How to Make an Alors on Danse AI Video From Two Photos

1. Place the photos in the correct performance roles

In the Alors on Danse AI Video Generator, the opening scene is a white studio with a keyboard and laptop. Use a separate JPEG, PNG or WebP portrait for each intended person, up to 10 MB per image. The upload labels identify the short-haired left role and long-haired right role.

Choose roles by following each person through the scene, not just by comparing one opening frame. Longer cuts continue after the keyboard section. The model is asked to keep both reference identities distinct when the people move and interact.

2. Preview the part included in your chosen cut

Every offered cut starts at the original opening. Five seconds keeps a short opening exchange; 12 seconds includes more of the performance; 30 seconds is the longest prepared choice. The displayed completed case is a separate example, rather than a preview of the selected private source segment.

You cannot upload another keyboard performance or request a different starting point in this studio. The supplied scene controls the motion and shot sequence. Your portraits provide the two identities.

3. Compare the source frame with the example frame

The current generation source is 16:9. The separate completed example lower on the page is 4:3. That displayed example is useful for seeing an AI interpretation, but its dimensions do not add a 4:3 option to the current studio.

Choose original, 1:1 or 9:16. Square and vertical output use a centered crop after generation; the narrow frame may remove part of either performer. If seeing both people's interaction matters, inspect the whole preview before accepting a crop.

4. Review the credit quote and confirm

Choose 480p or 720p and read the quote for your selected length. The same credit balance is shared with other EasyVideoAI studios. Changing the crop does not change the generation quote; changing source length or resolution does.

Confirm permission for both photos and review the request once more before submission. Previews are free to watch, while creating a new personalized MP4 requires the quoted generation credits.

View current credit rates and purchase options

5. Check both people after they leave the keyboard

Watch the generated MP4 from beginning to end. Check the reference faces separately, including turns, close-ups and later movement away from the keyboard. AI can change details or miss an identity even when the opening looks right.

The result uses restored audio from the matching source segment. Listen for timing differences, then download from the completed private result page. If you need another attempt, first check the photo and role assignments rather than repeatedly submitting the same uncertain casting.

Choose clearer portrait references