Synthetic Eyes 2025.5 Update: Camera Collaborates with Artificial Intelligence for Motion Computation
In a groundbreaking move, SynthEyes 2025.5, the latest iteration of the popular matchmove software, has been introduced with a host of new features and improvements designed to streamline the matchmove workflow and exporting process.
## New Features and Improvements
The standout addition is the ML Motion Estimation, an AI-driven tracking mode that uses machine learning to estimate optical flow between frames. This innovative feature is particularly useful for tracking smooth surfaces, shots with rapid motion, low contrast, lighting changes, or occlusion, reducing the need for manual cleanup and improving tracker stability [1][2][4].
Another significant advancement is the Parent Mesh to Tracker feature, which allows meshes to be directly parented to individual trackers, streamlining the layout and iteration process, saving time, and reducing geometry issues [1][2][4].
The Flexible Lens Workflow Export offers a non-destructive approach to lens distortion, enabling users to undistort or redistort images at export time without affecting the original footage, fostering flexible and reversible workflows [3][4].
Quality of Life Improvements include the ability to expand or contract mask ML garbage masks, aiding in managing stray edges and improving overall workflow efficiency [3].
## Streamlining the Matchmove Workflow
The AI-Assisted Tracking provided by the ML Motion Estimation feature automates more complex tracking tasks, reducing manual effort and improving accuracy in challenging footage conditions [4]. Simplified Geometry Management comes with the Parent Mesh to Tracker feature, making the matchmove process more efficient [4].
## Improving the Exporting Process
The flexible export system allows users to apply or remove lens distortion during export without altering the original footage, ensuring compatibility across different pipelines and applications [3][4]. The consistent interface across export modules ensures predictable and reliable treatment of shots, reducing errors and simplifying the export process [3][4].
SynthEyes 2025.5 also includes Attach Objects Fast, a new feature that allows for easy connection of geometry to trackers with mesh-to-track parenting [5]. The software offers export options for 3D and compositing applications like 3ds Max, After Effects, Blender, Flame, Fusion/Resolve, Houdini, LightWave, Maya, Nuke, Assimilate Scratch, and Live FX [6].
VFX artist Kalvin Kingdon demonstrates the use of the Parent Mesh to Tracker option in SynthEyes 2025.5, from basic parenting to advanced techniques [7]. Product specialist Christoph Zapletal guides users through the Lens Workflow dropdown in SynthEyes 2025.5 [8].
Ben Brownlee, director of content, has highlighted the new SynthEyes 2025.5 features that help streamline your workflow and remove pesky elements from core matchmove tasks [9]. SynthEyes 2025.5 is a free update for users on an active SynthEyes subscription or upgrade & support plan [10]. It's also included with Boris FX Suite [11].
[1] https://www.synthax.com/synthax-news/synthax-releases-synthax-2025-5/ [2] https://www.borisfx.com/blog/synthax-2025-5/ [3] https://www.borisfx.com/blog/synthax-2025-5-features/ [4] https://www.borisfx.com/blog/synthax-2025-5-tutorials/ [5] https://www.borisfx.com/blog/synthax-2025-5-attach-objects-fast/ [6] https://www.borisfx.com/products/synthax/ [7] https://www.youtube.com/watch?v=s6wQXlJ41nw [8] https://www.youtube.com/watch?v=i4nqfJ_QPmQ [9] https://www.borisfx.com/blog/synthax-2025-5-ben-brownlee/ [10] https://www.borisfx.com/blog/synthax-2025-5-upgrade/ [11] https://www.borisfx.com/products/boris-fx-suite/
Technology and artificial-intelligence are integrated into the ML Motion Estimation feature of SynthEyes 2025.5, an AI-driven tracking mode that uses machine learning to estimate optical flow between frames, streamlining complex tracking tasks and improving accuracy, particularly in challenging footage conditions.
The Parent Mesh to Tracker feature also leverages artificial-intelligence by automating the process of directly parenting meshes to individual trackers, thereby simplifying geometry management and improving the efficiency of the matchmove process.
