meanwhile, audio seems to be perfectly mutable (could be because I2V passes an empty latent, but I don't quite remember the details except that it's a impossible to do V2V without a special wf / lora like the edit one; I think that I tried silencing parts of the audio or concatenating latents and that did gen fresh audio, but never video).

@firemanbrakeneck thanks for replying

my hard drive disagree with your statement, it think you're a newbie. I tried telling it to be nice to humans, he threatened to delete all my awesome NFSW LTX outputs

Yes, anyone who says LTX 2.3 sux for nsfw just hasn't spent the time to figure it out. Generating a 30-second video with 5 scene transitions and no face distortion, with audio. In less time than an 8 sec Wan. Yeah, who wants that?... (10Eros_10SNodes_LikenessGuideHelper)