Foundry has launched the new release of their flagship product Nuke, with lots of news in this version of Nuke 13. The Foundry team has done a great job, let’s see all the news of this new version.

The CopyCat node is incredible! Can’t wait to use it in production!

1 – Making a Data Set | Creating Mattes Using CopyCat | Machine Learning in Nuke

In this series, you will be introduced to the CopyCat node in NukeX before learning how you can use it to create a garbage matte. Using CopyCat you can train a neural network to replicate various tasks, dramatically speeding up your workflows. In this video, you will learn how to create an effective data set of images to use to train the network.

2 – Training and Monitoring the Network | Creating Mattes Using CopyCat | Machine Learning in Nuke

3 – Applying and Improving the Results | Creating Mattes Using CopyCat | Machine Learning in Nuke

Nuke 13.0 | Overview

Introducing Nuke 13.0

In Nuke 13, VFX artists have the ability to work on the look of the final image with the introduction of Hydra support in Nuke’s 3D viewport, delivering an improved quality image much closer to rendered output.

Foundry has also extended Sync Review and added annotations in HieroPlayer, improving review workflows for teams so they can easily collaborate even when working remotely.

Introducing a new machine learning framework, which allows artists to create bespoke effects directly in NukeX, with a high quality result in a relatively short time.

With CopyCat and Inference, the digital compositor artists are able to harness the power of machine learning to accelerate tasks from removing motion blur to upres and more.

Nuke 13.0 also offers artists performance optimizations and updates to industry-standard libraries, so artists and teams can get the final pixel-perfect image faster than ever.

Download Nuke 13.0
Article Comments
Nuke 13 Ratings
Reader Rating6 Votes
New Machine Learning Framework
Great Control on Display
GPU, High Quality Result in Short Time
Stability and Usability are Improved