Create a Missed Train Dance Scene

Watch a full-body movement reference

Create your station dance

1. Prepare the two inputs

   This workflow uses exactly one character image and one user-uploaded MP4 or MOV dance-reference video. For a Missed Train Dance, choose a clear image of the person whose appearance you want to guide the scene. Pick a one-person, continuous routine with face, body, and feet visible, no rapid cuts, and a camera orientation you can read. If station context matters to the movement, a reference filmed in a station can help show it—but neither the reference nor the character image guarantees an exact match to a setting or routine.

2. Set the station story in the prompt

   Edit the starting prompt to describe the moment, setting, framing, and ending you want. For example, have a traveler begin after a missed departure, make a small step and shoulder move in clear pedestrian space, and finish calmly. Keep a bag or case still and ask for the full body to remain visible. The uploaded video guides movement order and camera orientation; your prompt and character photo add creative direction. Review the result rather than assuming it will replicate the choreography, background, or identity exactly.

3. Check the quote and generate

   Review the upload control for current limits: the video must be MP4 or MOV, 3–30 seconds, and up to 32 MB. The server reads the file’s actual duration and calculates the charge using the rounded duration and selected model and quality. Check the quote beside Generate before starting; do not rely on a fixed price. The motion model is currently a paid option, so check your access to that model, your purchased credits or account balance, and the live quote. Eligibility for any free use depends on the account and current offer; there is no free-credit promise.

4. Review the task and finish the video

   Find the task and its status in History. Wait times can vary, so there is no fixed completion estimate. When the video is ready, review the full playback: check the person’s appearance and outfit, whether the movement follows the reference clearly, whether the station context reads, and whether the ending feels clean. Also inspect the framing, feet, bag, and pedestrian space. If needed, adjust the prompt or inputs and try again. Handle any additional crop or audio finishing elsewhere; those are not controls in this workflow.

What the Missed Train Dance workflow does

This page is for making a short, character-led comedy scene from one character image and one dance-reference video. The chosen video guides movement order and camera orientation, while your image and editable prompt provide the character and story direction. The imagined scene here is deliberately restrained: a traveler realizes her train has gone, makes a small step and shoulder move, and ends clearly in open station space. The idea is to make the moment readable, not to imply an automatic missed-train transition or a separate train animation.

Choose a reference that is easy to follow

Prefer a single person moving continuously, with the face, body, and feet visible for most of the clip. A steady view makes the sequence and camera orientation easier to interpret; rapid cuts can make them harder to read. For a subtle station comedy, a simple routine with a few distinct movements is often a better creative fit than a crowded or fast-changing scene. Keep the reference within the upload limits and consider whether its background affects the action. A character image and prompt can suggest a warm wood-and-glass hall, broad windows, a side bench, and late-afternoon light, but you should review the output for how well the setting and identity carry through.

Make the ending and framing count

A clean finish helps a short comedy scene feel intentional. Describe a calm final pose or moment of stillness, leave clear space around the character, and keep the full body in frame so small footwork remains visible. In this example, the traveler stays in pedestrian space and her navy bag remains still as she makes a restrained movement. Use the prompt to clarify these choices, then judge the actual playback: results may vary in how they follow the reference, preserve the character’s appearance, and convey the station setting. Do not assume exact choreography or background replication.

Understand the controls and review before sharing

The page provides an editable starting prompt, the currently enabled motion model, output quality options, a quote beside Generate, and task status in History. Upload the character image and reference clip, then edit the direction before generating. Check the upload control and live quote before you generate; duration-based billing uses the server-read duration rounded according to the selected model and quality. After completion, review the whole video for movement, appearance, framing, and ending.

Missed Train Dance questions

More ideas to explore

Video-to-Video AI

Explore another video workflow when you want to work from video input.

AI Video

Browse the AI video tools available on the site.

AI Image Generator

Explore image-generation options for visual concepts.

Earth Zoom Generator

See a different AI video concept built around an Earth-zoom effect.