Choose the photo where you can really see them.

A quick source-photo checklist

• Clear eyes and mouth: avoid heavy blur, deep facial shadows, or glare across the face.
• Enough resolution: at least 512 pixels on both sides. Enlarging a tiny screenshot does not recover missing detail.
• A natural pose: a face looking roughly toward the camera makes a subtle smile easier to judge.
• Original file where possible: repeated screenshots and messaging-app copies may lose detail.
• Usable format: JPG, PNG, or WebP, less than 10 MB. Convert unsupported formats before uploading.

Leave room for the movement you want

A close crop around the face suits a gentle head inclination, eye contact, and a growing smile. It is not a good starting point for a hand gesture if the hands are absent.

For a ten-second wave, choose a photo of one person with the hands visible and enough room around the upper body. Even then, AI can distort fingers; a person reviews the result before delivery. A simpler movement may be the better choice.

Take extra care with group photographs

Each person needs a face large enough to see clearly. A wide group shot with many tiny faces gives the model less detail to work with. If appropriate, crop a copy around the people you want to include while keeping enough resolution.

For groups, restrained individual movement is preferable to asking everyone to wave or turn at once. You must have permission or the appropriate authority to create a portrait of everyone shown.

Avoid changing the face before animation

Strong beauty filters, aggressive smoothing, or heavy restoration can change someone’s features before the video is even made. Keep an untouched original and compare edits with it.

If the only photograph is damaged or very blurred, contact support before ordering. Animation cannot reliably reconstruct details that are no longer present, and the standard order does not guarantee restoration.