VP Land
Posts
Oscar Winning VFX Legend Joins Stability AI

Oscar Winning VFX Legend Joins Stability AI

Sony's New Camera Tracker, NVIDIA's GTC Keynote

March 21, 2025

Powered by

Out with the NFTs, in with the AI: LG’s research arm just released their own AI model (maybe it’s how they’ll run their selfie-taking, AI-enhanced microwave/oven). They also killed their NFT marketplace.

Also in today’s issue:

VFX Legend Rob Legato joins Stability AI (we’ve also got a new interview up with Legato from the AILA Camera Control Hackathon)
Sony has their own camera sensor
GTX Keynote breakdown (for non-AI wizards)

And more - let’s get into it!

Joey

Rob Legato's Next Frontier: Stability AI

Oscar-winning visual effects supervisor Rob Legato, ASC, has joined Stability AI as Chief Pipeline Architect. This collaboration signals AI's growing role in production workflows and highlights how established industry veterans are adapting to technological change.

Legato brings decades of experience from groundbreaking films including Titanic, Avatar, and The Jungle Book, where he pioneered a variety of virtual camera and production techniques.
Stability AI, known for its Stable Diffusion model, has been rapidly developing AI tools that can generate and manipulate images based on text prompts—technology with significant implications for production workflows.
Recently Stability AI joined the Academy Software Foundation to play a key role in driving open-source innovation in visual effects and generative AI.

In our interview with Legato at VES last year, one of his main issues with AI was its lack of control and consistency, especially over the camera, which he echoed again last month and described AI’s ‘90% issue.’

Stability is already addressing that with their new Stable Virtual Camera, which transforms static 2D images into dynamic 3D videos with camera control options.

The AI model uses multi-view diffusion technology to generate videos up to 1,000 frames long from still images, allowing creators to define custom camera trajectories or choose from 14 preset paths.

Read more here about Rob Legato joining Stability AI.

SPONSOR MESSAGE

Descript is an easy to use text based video editor for podcasts and talking head videos.

But we've been using it as part of our longer video editing process with Resolve: paper edits.

Sorting through hours of interview footage and building out a radio edit is a pain.

With Descript, I edit videos as quickly as editing a document. I import my footage, get instant transcriptions, and search through hours of content in seconds.

When I hear something I like, I highlight the text and add it to my rough-cut composition. Then I build everything out, copy and paste soundbites like I would with a text document - but the video updates too.

I can also add temp VO, scene notes, and comments. The whole team can be on the same project, working on this in real-time.

When we're done, we export an XML from Descript and bring it into our NLE of choice.

Yes, most editing apps have added some transcription support, but IMO, none of them come close to the speed and ease of use in working with text in Descript.

Try Descript today.

Sony's OCELLUS: A Marker-Free Tracking System for Virtual Production

Sony has announced OCELLUS, a new camera tracking system designed for virtual production and augmented reality workflows. The marker-free system offers real-time position tracking, works with both Sony and non-Sony cameras, and integrates with software like Unreal Engine to streamline production processes.

Key features include:

Marker-free tracking technology that eliminates the need for physical markers, reducing setup time and reflections in post-production
Three-component system consisting of a sensor unit with multiple image sensors, a processing box for data handling, and lens encoders for capturing precise lens movements
Environmental versatility with dust and moisture resistance for both indoor and outdoor shooting conditions
Real-time data transmission capabilities enabling immediate integration with CG rendering software
Camera-agnostic design that works across various camera brands and models, expanding its utility across production environments

New model roundup

LG, maker of Selfie Micrwaves, has now entered the AI game with their own model EXAONE Deep (which tracks with DeepSeek).
Adobe pushed more into agentic - its AI platform is now capable of managing AI agents across multiple ecosystems, enabling businesses to create more cohesive and personalized customer experiences through unified marketing and creative efforts
xAI is ramping up its push into AI video with its Hotshot acquisition.
ReCamMaster allows media professionals to generate videos with new camera angles from a single input video.

We break down NVIDIA's key announcements at GTC—AI factories, Blackwell chips, and Project Newton—on the Denoised Podcast.

Stories, projects, and links that caught our attention from around the web:

🔵 Sony Electronics has launched the mocopi Pro Kit, a professional-grade motion capture system designed to enhance accuracy and workflow efficiency for creators in animation, gaming, VR, and live-streaming.

🃏 Village Roadshow, a prominent independent film production company behind Joker and Mad Max: Fury Road, has filed for Chapter 11 bankruptcy, citing significant financial liabilities.

🎥 Panavision closes its Hollywood office, consolidating to their Woodland Hills location. Will they make the shift to selling their cameras and lenses, versus their rent-only model?

🚔 Carl Erik Rinsch, director of 47 Ronin, has been charged with wire fraud and money laundering for allegedly swindling $11 million from Netflix intended for an unfinished sci-fi series called White Horse.