Skip to content

Upcoming Release of Silhouette 2024.5: Enhanced Rotation and Painting Functionality Through New Machine Learning Features

Instantly generate mattes, add realistic motion blur, and streamline your rotoscoping quality assurance process.

Update on Silhouette 2024.5: Enhancements in Rotation and Paint featuring Innovative Machine...
Update on Silhouette 2024.5: Enhancements in Rotation and Paint featuring Innovative Machine Learning Resources

Upcoming Release of Silhouette 2024.5: Enhanced Rotation and Painting Functionality Through New Machine Learning Features

Silhouette 2024.5, the latest release from Boris FX, brings a host of advanced machine learning (ML) features to revolutionise roto, matte, tracking, and retime processes in visual effects workflows.

Streamlined Roto and Matte Creation

The new version introduces several ML-powered tools designed to speed up and simplify the rotoscoping and matte creation processes. Notable additions include Matte Assist ML, Mask ML, and the Roto > Markup Tool.

  • Matte Assist ML allows for quick creation of animated mattes from a single frame using local ML models, significantly reducing the time required for rotoscoping.
  • Mask ML enables point-and-click matte generation with ML-driven object segmentation for precise masking.
  • The Roto > Markup Tool simplifies rotoscoping quality assurance by adding trackable strokes and linked notes for easier communication.

Improved Tracking and Retime Capabilities

Silhouette 2024.5 also introduces an Optical Flow ML Tracker that generates optical flow data to track shapes, objects, or points with integrated realistic motion blur. Additionally, Retime ML creates clean time warps and animated ramping effects using AI for smooth temporal adjustments.

Enhanced User Interface and Collaboration Tools

The UI in Silhouette 2024.5 has been made more responsive, with rendering implemented in a thread so the UI remains responsive during playback and when dragging through render-intensive nodes/trees. Progress Bars are displayed for EZ Mask, Flow->ST, Matte Assist ML, and Power Matte above the status bar as well as within these nodes in the tree, providing better feedback during the rendering process.

Moreover, the new Markup tool allows for detailed notes to be left directly on shots for improved collaboration among roto artists and supervisors. Node Highlighting is also implemented, providing a tree render status by highlighting each node as it processes using a blue outline.

Availability and Pricing

Silhouette 2024.5 is available as a standalone application and plugin for Adobe After Effects/Premiere Pro, Autodesk Flame, Blackmagic Resolve/Fusion, and Foundry Nuke. To download the free trial, visit Boris FX’s official website.

Pricing for Silhouette 2024.5 starts at $495/year or $100/month. Existing customers with an active subscription or upgrade & support plan can upgrade to Silhouette 2024.5 for free.

Learning Resources

For those wanting to make the most of the new features, Ben Brownlee, director of content, offers tips on using the new Mask ML node together with Matte Assist ML to produce precise animated mattes in seconds.

Conclusion

With its focus on ML-powered tools, Silhouette 2024.5 promises to enhance productivity, accuracy, and creative control over the roto, paint, tracking, and retime processes in visual effects workflows. Download the free trial today and experience these powerful new capabilities for yourself.

  • The advanced machine learning features in Silhouette 2024.5, such as Matte Assist ML, Mask ML, and the Roto > Markup Tool, demonstrate the intersection of data-and-cloud-computing technology and the visual effects industry.
  • The integration of technology like the Optical Flow ML Tracker and Retime ML in Silhouette 2024.5 exemplifies how cloud computing and artificial intelligence are revolutionising data workflows in visual effects.

Read also:

    Latest

    Enhancing Accuracy in Brain Tumor Identification

    Enhancing the Identification of Brain Tumors

    Researchers at Dalhousie University, Canada, have made public a collection of over 1,800 ultrasound images from mouse brains for the purpose of educating Artificial Intelligence systems in identifying brain tumors more precisely. The images, sourced from live and deceased mice, feature various...