ats2man

I used tensor.art and wan for this clip, but its no longer possible to render such clips due to recent policy change

absolutely incredible!!

Thanks, that means a lot, I am glad you like it. I presented the screenshot to chatgpt and asked for a detailed prompt describing the scene. after a few back and forth (e.g. working out details such as for the Korean characters in the headband) it came up with a reasonable prompt. Then I fed that into Sora to get these images.

I'm sorry, I don't know how to give official credit on this platform, I tried to make it more explicit in the title.

Yep, its still very difficult to get a natural flow of action, the current models don't quite deliver in terms of basic physics. Not even close to the beauty and the intensity you created with your marvellous pixel art!