Understand the Reference Before Creating
The October 13 campaign post is an AI dance edit referring to Kendrick Lamar’s 2025 Super Bowl performance and Not Like Us. The production model is not verified in the reviewed report.
Appearance comes from portraits
Use your own portrait or participants who agree to take part. A face reference does not establish choreography.
Movement comes from the MP4
Prepare the performance, camera framing, and scene order in the source video. Test one clear solo shot first.
Audio is a separate input
Source audio is merged at export. The tool does not automatically supply a trending song or a new singing voice.
How to Make a Stage-Style AI Dance Video
A practical MoveFrame AI workflow for your own performance footage; not the verified recipe used for the campaign video.
Prepare footage and a portrait
Record a short solo performance with a clear face and stable framing. Use a standard MP4 up to 100 MB and JPG, PNG, or WebP portraits up to 10 MB each. Prepare the stage and soundtrack in your own source footage.
Upload and inspect the timeline
Sign in, choose the MP4, and start upload and scene analysis. Review the original segments before assigning portraits. Look for face turns, hands, microphones, and extra people entering the shot.
Assign characters scene by scene
Add portraits to the character library. Apply a portrait to the current scene, selected scenes, or all generated scenes. For several characters, use separate solo segments; independent targeting within one group shot is unsupported.
Keep original scenes when needed
Choose Keep original for titles, scenery, or difficult overlapping dance shots. Confirm that every generated scene has a saved portrait assignment before submitting the plan.
Generate and review moving frames
Generate the video and compare each result with its original. Check faces, hands, clothing, backgrounds, and transitions. Regenerate a poor segment or retry a failed task; simplify footage when problems persist.
Check source audio and export
Listen to the full preview, then download the MP4. The service merges the uploaded audio; it does not supply Not Like Us, clone voices, or guarantee singing lip-sync. Label the shared work as AI-edited.
Full Tutorial & Troubleshooting
Follow the detailed scene controls, choose better references, and troubleshoot a segment before trying again.
Original Video & Source Notes
The X link identifies the published post. Mediaite’s October 14 report provides the musical and performance context. Neither establishes a MoveFrame AI production claim or supplies reusable source footage.
Marsha Blackburn AI Video FAQ
Source, workflow, group shots, and soundtrack.
Create Your Own Stage Video
Start with a clear solo MP4 and portrait. Review the generated result and disclose that it is AI-edited, without implying a real performance or endorsement.