The grind continues. One day, we will be able to make accurate text consistently, but that is not today.😵

vidu-q2 for character refenrence? really nice video

yessir. Q2 is a godsend

How about using Nano banana (on Adobe) on still frames as a temporary measure?

If it’s a still frame I could edit itin capcut too, but the fact it’s not still frames and not in separated layers makes it difficult…

aiden can you tell me what you know about achieving setting consistency when generating at different angles or when characters leave & return to settings in subsequent scenes (possibly in different wardrobe)?

Vidu is really a great tool, tbh 🫡🫡🫡

You need to try it. I think it is an underground, yet insane model.

i've been cooking with it! when i finish my first project with it i'd love it if you check it out 🙏🏻🪽

Would love to!

For this kind of thing we are still using camera tracking and post, but someday it will improve for sure

this looks FANTASTICCCC omg!

Thank u so much!!!

I would greatly prefer if Clip or photoshop just auto completed frames on sketch,that would make the work of animating easier without reducing my ability to fix the artwork

Oh... Just two legs left in the empty corner of the last frame.

you're right. Ti's why none of these shots are final, and can be changed for better fit. For now though, you will probably always find imperfections here and there. Perhaps wait another year.

Have you tried Google flow. Is it any good

This is really impressive, how long did this take?

Mostly just today, but storyboards took a little bit to get to where i wanted it to be

who are you and why are you only appearing in my feed now. this is some great looking anime.

老哥，国内平台怎么搜不到你的账号

还没有好好搞。想创建。 bilibili上叫sishenfan1

Looks promising! How you going to do voices?

Tbh, I am not sure yet. I know what they’re going to say, but how to get good quality delivery is something I’m thinking about how to di

@gantetzu Eleven Labs, i could show you

@gantetzu Yeah, that would be cool!!

You can make empty billboards and put the text with After Effect.

You are right, but currently I am not paying for Adobe, so I do not have AE. Any alternatives? Can the same operation be done in Capcut itself?

@FLaMeOfNoVa Capcut is capable of tracking and masking, so I would assume that it is possible. Awesome content!

@FLaMeOfNoVa seems I need to know more about capcut 🥺🥺🥺

looks like something I would watch

Much better than eleven labs btw:

Try Seedream!!! It can make really good text and if you give an image as reference it will keep style consistency too

Your work is inspiring. Thanks

@DeepThinkerAI From Silos: but, not today 😉

i think we are no to far away from a framework in which you could overlay the text and then pass it through ai again to make it blend naturally in the video. In any case your work looks already better then , one frame man

Wich model for creating image ? And wich tools to change angle etc

Can you create one for me ?

Check out @jetsydev my man

What model is this?

'grind'

Can you share your formula?