Two separate photos work

The two people do not need to appear in the same original picture. Upload one portrait for each person. The generator uses them as separate identity references and places both people into the prepared performance.

This makes the workflow suitable for a couple or two friends who do not already have a well-lit photo together. Both people must be adults and must agree to their photos being uploaded and processed.

  • Photo 1 and photo 2 can come from different cameras, places, or dates.
  • Use one person per image rather than a group photo when possible.
  • A public social profile photo is not permission to use someone's likeness.

Choose the performer and partner roles

Before generating, assign one photo to the performer and the other to the partner. The performer receives the featured performance moments and close-ups; the partner appears in the shared scenes. These are video roles, not labels for the people's relationship or gender.

Check the two photo slots before submitting. AI can still change identity details or role consistency, so review the complete result before downloading or posting it.

Use one clear portrait for each person

Choose recent, well-lit photos with the face and hair visible. Front-facing or gently angled portraits usually provide a clearer reference than strong profiles, tiny faces, or heavily filtered screenshots.

If the only source is a photo of both people together, make a separate crop for each person. Keep the whole face and hair in each crop, and do not enlarge a very small face until it becomes blurred.

  • Prefer chest-up or waist-up portraits with a natural expression.
  • Avoid sunglasses, hands over the face, strong beauty filters, and motion blur.
  • Use comparable lighting when possible, but the two photos do not need to match exactly.

What the current duet workflow produces

The current public MV workflow uses two portrait references with prepared performance footage and returns a private 10-second 768p MP4. It is an AI video edit, not a custom model trained on either person and not a promise of frame-perfect face replacement.

The Studio is the source of truth for current availability, framing choices, the displayed credit quote, and the exact route offered before submission. Do not rely on another Raindance AI site's duration or price because similarly named services have different workflows.

  • Input: two consented adult portraits.
  • Roles: one performer and one partner.
  • Current output: one private 10-second 768p MP4.
  • Known limits: likeness, expressions, roles, framing, and motion can vary.

Review both people before you post

Watch the entire result and check that both people are present, the roles are recognizable, and the faces, hands, framing, and movement are acceptable. A completed generation is a new AI output; a different take may look different and may use additional credits.

Use audio you have permission to use. This service does not grant a license to the original Raindance recording. Follow the destination platform's rules for labeling realistic AI-generated media.

Common questions

Do we need a photo together for a Raindance AI duet?

No. Use two separate portraits, one clear image of each consenting adult. The generator assigns the photos to separate performer and partner roles.

Can two friends make a Raindance AI duet?

Yes. The roles describe the video rather than the relationship. A couple or two friends can use the same two-photo workflow with both people's permission.

Which person should be the performer?

Choose the person you want to receive the featured performance moments and close-ups. The other photo becomes the partner in the shared scenes. Check the assignments before submitting.

Will both faces look exactly like the photos?

Not always. Clear portraits help, but AI can change faces, expressions, roles, framing, and motion. Review the finished video before sharing it.

How long is the Raindance AI duet video?

The current public workflow produces a 10-second 768p MP4. Check the Studio before submitting because available routes and specifications can change.

Sources and review

Reviewed on . Sources cover the music and platform context. Check the generator for current availability, output settings, and credit costs.