Hotel Lobby AI and AI Rap Duo: A Two-Portrait Guide
Prepare two portraits, assign their roles, and review an orange-studio duo video. Learn what the preset does and what it does not promise about music or likeness.
Hotel Lobby AI and AI Rap Duo describe the same creative idea on Dola AI: two portrait references become a short performance in an orange studio, with a central microphone and coordinated movement. Start in Hotel Lobby AI. You do not need a photograph of the two people together.
The preset requests a ten-second, 16:9 video through the configured MiniMax H3 route. It supplies a scene and movement brief. It does not include the original song, copied lyrics, or an exact recording of named performers. Any generated sound and lip movement need to be reviewed.
Prepare two separate portraits
Choose one adult per photo. A clear, front-facing image with visible eyes and an unobstructed mouth is easier to assess than a group photograph or a heavily filtered selfie. Use similar face sizes where possible. The backgrounds do not need to match.
Avoid cutting through the forehead or chin. Keep an original copy of each photograph and confirm that both people agree to having their likeness animated.
Give each person the right role
- Add the first portrait to Left person.
- Add the second portrait to Right person.
- Check the thumbnails. Use the swap control if the order is wrong.
- Review the actual credit quote before creating.
Uploading a reference is preparation. The generation starts when you submit the request. Read the form's current account and credit requirements rather than assuming that viewing the public examples means a new video is free.
Review a whole clip, not its best frame
Watch the first second to establish which person is on each side. Continue through the middle and final frame: faces can drift during movement even when the beginning looks convincing.
Check four details:
- Each identity stays with the intended body and position.
- Hands do not merge into the microphone or the other performer.
- Both people remain inside the frame.
- Any generated audio suits the clip and contains no unexpected material.
Download an accepted result before making your final edit. For a vertical post, fit the landscape video inside a vertical canvas so both people stay visible; cropping tightly around one person changes the point of the duo.
Choose a different duo when the scene calls for it
The orange studio is useful for a performance. Smalltown Boy AI starts with a reaction and moves into a shared dance. APT-Inspired AI Duo uses a brighter pop-performance setting. Crazy Frog AI is a two-character chase rather than a dance.
These are separate visual briefs. None promises exact original choreography or a licensed commercial recording. If you want a different scene, use AI Video, choose an available model that supports your reference inputs, and describe one short interaction.
A useful custom duo brief
For a new scene in the general video workspace, start with roles, then action, then camera:
Use the first portrait for the performer on the left and the second for the performer on the right. Keep their identities distinct. They stand together in a warm minimalist studio, exchange a quick amused glance, and make a small coordinated shoulder movement. Keep both faces visible in one stable medium-wide shot. No captions, logos, or copied recording.
This is an original starting prompt, not a tested guarantee. Keep the motion simple enough that you can judge whether both references remain recognizable.