AI VFX with Seedance 2.0 Testing how far AI video transformation can stretch a single piece of footage. One video and a detailed prompt describing each sceneario, try it yourself, and make the changes needed 🔥 Original clip — subject standing still in a medium shot, static locked camera, no camera movement, indoor green screen backdrop. Wears a dark, near-black long-sleeve henley-style top and dark trousers, hands clasped together in front at waist height, soft even frontal light on the face and torso. Preserve the exact standing pose, wardrobe, hand position, and centered medium-shot framing throughout; replace the green screen with a sequence of three environments Photoreal. 16:9. 8s. Grade shifts to match the color temperature and light quality of each environment. NON-IP — generic environments only. SFX only. Static locked shot, camera completely still for the full 8 seconds, identical centered medium-shot framing The standing pose, clasped hand position, and dark henley top and trousers never change across any cut. The background cuts hard between three environments, camera and subject never moving, only the background and its lighting change at each cut: At 0.0s: a crowded metro train car, filled with Asian commuters, the subject standing shoulder to shoulder among many passengers, fluorescent interior light. At about 2.7s: a narrow convenience store aisle in Japan, tall shelves packed with snacks, drinks, and bento boxes on either side, cool bright fluorescent overhead light, small handwritten price tags and Japanese product labels visible on the shelves. Japanese shoppers pass through the aisle in the foreground between the camera and the subject, walking at a natural unhurried pace, briefly and partially occluding the lower frame as they cross but never fully blocking the subject for more than a beat; they pay no attention to the subject, going about their own shopping. At about 5.3s: the summit of a snow-covered Mount Fuji at sunrise, subject standing on snow and rock, low warm sunrise light raking across the peak, thin cold haze in the distance. Lock-down: standing pose, clasped hand position, dark henley top and trousers, and static locked camera framing identical to the source across every environment cut; only the background and its lighting shift at each hard cut. SFX only: faint layered murmur and metro rail rumble under the train scene; murur and the ring of the door opening at the store; thin wind gusts under the Mount Fuji scene.

Looks awesome!

Before you bring a video idea to any creative team, answer these six things first.

I put together a short video walking through answers you should be able to answer before starting a video project. Things like what the video actually needs to accomplish, where it's going to live, what raw materials you should have ready, and who's making the final call on feedback.

None of it is complicated. But most teams skip at least two or three of these and then wonder why the project takes twice as long as it should.

This applies whether you work with us or somebody else. Having these answers ready is the fastest way to get better work out of any creative team you hire.