Andrey Lebrov Building Hollywood-Scale Pipelines with Render Network
Imagine rendering a 15 kilometer forest scene with millions of animated instances directly from your MacBook in 37 minutes without crashing your system. That’s not a dream anymore. It’s the real-world workflow of filmmaker, VFX trailblazer, and the Render Network power user Andrey Lebrov.
In this week’s X Spaces conversation hosted by the Render Network Foundation, Andrey pulled back the curtain on his latest production pipeline, one that brings together Octane Standalone, Houdini, and Render Network into a workflow that feels like a huge breakthrough for 3D creators.
Andrey shared his inspiring journey from running a studio using Cinema 4D and Octane for beauty shots and product commercials to selling the studio in 2021 to focus on movies, immersive museum experiences, and 20K resolution theme park content.
He was looking for more scale and control over things that traditional DCCs like Cinema 4D or Maya struggle with at massive resolutions. That search led him to Houdini — the go-to for complex procedural 3D work.
But Houdini alone wasn’t enough. He returned to Octane but this time, Octane Standalone not as a plugin renderer, but as a scene compiler — where asset assembly, lookdev, and render execution are modular and decoupled.
His workflow borrows from the same philosophy behind tools like the now-deprecated Clarisse (used in Avengers: Endgame and Blade Runner 2049), which separate asset assembly from rendering to reduce load on any single DCC. Using Octane Standalone and the Render Network, Andrey built a scalable workflow for delivering production-level scenes across GPU-distributed infrastructure.
To bring it all together, Andrey and his team developed a toolkit designed for scalable rendering on the Render Network called LMI Houdini Tools (fun fact: LMI stands for Lebrov Motion Imagery, his studio’s name). These tools:
- Tag and modularize Houdini scenes into manageable components
- Compress Houdini projects into smaller ORBX files that are easier to manage
- Split sequences into 10–25 frame chunks, enabling frame-level batching
- Support differential uploads, where unchanged assets aren’t re-uploaded — saving time, bandwidth, and cost on large jobs
These tools make rendering massive projects faster, cheaper, and more seamless.
Using these tools, Andrey and his team rendered a shot that would’ve taken 275GB and split it into 30GB chunks to make it easier to upload to the Render Network. Thanks to the differential uploader, which only sends what has changed, the final scene update required just 1.5GB to upload.
Andrey walks through the full Houdini to Render Network workflow, showing how these tools make baking huge scenes into ORBX simple and how teams can now render at scale with way less friction.
During an interview with Jules Urbach (founder of Render Network) covering neural rendering and more, he described Andrey’s technical innovation as groundbreaking.
Jules: Standalone is how I use Octane myself. Pure GPU power, no friction. What Andrey’s doing is what I dreamed of when we built the Render Network.
Andrey: This workflow lets artists go from product commercials to cinema-scale storytelling. You don’t need to be a big studio you just need the right tools.
Whether you’re a solo artist on a Macbook or a studio building cinematic worlds, this breakthrough highlights what’s possible for scaling up your renderings, keeping costs low, and staying in control while working across tools like Blender, Cinema 4D, Houdini, and Octane.
The LMI Houdini Tools: Unlocking Octane & Render Network Capabilities Knowledge Base tutorial Q&A can also help you learn more about the toolkit.
If you want to experiment with it when it launches, the Render Network Foundation encourages you to apply for a grant like Andrey Lebrov did at renderfoundation.com/grants and stay in connected with the Render Network community.
