Pick two clear portraits
Choose one clear photo for each adult performer. A visible face, natural expression and similar camera angle give the video model a stronger reference for keeping two people distinct.

TWO FACES. ONE MIC.
Hotel Lobby AI is the two-person duet trend set in an orange recording booth. Upload two photos and we generate the duet video for you — scene, moves and audio.
Make your duetTHE LOOK
Two subjects stand on opposite sides of one hanging microphone while the camera holds steady. One leads, the other reacts, and then they trade the moment. Friends, partners or family members can appear together even when their source photos were taken separately.
This page makes that look for you. Choose two portraits, decide who appears on each side and optionally add a styling note. We compose the scene with GPT Image 2 and animate it with DreamActor motion transfer, which carries over the reference performance and its audio.
Think of it as a short scene. A good result keeps each person recognizable, gives both people room and places the microphone between them. Orange walls, warm light and a steady camera unify the setup. Results vary with the model and photos, so review the clip before posting.
MAKE THE SCENE
Choose one authorized adult portrait for each side and add an optional styling note. We compose both people into the orange booth, then transfer the Hotel Lobby performance — moves and audio — onto them.
Video size
Pick a photo for each side and confirm consent to start.
THE PROCESS
Choose one clear photo for each adult performer. A visible face, natural expression and similar camera angle give the video model a stronger reference for keeping two people distinct.
Write a simple sequence: person A begins, person B reacts, then the lead changes. A fixed camera and small gestures usually make the first attempt easier to judge.
Press Generate. We first compose both people into the orange booth, then transfer the Hotel Lobby performance onto them. Review identity, hands, the center microphone and left-right placement before you share.
MAKE IT YOURS
The duet works best when the relationship between the two performers is easy to understand. Start with one moment, one camera angle and a clear handoff.

Make a call and response: one friend takes the first line, the other answers at the mic. Use eye contact, smiles and small gestures to show their connection.

One sibling starts confidently; the other joins on the next beat. Keep them on opposite sides of the orange booth and let their expressions tell the story.

Give a couple a quiet first verse and a shared chorus. Keep the gestures small and let a glance across the microphone mark the moment they come together.
BETTER SOURCE MATERIAL
A precise prompt helps, but the photos you choose set the ceiling for the final scene. These checks can reduce common identity and composition problems before you generate.
Use sharp images with full faces and enough clothing to guide the scene. Avoid heavy filters, sunglasses, crowded backgrounds and cropped faces. Similar lighting and perspective can make the duet feel more consistent. If one portrait is a close-up and the other shows a full body, crop or replace one so the references match more closely. Neutral expressions leave more room for a believable reaction. Only transform images you have permission to use.
Describe one short exchange: a lead singer, a listening reaction and a handoff. Ask for natural mouth movement and small gestures while the camera stays centered. State whose lips move in each beat so the performers do not appear to sing the same line at once. Fast turns and sudden cuts can make faces or left-right positions harder to preserve.
Watch the clip before sharing. Confirm that A stays left, B stays right and the microphone stays between them. Check for warped hands, merged faces, extra people or changing outfits. Watch once with the sound off: the handoff should still make visual sense. If a face changes midway, shorten the scene before adding more movement. If the scene breaks, simplify your direction and try again.
PROMPT NOTES
For a Hotel Lobby AI duet, the starter prompt names the main visual anchors: a vertical 9:16 frame, a matte-orange booth, two separate people, one black microphone, soft light and a locked camera. It also asks the model to preserve faces and outfits and avoid side swaps, added people, captions and watermarks. These constraints give the generator a clearer shot list than a vague request to make two photos sing together.
Use the direction field for one styling detail, such as outfits, accessories or expressions. Keep it short so the core composition still leads — the booth, the center microphone and left-right placement are fixed by the base prompt. If the first result is off, change one thing at a time and generate again.
Buy credits once or subscribe. Credits are only spent when you generate.
A 5s 720p video = 140 credits · 1 image = 160 credits
Basic
For casual creators
Pro
For regular posting
Studio
For creators and small teams
GOOD TO KNOW
Yes. Sign in, upload one photo for each side and press Generate. The scene is ready in under a minute and the finished video in about 3–5 minutes. Each video uses credits.
When you press Generate, your two photos are resized in your browser and sent to our AI provider (fal.ai) to create the scene and the video. They are not uploaded before you press Generate. Only use photos you have permission to use.
Yes. Use one clear reference for each adult, with A on the left and B on the right. Identity preservation depends on the generator and photos, so inspect the result.
Models can drift when references are unclear or actions are busy. Use visible faces, remove camera moves and describe a shorter exchange. Repeat left-right placement and request one continuous shot. A usable clip may take several tries.
Yes. The orange booth can frame a couple, friends or adult family members. Change the performance direction but keep identities separate. Get permission from everyone pictured and identify the result as AI-generated when sharing.
Planning and previewing are free. Generating a video uses credits; the exact amount is shown next to the Generate button and on the pricing section. Buy a one-time pack or subscribe monthly.
It is a two-person duet scene in an orange recording booth. One person starts a line beside a hanging microphone, and the other responds. Creators often use two separate photos to plan a shared AI video scene.
Yes. The look is tool-agnostic: any image-to-video tool that accepts two reference photos and a vertical 9:16 clip can recreate it. Generating here just saves you the manual steps.
Choose your two performers and make the lobby scene your own.
Make your duet