Fast 4× super-resolution

Refocus demos

The first act of the flagship upscale, isolated. These clips run the Refocus stage only — motion-aware 4× super-resolution with no generative pass — so you can see exactly what the model recovers from the source. Then scroll down for head-to-head comparisons against Topaz Video AI on identical footage.

Refocus stage only

Original vs. Refocus

No generative pass — just the motion-aware super-resolution stage. Drag the wipe, zoom in, and step frames to inspect what gets recovered.

Action on water

Surf footage — spray, foam, and fast motion

ORIGINALREFOCUS
0:00.00 / 0:00.00
1×
A B
Drag to wipe · Space play/pause · ← → frame-step

Select clip

Head-to-head

Refocus vs. Topaz Video AI

Same source footage, processed through Refocus and through TopazLabs Proteus. Left is Refocus output, right is Proteus. Slide, zoom, and frame-step to judge for yourself.

Portrait

Skin texture and facial detail on the same source clip.

REFOCUSTOPAZ PROTEUS
0:00.00 / 0:00.00
1×
A B
Drag to wipe · Space play/pause · ← → frame-step

Select clip

Try it on your own footage.

The free trial runs every model locally — full videos, watermarked, no credit card.