Skip to content

Conversation

@kijai
Copy link
Contributor

@kijai kijai commented Jan 8, 2026

Quickly trained Latent2RGB factors, feels good enough considering how small resolution it's for LTXV anyway.

I'm marking this as draft because I'm unsure how to best handle the nested latents when audio is used, but this way it does work including VHS animated previews:

chrome_xWNcOwSK64.mp4

@zwukong
Copy link

zwukong commented Jan 12, 2026

I tried ,very good , a little blurry but can see the processing.

@nistvan86
Copy link

Tried myself aswell, and it works properly, though this resolution obiviously doesn't reveal small changes. Still, it's better than nothing, at least the motion trajectory is visible. And it's quite fast.

@kijai
Copy link
Contributor Author

kijai commented Jan 12, 2026

The resolution can't really be helped, that's just how the LTX latent space is, highly compressed. Unless someone trains a tiny VAE or something.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants