Frequently Asked Questions

Refocused is an AI-powered video enhancement tool that uses advanced machine learning to upscale and improve video quality. It can increase resolution up to 8K, reduce compression artifacts, enhance sharpness, and improve overall visual quality while preserving natural motion.

Refocused works well with most video content including old footage, compressed videos, screen recordings, and standard definition content. It's particularly effective at restoring detail in videos that have been compressed or recorded at lower resolutions. Results may vary depending on the source material quality.

Refocused supports most common video formats including MP4, MOV, AVI, MKV, and WebM. The output is typically delivered in high-quality MP4 format with H.264 or H.265 encoding for optimal compatibility and file size, but users can choose to encode it to their preferred format.

Yes, Refocused can output the enhanced video as an image sequence - we support most image formats, including PNG, JPG, TIFF, and more.

Processing time depends on video length, input/output resolution, enhancement model, and server load. Here are some estimates, based on our testing: Cloud (GH200): • 4K output: 4–8 fps • 8K output: ~1 fps Local (M4 Max, 48GB RAM): • FHD output: 3–4 fps • 4K output: 1–2 fps • 8K output: ~0.5 fps