What a Hotel Lobby AI prompt should specify
A description of a microphone cannot reproduce a particular sequence of gestures on its own. A video reference provides more direct motion guidance, while the portraits supply the intended identities. Results still depend on the model and inputs.
- Scene: a flat orange studio and one hanging microphone between two performers.
- Identity: one separate image reference for each performer.
- Roles: first portrait on the left, second portrait on the right.
- Performance: use the supplied video reference for staging, motion and pacing.
- Output: avoid unwanted text or logos, then review the generated details.
A prompt example for a compatible reference tool
Use this wording only in a tool that supports both multiple identity images and a video reference. Its syntax and input limits may differ. The example is guidance, not a tested guarantee for another provider.
This generator already applies its fixed prompt. You cannot replace the reference, change the lyrics or paste this example into the page. Your controls are the two portraits, duration, resolution and frame.
Create a two-performer rap scene in a flat orange studio with one hanging microphone. Use image reference 1 for the left performer and image reference 2 for the right performer. Use the provided performance video for staging and motion guidance. Keep each performer recognizable and assigned to their side. Avoid text overlays and extra performers.
Improve the references before adding more words
- 01
Check each face
Eyes, nose and mouth should be visible without sunglasses, masks or motion blur.
- 02
Keep the people distinct
Choose different clothing or hairstyles where possible. Similar references can make roles harder to follow.
- 03
Include useful outfit detail
An upper-body image gives more clothing context than a tight face crop. Keep the face large enough to remain clear.
- 04
Review the chosen output
Check all scene cuts and the sound. A stronger prompt is not a substitute for inspecting the finished MP4.