Magnific shows multiverse panel technique for AI video with masked reveals

Magnific (@magnific) shows an AI video method that generates six synchronized clips of one rider on a galloping horse, each set in a different environment.

The clips are stacked and overlapping rectangular masks animate to reveal each one. Magnific says the prompt keeps the horse, rider posture, timing and camera static, so the subject reads as one continuous figure across the panel boundaries. The post is a demonstration from Magnific; no further details on availability or results are reported.