Hotel Lobby AI prompt: what text can and cannot control

Do you need a prompt on AI Hotel Lobby?

No. The Hotel Lobby generator on this site uses a prepared video and two photo slots, not a free-form text scene. You control the source photos, role order and output quality. The template controls the scene and performance reference. Pasting a long prompt into an unrelated field will not change the fixed duo template.

When is a text prompt useful?

A prompt is useful if you are using a separate video service that accepts reference images and a scene description. That route gives you more scene freedom, but each model supports different inputs, lengths and audio options. Check the chosen service's current documentation. A text description alone cannot guarantee two stable identities, the same choreography, an existing song or accurate lip sync.

An original scene brief you can adapt

The example prompt below describes a general studio duo; it does not contain copyrighted lyrics or claim to reproduce a named artist. Treat it as a starting brief for a service that actually supports two reference images. Replace the framing or movement only when that service accepts those controls. Use authorized adult portraits and your own or licensed audio.

Create a landscape studio performance with two consenting adult subjects from separate reference photos. Put the first subject on the left and the second on the right. Use a plain orange background and a single microphone suspended between them. Keep a fixed medium-wide camera, both faces visible, natural hand gestures and subtle rhythmic movement. Preserve each subject's facial features, hairstyle and clothing separately. Avoid extra people, added text and logos. Use only audio I have permission to include.

If the two faces blend or swap

Confirm that the service accepts two distinct references, then explicitly assign left and right roles. Remove group photos and choose two clear portraits with visible hair and shoulders. If the model still cannot preserve the subjects separately, a more elaborate prompt may not solve its capability limit. Compare a new attempt against the previous output before paying for further retries.

If the rap or lip sync is wrong

A silent video model does not acquire speech from a scene prompt. Audio generation, audio-to-video and reference-motion editing are different capabilities. For this site's fixed template, the original template soundtrack accompanies the recast clip, but precise lips and likeness remain model-dependent. Custom voice, song upload and arbitrary lyrics are not available in the current duo workbench.

Choose the workflow before writing more text

Use a fixed template when you want the demonstrated scene and a straightforward photo-to-video workflow. Use a separate prompt-based model when you need a different setting and can review its input support, costs and uncertain results. Neither workflow grants rights to another person's likeness or a third-party track. Compare the sample, actual output format and final price before choosing.