How to Make a Hotel Lobby AI Video

1. Prepare the references

Use one clear image for each performer. Keep faces visible, avoid heavy filters, and choose images with similar crop and lighting when possible. Only upload images you own or have permission to use. Do not use a real person's likeness without consent.

2. Assign the positions

Write the roles explicitly before generation:

Reference image 1 = the performer on the LEFT.
Reference image 2 = the performer on the RIGHT.
Keep both positions fixed for the entire clip.

This simple constraint prevents one of the most common failures: identity and position swapping between shots.

3. Describe the invariant scene

State the orange studio, central hanging microphone, warm lighting, framing, pacing, and camera behavior. Then describe the action order. Keep the prompt focused on what must remain stable rather than adding unrelated style adjectives.

4. Generate a short draft

Start with the shortest supported duration and a vertical aspect ratio. Review the draft before spending more credits or adding complex motion. Check faces, hands, microphone placement, lip sync, audio, and side assignments frame by frame.

5. Export and label

Export a version suited to the destination feed, preserve the source and tool metadata, and label synthetic media clearly when publishing. Do not distribute the original song or other copyrighted audio unless your use is licensed.