You've seen that: Midjourney just dropped a game-changer — one-click image-to-video, and the results are insane for anything human-shaped.
But what happens when base image is more abstract, less perfect? 🤔
I put it to the test with some interactive 3D visuals I've made last year (which are not perfect).
I took an image from my videos of the scene and dropped it in midjourney.
I made a video of it to compare it to the actual video I've made.
Can you spot which one is the real video?