That is how I understand it as well. However, this might only be for existing V3 users, I’m not sure. I’m very excited about this new Brigade integration, because with the latest adaptive sampling, I’m already cutting my interior render times by 2/3rds.
Yes for me too. The scene speed loading coupled with the AI lights and denoiser speed – I mean WOW!!!
I’m already getting way faster renderings on the latest version with the adaptive sampling – that’s it’s been a huge help. The only slowdown now with Octane (for us) are HUGE outdoor scenes, and interiors. Looks like these updates may correct those slow downs.
And the scene load speed up will be awesome for animations too. Maybe we won’t have to export to the standalone for large scenes anymore.
The pricing is a bit strange to me . . . but no big deal. I have 4 1080’s and seeing the speed on 2 GPU’s I’m thinking what I have might be overkill now . . .
Same here, I agree that with 2 is already really fast, but it takes like a good 60 seconds to get up to 3000 samples in our scenes. So 30 seconds with 4 is better
In the past it was a problem that the geometry needs to sent from Rhino to Octane and the geometry wasn’t moved at the Octane side only. Will this be solved?
That’s a good question. I’ll see if Paul can respond to this one. I believe that Octane V4 has some vertex animation additions - if the Octane for Rhino plugin can read that data is the question. Also, I know that the Rhino V6 RDK has had some major changes that supposedly open up some doors with plug ins. It’s always been my hope that Rhino won’t have to reload a scene for every frame like it currently does. I usually end up exporting the Bongo Animations to the Octane Stand Alone package to avoid the issue you mention.
However . . . if the scene load time is cut down this much in V4 . . . it’s possible that reloading a scene might not be nearly as huge of a deal.
I was hoping that the new Rhino V6 could use Octane’s Render window to process animations, then no reload is necessary, as even with the V3 build proxies are now transformed live and any geometry (or all geometry if you wish) can be transformed live with no re loading. But . . . that requires the Octane render window to be open and Bongo always pulls up the Rhino Render Window. This may be more of a Bongo issue than an Octane issue, I’m not sure.
In any case, being able to create interior renderings in just a few seconds in Octane V4 is HUGE!!!
I’m very curious to see it. For me Octane was my product design animation tool, since there was no expensive GI needed and so it was faster than Vray. (scenes with a lot reflections like from metalls)
I am curious as well. I’m really hoping it’s all everyone has been expecting it to be. Otoy has been showing clips of Brigade for quite some time so the anticipation is high since they announced it would be integrated into Octane. If scenes are truly rendered as fast as they are showing with Pathtracing and not losing quality . . . I think the Otoy will turn a lot of heads.
In the past it was a problem that the geometry needs to sent from Rhino to Octane and the geometry wasn’t moved at the Octane side only. Will this be solved?
I am not completely sure what you mean by this. However with Rhino6 the new Render Queue provides the ability to determine exactly which geometry has changed each frame, so that should provide additional benefits to users once I incorporate it into the plugin. I think long-term the idea of having “Non-Live” and “Live” geometry will probably be removed, and all geometry will be “Live”, and that way, when you add geometry to the Rhino scene, it will automatically be added to the Octane Viewport.
HI guys, I already tested Octane 4 - experimental build.
Besides many improvements under the hood - the only speed gainer at the moment is the AI denoiser (and AI Light which I didn’t test)
But denoiser is not bulletproof - at least not yet.
First, it doubles the amount of vram used - so if your scene uses 800MB - with denoiser it is 1.6GB
It’s is not 100% perfect solution - some artifacts are still visible in some cases - especially when specular material is close to glossy or diffuse surface.
It also slows down render speed a little - like 1 - 2 %.
Overall it is a speed gain, but not always, and it is not some magic 2h to 2m render time trick. But it is experimental build - things will change I presume.
And free Octane4 license for 2GPUs is only for upgrading from earlier versions - that’s how I understood the video - but now I’m not sure.
Thanks for sharing your insight on the tests you’ve done. I hate to hear that it doubles the amount of VRAM, I see a lot of people having big issues with that, but maybe the Otoy team will be able to resolve that issue in the future.
I’ve never seen a denoiser that works perfectly in all cases, so I would have been surprised if Otoy’s did.
The AI light is of great interest to me because the large number of lights is what often makes interior renders take so long to clear. I’m eager to test out V4 but will probably wait until the plugin for Rhino is released.
Thanks again for your feedback, it’s great to hear. If you do any more interesting tests please let us know!