Skip to content

Photo preparation

Photo requirements for your AI dance video

Upload a clear photo of one person in JPG, JPEG, PNG or WebP format, up to 8MB per file. A regular selfie or head-and-shoulders portrait is enough. Solo needs one photo; Duo needs two separate photos.

Check the face, file format and file size before uploading. A full-body photo is not required.

Supported photo formats and the 8MB limit

The uploader accepts JPG, JPEG, PNG and WebP. Each photo can be up to 8MB. In Duo, the same limit applies to each of the two files; use a separate photo for each role.

HEIC is not supported. If a phone photo is saved as HEIC, export or convert a copy to JPG before uploading. Renaming a .heic file to .jpg does not convert the image. Open the exported copy to check that it still shows the face clearly.

CheckRequirement
File formatJPG / JPEG, PNG or WebP
File sizeUp to 8MB per photo
Solo1 photo for the dancer
Duo2 separate photos: dancer, then onlooker
HEICConvert or export to JPG first

What makes a useful selfie for an AI dance video?

Use a photo where one face is easy to recognize without zooming in. A front-facing view or slight angle gives the model a clear reference. Ordinary daylight or an evenly lit room is enough; a professional portrait is unnecessary.

A head-and-shoulders crop is suitable. Leave the face intact instead of cropping through the forehead or chin. The template creates the body, costumes and barn, so you do not need to photograph a dance pose or find a matching background.

  • Keep the eyes, nose and mouth visible.
  • Choose a sharp original photo where possible.
  • Avoid sunglasses, masks, hands across the face and heavy shadows.
  • Avoid group photos and heavy filters that change facial features.

Preparing two separate photos for Duo

For Duo, pick one photo of the dancer and a second photo of the person who will react. Upload them to their labeled slots. A picture of two people together is not a replacement for the two separate uploads.

Review both previews before continuing. If the people are in the wrong roles, replace the photos before starting generation. Both photos should meet the same clarity, format and size requirements. The second person is an onlooker, not a second dancer.

What your photo controls, and what it does not

Your photo guides the character’s facial appearance. The template supplies the setting and clothes: a green tailcoat for the dancer and a white dress for the onlooker. Changing the background or outfit in your input photo does not request a new video setting or costume.

Clear input helps, but an exact facial match is not guaranteed. AI can change facial details, expressions and movement. Only upload a photo you have permission to use, and consider whether the person pictured is comfortable appearing in a dance video.

If a photo will not upload

First check the actual file format and size, rather than relying on the filename alone. Export a supported copy if needed. If the file is too large, resize or compress a copy and inspect the face again before using it.

A screenshot may make a smaller file, but it can also reduce detail. Prefer a resized original when available. Fix any visible upload error before continuing, then confirm the correct photo appears in each role preview.

Questions and answers

Can I upload HEIC from my iPhone?

HEIC is not supported. Export or convert the photo to JPG, then upload that copy. Each file must be no larger than 8MB.

Do I need a full-body photo?

No. A clear selfie or head-and-shoulders portrait is enough. The template creates the dance, body, costume and scene.

Can I use one group photo for Duo?

Duo requires two separate photo uploads, one for the dancer and one for the onlooker. Choose a clear single-person photo for each slot.

From photo to MP4

Upload photos to generate your video.

Solo uses one photo. Duo uses two: a dancer and an onlooker.

Check the generator for current availability.

Generate video