Alors On Danse AI Video · Prompts
Understand the two-photo prompt, character positions and ways to improve your references.
Last updated:
Do I need to write a prompt?
No. Alors On Danse AI Video uses a built-in prompt and a studio motion reference. Upload two photos, choose your resolution and aspect ratio, and generate. The current creator does not have a custom prompt field. The examples below explain the intended scene; they are not an additional setting you must enter.
How the two photos are used
The first photo supplies the subject on the left at the keyboard. The second supplies the partner on the right. The motion reference guides the scene, camera and performance. Clear, separate references help distinguish the two subjects, but facial resemblance and gestures can still vary.
Example scene description
Create a two-person studio performance. Use the first photo for the keyboard player on the left and the second photo for the partner on the right. Keep the subjects distinct throughout the clip. Follow the reference composition and motion while preserving the identity cues from each supplied photo.
Portraits or full-body photos?
A clear portrait gives the model more visible facial detail. A wider photo may also influence hair, clothing and overall appearance. Avoid collages, heavy filters, covered faces and photos containing multiple people. Use JPG, PNG or WebP, up to 10 MB per photo.
Ideas for a contrasting duo
Try your own authorized photos with formal clothes, streetwear or traditional outfits. Pet examples are also available in the gallery. These are creative directions, not guarantees that every garment, face or animal marking will match exactly.
If the face does not match
Check that each preview shows the intended photo and wait for both uploads to finish. Try a clearer reference with even lighting. The system starts the provider task after preparing both image references; a successful upload alone cannot guarantee identity accuracy. Review the entire output before sharing.