The recent announcement that James Cameron’s Lightstorm Vision has acquired Outsyders could have significant implications for the stereoscopic 3D and visual effects industries.
For years, stereoscopic content has largely followed two different paths: native 3D capture using dual-camera systems, and 2D-to-3D conversion during post-production. The acquisition brings Outsyders’ conversion technology into Lightstorm Vision’s existing 3D production ecosystem, potentially bringing these two workflows closer together.
How Outsyders works for 2D to 3D conversion?
Founded by industry veterans Paul Becker, Chris Harvey, Josiah Reeves and Jared Sandrew, Outsyders has focused on one of the most demanding areas of stereo post-production: converting conventional 2D images into convincing stereoscopic 3D.
Traditional conversion can involve extensive rotoscoping, depth-map creation, paint work and compositing. Each shot may require artists to define depth, separate foreground and background elements. It also had to deal with hidden areas and make sure the resulting stereo image remains stable throughout a sequence. For large productions, that can mean significant artist time and long delivery schedules.
Outsyders developed a machine-learning-based approach aimed at reducing some of this manual work. Its technology uses real and synthetic training data to help with depth estimation. It also supports technical details of occlusion and motion, while artists and supervisors remain involved in the process.
One of the more notable demonstrations of the approach came with Disney’s traditionally animated Lilo & Stitch, which Outsyders converted to stereoscopic 3D in eight weeks. The company has also worked on projects including Tron: Ares and Moana.
The important point is not simply that AI is being used in the process. It is that machine learning is being applied to a part of stereo conversion that has traditionally required a large amount of manual artist intervention.
Why the acquisition matters to Lightstorm Vision?
Lightstorm Vision has been developing technologies around native 3D capture, camera systems and real-time viewing. Native stereo capture works particularly well when a production is designed and shot for 3D from the beginning.
But not every project starts that way.
Existing films, television content, traditional animation and CGI projects may not have been created with stereoscopic delivery in mind. In those cases, conversion remains necessary. This is where Outsyders becomes particularly relevant.
Bringing the company’s technology into Lightstorm Vision gives the studio access to a conversion pipeline alongside its work in native 3D capture. It also creates the possibility of using information from the original production—such as camera data, tracking information and other scene information—where available, to improve the conversion process.
The broader goal is to make converted material look less like a collection of artificial depth layers and more like a properly designed stereo image. That remains one of the biggest challenges in 2D-to-3D conversion. Poorly handled conversion can produce familiar problems such as flat-looking depth, unstable edges, excessive image separation and visible warping.
What this acquisition means for VFX Artists?
The acquisition is also interesting from the perspective of artists and VFX studios.
Less manual Rotoscoping
It will lead to more technical supervision for better quality control. Machine learning is unlikely to eliminate the need for stereo artists. Instead, some of the work may shift from creating every depth decision manually to supervising, correcting and refining automated results.
Artists who understand depth, compositing, tracking and stereo principles could increasingly find themselves working alongside machine-learning tools rather than performing every step manually.
More 2D content could be converted
There is a large amount of existing 2D content that could potentially be repurposed for 3D viewing.
If conversion can be made faster and more cost-effective without compromising visual quality, studios and platforms have an obvious incentive to revisit older libraries. This could become particularly relevant as 3D viewing develops across cinema, headsets and other display technologies.
Stereo Conversion could become more closely integrated with production
For new productions, the distinction between capture and conversion could also become less rigid.
A project may contain footage captured natively in stereo, CGI rendered with stereo information and shots that need to be converted from 2D. Having these processes operate within a connected pipeline could make it easier to maintain consistency across the final project.
Conclusion
The Lightstorm Vision and Outsyders deal is significant because it brings together two areas that have traditionally operated separately: native stereoscopic capture and post-production conversion.
The real test will be whether this combination can deliver the visual quality expected from high-end stereo productions while reducing the amount of manual work required for conversion. If it succeeds, the impact may extend beyond individual films. Faster and more consistent conversion could make it more practical for studios to create 3D versions of existing films, animation and other content that was never originally produced for stereoscopic viewing.