RuneXX

/

LTX-2-Workflows

like 64

can set the strength stronger at the node connected to the final image. Think it was set to 0.7 or something by "default" to let the AI have a bit of freedom

Thanks for the reply. Tried set the strength to 1 and insert to different position like middle of the video. Still not match the image I used.

However, workflow using TTP toolset does not have this issue. Also wan2gp does not have this issue.

Try set 0.8 for the first image, and 1.0 for last image as the strength.
Works for me at least

Since not a lot is happening, you could also try reduce the frames to 121, so the model doesnt start to hallucinate new things in lack of anything "to do"
Often LTX ends up more like a slide show though, but with better prompting you can get it to animate and camera slowly move

But i'll try TTP toolset node too, and upload a workflow variant with that node, if it is easier to use or working better

Here is the result with TTP node:

Looks the same to me, but will try look at it, see if something can be improved ;-)

Also tried the workflow from TTP, with same result.

Maybe just try different seeds, and play around with the strength value.
At least my end, no difference

But the workflow was a quick one to experiment, i'll take a look at it, and see if it can be made to work even better ;-)

As per suggestion from Kijai at discussion at KJnodes, I also uploaded a workflow that is not using image guide node, but rather image in-place node. For a more strict first and last frame.
You can try that.

https://huggingface.co/RuneXX/LTX-2-Workflows/blob/main/LTX-2%20-%20First%20Last%20Frame%20v2%20(in-place%20node).json

thanks you !! the in place node works.