REZup increases the resolution of video without loss of detail or introducing artifacts. It contains two plug-ins, REZup Enhance and REZup Resize, which pair video image processing with machine-learning techniques. Enhance guides you through the steps to improve footage, with control over the look, sharpness, detail, and contrast. Resize adds zoom, ratio, and a reframing window for different upscaling scenarios, and supports fractional scaling, semi-transparencies, and extended dynamic range, with a denoise option included.
It suits upscaling standard-definition content to 4K, improving the resolution of low-res generative AI videos, creating high-resolution LED video wall installations, and remastering older 720 and 1080P movies. It also works with long-to-render CG and pipelines without loss of bit depth or intermediate video files. Combined with Twixtor, REZup Resize can convert VR footage for newer devices and reframe video without loss of detail.
REZup Resize is trained and tested with 2X, 3X, and 4X models; larger factors may need preprocessing with a temporal denoiser or REZup Enhance first. The upscaling models are demanding on GPU processing and memory—upscaling 1080p to 8K needs at least 8GB of VRAM.