3D Gaussian Splatting

[TomG]

Umm, it's not ignored - we've mentioned repeatedly it is going to come to Corona, and that the decision is whether that's 14 later this year, or 15, depending on what we map out once we sit down after the release of 13. All that is said in this very thread, in fact! So it is right there at the top of the chain of things we are considering :)

Do use the AHA portal to vote, of course. In particular there is a bit of a contradictory message - "Stop just adding what V-Ray just added, and innovate!" vs "V-Ray just added Gaussian Splat rendering, add it!". So the voting on the portal will sort out which of those takes real priority for the users :)

[TomG]

It's 5th on the list so far, if you remove clusters and toon shading which are due in 13:
https://chaoscorona.ideas.aha.io/?sort=popular

[tolgahan]

Thanks for the reply Tom — fully respect that you and the team are tracking this, and that it’s in the AHA roadmap.

That said, I want to clarify one key thing:
This isn’t about “adding what V-Ray added.”
It’s about not missing what real-time, capture-based workflows are becoming for us in 2025.

Gaussian Splats aren't just another rendering trend — they're a new layer of creative input.
For many of us working in architectural and environmental visualization, the ability to bring scanned or AI-generated splats into a physically-based renderer like Corona would be a gamechanger.

Octane and others have already opened this door. Corona has a chance to do it with better lighting realism and artistic control.

So yes, we want innovation — and in this case, innovation also means adapting to how the craft is changing.

Appreciate all the work the team’s doing. Looking forward to seeing how you approach it!


As a fanatic corona user, I can sometimes overreact a little. :))))))))  Thank you for your understanding.

[TomG]

Keep up the enthusiasm and passion!

I personally agree about the rendering of Gaussian Splats, and see it as useful regardless of whether V-Ray has recently done it or not. But we do get that criticism leveled at us, which makes it tricky, how will users interpret it - as a fantastically useful new feature and who cares if someone else did it first it's about time Corona had it regardless, or as another err... reason? to just complain that we don't innovate any more.

It is still high in the user votes, I do think it is incredibly useful for archviz, so it is very high in the running either way :) I do encourage folks to vote over the next week or two, as it won't be long before we sit down and decide what is happening next, and we'll be pulling up those votes during that process. So if you want Gaussian Splats, go vote! Or indeed vote for whatever other features you'd like to see given priority. Influence what 14 will be :)

[tolgahan]

Keep up the enthusiasm and passion!

I personally agree about the rendering of Gaussian Splats, and see it as useful regardless of whether V-Ray has recently done it or not. But we do get that criticism leveled at us, which makes it tricky, how will users interpret it - as a fantastically useful new feature and who cares if someone else did it first it's about time Corona had it regardless, or as another err... reason? to just complain that we don't innovate any more.

It is still high in the user votes, I do think it is incredibly useful for archviz, so it is very high in the running either way :) I do encourage folks to vote over the next week or two, as it won't be long before we sit down and decide what is happening next, and we'll be pulling up those votes during that process. So if you want Gaussian Splats, go vote! Or indeed vote for whatever other features you'd like to see given priority. Influence what 14 will be :)

Totally agree Tom — and appreciate your transparency here.

I get that the "innovation vs following V-Ray" narrative can create friction, but to me, Gaussian Splats are not about competitive feature matching — they’re about embracing where visual workflows are going.

For architectural visualization especially, this isn’t just a new toy — it’s a tool for speed, iteration, and storytelling. The ability to scan a context or object on-site with a mobile phone and light it in Corona's beautiful renderer? That’s not a gimmick. That’s power.

I’ll definitely encourage my team and peers to vote on AHA. Because if this is a matter of priority, I believe the future of context-aware, capture-driven rendering deserves to be high on that list.

Thanks again for being open to the feedback — it’s this kind of conversation that keeps us all excited about Corona’s potential. 🙌

[TomG]

Being able to capture the surroundings, in 3D, with just a regular phone or drone, and have it process quickly to give a relatively light and easy to work with model, is pretty astounding really. No more reliance of specialist tech, or on software that produces billions of polys and takes hours to do so, or just relying on flat 2D imagery of a surrounding area - all that kind of thing is the bread and butter of many 3D projects :)