OctaneRender™ 2019.1 XB1

Tue May 14, 2019 8:33 am

Hey, developers, anybody here?

Tue May 14, 2019 9:54 am

divasoft wrote:Hello guys. I had a problem while I rendering a lot of particles. Octane render time is constant per-frame, but frame-processing time drammaticaly increasing frame by frame. Scene consist same quantiti particles in every frame, but render time increasing for 2 times after each 10 frames. But when i restart render, render time going small and start increasing again. On screenshot you can see when same frame (114) can render 4:17 and after restart 1:01. It's a madness. (x-particles are cached)
Is it Cinema related problem?

Is this problem related to 2019.1 XB1? If not, please get in touch with Ahmet on the forum, or with our support team.

Otherwise, if it is possible to reproduce this problem in standalone, please let us know.

Tue May 14, 2019 9:56 am

Despot wrote:My 2019.1 XB1 build expires in 5 days or so, and whilst I'm obviously not using it for production work - I'm still enjoying testing out the new features.

Will there be an extension before it expires ?

Thanks :)

It is possible to extend it, but since XB2 is almost here, we'd much prefer if you could take a look at that instead... It will have heaps of fixes, and also some very interesting additions - you'll see AI Upsampling for the first time in XB2.

Tue May 14, 2019 10:13 am

haze wrote:
divasoft wrote:Hello guys. I had a problem while I rendering a lot of particles. Octane render time is constant per-frame, but frame-processing time drammaticaly increasing frame by frame. Scene consist same quantiti particles in every frame, but render time increasing for 2 times after each 10 frames. But when i restart render, render time going small and start increasing again. On screenshot you can see when same frame (114) can render 4:17 and after restart 1:01. It's a madness. (x-particles are cached)
Is it Cinema related problem?
Is this problem related to 2019.1 XB1? If not, please get in touch with Ahmet on the forum, or with our support team.

Otherwise, if it is possible to reproduce this problem in standalone, please let us know.

Sorry, but i can't reproduce this problen in standalone because octane can't work with particles as particles in standalone and convert all of them in geometry. But scene consist so many particles and export process crashes every time. I tried 5 times, unsuccessfull.

Tue May 14, 2019 7:53 pm

haze wrote:
Despot wrote:My 2019.1 XB1 build expires in 5 days or so, and whilst I'm obviously not using it for production work - I'm still enjoying testing out the new features.

Will there be an extension before it expires ?

Thanks
It is possible to extend it, but since XB2 is almost here, we'd much prefer if you could take a look at that instead... It will have heaps of fixes, and also some very interesting additions - you'll see AI Upsampling for the first time in XB2.

Hi, thanks for the reply - that all sounds cool, but my next question is will there be a gap between XB1 expiring (clock reads 3 days 5 hours for me) ? or will it be a seamless transition to XB2 ?

Thu May 16, 2019 1:20 pm

Any chance of an extension for XB1 please ? just over a day left here...

Fri May 17, 2019 12:40 am

Yes - we are extending it a bit, and XB2 is just waiting for me to review AI upsampler (now in imager node) before it goes out on the forums.

Fri May 17, 2019 9:56 pm

Goldorak wrote:Yes - we are extending it a bit, and XB2 is just waiting for me to review AI upsampler (now in imager node) before it goes out on the forums.

Thank you for doing that, it's appreciated

Mon May 20, 2019 3:35 am

please add an option to nickname GPUs.

Tue May 21, 2019 3:39 pm

Regarding my huge post about normal mapping issues, in the mean time, I've made some progress implementing normal mapping FROM SCRATCH using OSL.

Here's a video demonstrating some of the improvements to diffuse shading. I am simply using view direction to gradually reduce intensity of normal maps only in problem areas. The results are surprisingly good so far. In real world tests overall normal maps appear more "3D" (for lack of a better term) compared to what we have now using material normal map pin only.

I'm now adding more features to this OSL script and I'll be posting it here on the forums along with a bunch of other useful utility scripts for working with normal maps.

Also here's a research paper that offers a microfacet-based approach to solving this issue.
https://blogs.unity3d.com/2017/10/02/mi ... h-tracing/

Cheers
Milan