1.x Re:Vision Effects REZup

Enhance the quality (“IQ”) of your content, increase the resolution of your video without loss of detail or introducing artifacts.

Upscale your content with RE:Vision Effects REZup. Improve image quality (“IQ”) and get higher resolution video without compromising any details or introducing weird artifacts.

REZup contains two plug-ins: REZup Enhance and REZup Resize. These plugins provide a unique mix of state-of-the-art video image processing augmented with machine-learning techniques.

REZup Features

Enhance sharpness, contrast, and details

Improve your image “IQ”. Reveal details you did not even know were there.

Upscale with the state-of-the-art A.I.

Better than standard resizing filters like Lanczos or cubic.

Resize or Enhance

Resolution enhancement (same dimension in and out) or upscaling (downscaling too).

REZup features that are uncommon for software using AI Deep Neural Networks (DNN) models include:

• Supports fractional resizing factors
• Supports semi-transparent material (alpha sensitive)
• Supports extended dynamic range values
• Well-thought-out interface that guides you to address different issues
• Handles different scaling factors for X and Y dimensions (so will work directly on anamorphic video)

REZup Enhance

Smartly guides you through needed steps to improve the image quality of your footage. Control and adjust the look, sharpness, detail, and contrast to your liking. For example, for a close-up portrait, it might be preferable to tone down adding too much detail, as who wants to see super-clear skin blemishes? You have complete control of what it should look like.

REZup Resize

Adds zoom, ratio, and a reframing window for handling various upscaling scenarios. REZup Resize is trained and tested with 2X, 3X, and 4X models, however, at larger upscaling factors you might need to preprocess your clip with a temporal denoiser like DE:Noise (also part of Effections) or apply REZup Enhance before for best results.

REZup demo 2X

Side-by-Side comparison of a regular resizer versus REZup. Note we cranked up REZup settings on some shots so you can see it does something if you are watching on a small display.

Demo 2X – Lanczos versus REZup

Lancsoz (included for reference in REZup) is considered the best classical upscaling resampling method because it adds a bit of sharpening. Here we push REZup probably a bit more than you want so we can see.

REZup demo 1080P in

Starting from Hd and scaling up 2X and trying 3X and 4X. Note DE:Noise was used for the 4X example resulting in a bit softer image but without scaling up the noise.

Feature Specs

• Not limited by integer scale – Fractional resizing factor support (digital zoom).
• Alpha Weighted -Takes into account the alpha value when performing enhancement. Works as expected on semi-transparent values.
• Extended Dynamic Range – Supports arbitrary image ranges — that is pixel values under 0 and over 1.0 (white).
• Source Cropping -Supports a source window to frame your resizing.
• Adaptive UI – UI guides as you make changes, starting safe and simple. Supports various algorithms to handle a variety of edge cases and source footage.
• Tweak to your liking – Separate controls for sharpness and contrast, noise reduction – not a black box. Make it look just right.
• Machine Learning Models – Trained on 2X, 3X, and 4X. Larger scaling factors might require pre-processing (e.g. denoising)
• Functionality varies per Host* – Method to change output resolution of video varies between each app: Workarounds
• Any Resolution – Great way to upscale your HD video to UHD (4K)

You can also get REZup in RE:Vision Effects Effections Bundle and RE:Vision Effects Effections Plus Bundle.