OctaneRender™ 2020.1 [superseded by 2020.1.1]

VIP Information, news and announcements regarding new Octane Render commercial products and releases.
Post Reply
linograndiotoy
OctaneRender Team
Posts: 1354
Joined: Thu Feb 01, 2018 7:10 pm

coilbook wrote:Hi otoy,
I know that you know that rendering animation in path tracing is still not easy for the most of us even having tens of 2080TIs. Can you please add secondary direct light bounces since in AO light looses its strength quick

Anyway to create a hybrid kernel for animations with better denoiser for moving cameras and better faster light bounce ? We were waiting for crazy rt improvement and brigade and saw no mprovements for faster rendering for animations

Did you try Direct Light in Diffuse mode?
I'm normally working using Path Tracing and I think the speed is pretty awesome. This is a quick experiment I made using my 2 RTX 2080 Tis:

https://youtu.be/khdaL-jgW5U

And this:

https://youtu.be/7zGV2ce8B6E
User avatar
Notiusweb
Licensed Customer
Posts: 1285
Joined: Mon Nov 10, 2014 4:51 am

linograndiotoy wrote:
coilbook wrote:Hi otoy,
I know that you know that rendering animation in path tracing is still not easy for the most of us even having tens of 2080TIs. Can you please add secondary direct light bounces since in AO light looses its strength quick

Anyway to create a hybrid kernel for animations with better denoiser for moving cameras and better faster light bounce ? We were waiting for crazy rt improvement and brigade and saw no mprovements for faster rendering for animations

Did you try Direct Light in Diffuse mode?
I'm normally working using Path Tracing and I think the speed is pretty awesome. This is a quick experiment I made using my 2 RTX 2080 Tis:

https://youtu.be/khdaL-jgW5U

And this:

https://youtu.be/7zGV2ce8B6E
A couple things here (this is aimed at Otoy in general, not you Lino....!)

first off - it has to be 1080p or greater. Forget mini 1/3 or 1/2 screen renders. That's not what we use to make videos or animations. EVER.
second - when you animate frames and use denoiser, you get blurry pixelations once frames compile. And this isn't even grainy noise like that Disney noise solution thing. This is totally different - with denoiser - in fact it's worse I will say.
third - when you render organics - and I'm not talking toon / stylized - I mean as close as photo-real as we can get - you deal with displacement, hair, trees, subdivision - this becomes MEGA-ZEGA slow 1080p or >
fourth - RTX is not universal...And even with 10 Titan RTXs you *still* need an RTX-optimized scene for the speeds, so RTX is essentially not there more than half the time.
fifth - do what Coilbook said...NOW
sixth - thank you!
Win 10 Pro 64, Xeon E5-2687W v2 (8x 3.40GHz), G.Skill 64 GB DDR3-2400, ASRock X79 Extreme 11
Mobo: 1 Titan RTX, 1 Titan Xp
External: 6 Titan X Pascal, 2 GTX Titan X
Plugs: Enterprise
linograndiotoy
OctaneRender Team
Posts: 1354
Joined: Thu Feb 01, 2018 7:10 pm

A couple things here (this is aimed at Otoy in general, not you Lino....!)
Discussion is always a nice thing!
fifth - do what Coilbook said...NOW
sixth - thank you!
What Coilbook is asking for (adding secondary bounces to AO mode) doesn't make much sense to me.
"Adding Secondary Bounces" means Radiosity, and radiosity means Direct Light in Diffuse Mode or Path Tracing.
Ambient Occlusion is what it is, Ambient Occlusion!
You want Ambient Occlusion with Secondary Rays? Use Direct Lighting in Diffuse mode, which is there already.
itsallgoode9
Licensed Customer
Posts: 894
Joined: Thu Apr 03, 2014 9:04 am
Location: New York City
Contact:

Lewis wrote:Hi Devs/OTOY,

I have some problems with overlapping displacement materials when one material is transparent (water) then i get black edges around sections where they meet. On bigger scene my real shoot it's even worse and from distance i get thick 5-6 pixels black border around my coast line where gravel meets sea but even on close ups like here it's very noticeable and problematic as you can see from this screen grab and i simplified scene into ORBX so you can see for yourself.

Can you investigate and fix this ?

Thanks
This is one of those fun "If somebody brings it up, we just ignore the request without explanation" things. This issue has been here since well before v2020...I can't remember a time that it's NOT been an issue.
User avatar
Goldorak
OctaneRender Team
Posts: 2321
Joined: Sun Apr 22, 2012 8:09 pm
Contact:

coilbook wrote:Hi otoy,
I know that you know that rendering animation in path tracing is still not easy for the most of us even having tens of 2080TIs. Can you please add secondary direct light bounces since in AO light looses its strength quick

Anyway to create a hybrid kernel for animations with better denoiser for moving cameras and better faster light bounce ? We were waiting for crazy rt improvement and brigade and saw no mprovements for faster rendering for animations
I don’t know if it’s going to solve your problem, but the dirt shader is basically AO, and using that to modulate albedo in DL+diffuse or PT may get what you are looking for.

Regarding future work, there is a lot in development. Brigade , PPM kernel/ IC and more. See the GTC20 talk if you want to know what we are doing on all of those fronts: https://developer.nvidia.com/gtc/2020/video/s22153
User avatar
Greggers
Licensed Customer
Posts: 149
Joined: Wed Feb 15, 2017 9:51 pm

Have you been able to find what is happening with the Fog medium darkening when adjusting? The problem was reported a few days ago, apologies for the impatience.
User avatar
Notiusweb
Licensed Customer
Posts: 1285
Joined: Mon Nov 10, 2014 4:51 am

linograndiotoy wrote:
A couple things here (this is aimed at Otoy in general, not you Lino....!)
Discussion is always a nice thing!
fifth - do what Coilbook said...NOW
sixth - thank you!
What Coilbook is asking for (adding secondary bounces to AO mode) doesn't make much sense to me.
"Adding Secondary Bounces" means Radiosity, and radiosity means Direct Light in Diffuse Mode or Path Tracing.
Ambient Occlusion is what it is, Ambient Occlusion!
You want Ambient Occlusion with Secondary Rays? Use Direct Lighting in Diffuse mode, which is there already.
This was the part I was talking to:
"Anyway to create a hybrid kernel for animations with better denoiser for moving cameras and better faster light bounce ? We were waiting for crazy rt improvement and brigade and saw no mprovements for faster rendering for animations"
Win 10 Pro 64, Xeon E5-2687W v2 (8x 3.40GHz), G.Skill 64 GB DDR3-2400, ASRock X79 Extreme 11
Mobo: 1 Titan RTX, 1 Titan Xp
External: 6 Titan X Pascal, 2 GTX Titan X
Plugs: Enterprise
User avatar
wallace
OctaneRender Team
Posts: 188
Joined: Sun Dec 18, 2016 10:38 pm

linograndiotoy wrote:I think we may have some issue with Mediums when used with meshes to simulate fog.

I've made a simple plane with a Volume Medium applied to simulate this effect:
Fog_01.PNG
Lowering down the Scatter value is making the whole scene darker (Absorption is set to black, so there's none):
Fog_02.PNG
I made a video:
https://youtu.be/f-lY35im4Co

That's the scene:
Plane_Fog.orbx
Greggers wrote:Have you been able to find what is happening with the Fog medium darkening when adjusting? The problem was reported a few days ago, apologies for the impatience.
OK, I am assuming you are referring to the above post from Lino.

Basically the issue with this is that the step length is now being used for medium (previously it was just volume), and because it's set to really low (4.0) along with low scatter depth (8), so the scattering rays are often terminated before reaching the environment.

Increasing the step length would resolve this issue, and it makes sense to do so in this case because you are essentially using this as an environment medium.

Out of curiosity, why is this fog plane a preferred approach over the environment medium scatter node?
User avatar
grimm
Licensed Customer
Posts: 1332
Joined: Wed Jan 27, 2010 8:11 pm
Location: Spokane, Washington, USA

Hi Wallace,

The reason I like the plane method better is that it's so much easier to control how the fog looks. I spent about an hour just playing with the environment medium and could not get the same results that the plane method gets. In the environment medium, as far as I can tell, you can't control where the fog starts or where it reaches full density. With the plane you can place it where you want the fog to start and adjust the density for when the fog saturates. I might be setting it up incorrectly for the environment node?

The plane...
fog-plane.png
Environment node...
fog-environment.png
And, at least with the Blender plugin, changing the scene units (to something like kilometers), does nothing.

Jason
You do not have the required permissions to view the files attached to this post.
Linux Mint 21.3 x64 | Nvidia GTX 980 4GB (displays) RTX 2070 8GB| Intel I7 5820K 3.8 Ghz | 32Gb Memory | Nvidia Driver 535.171
linograndiotoy
OctaneRender Team
Posts: 1354
Joined: Thu Feb 01, 2018 7:10 pm

wallace wrote:
OK, I am assuming you are referring to the above post from Lino.

Basically the issue with this is that the step length is now being used for medium (previously it was just volume), and because it's set to really low (4.0) along with low scatter depth (8), so the scattering rays are often terminated before reaching the environment.

Increasing the step length would resolve this issue, and it makes sense to do so in this case because you are essentially using this as an environment medium.

Out of curiosity, why is this fog plane a preferred approach over the environment medium scatter node?
Thank you for your explanation!
I personally tend to use the environment medium scatter node in my scene, but many users find using a plane (or a box, or a sphere) to be a way to get more predictable and controllable results. And it is, especially if you need to get a well define look whatever the scale of your scene is.
Post Reply

Return to “Commercial Product News & Releases (Download here)”