OctaneRender™ Standalone 3.08 TEST 5

A forum where development builds are posted for testing by the community.
Forum rules
NOTE: The software in this forum is not %100 reliable, they are development builds and are meant for testing by experienced octane users. If you are a new octane user, we recommend to use the current stable release from the 'Commercial Product News & Releases' forum.
User avatar
Jolbertoquini
Licensed Customer
Posts: 1067
Joined: Sun Aug 31, 2014 7:08 am
Location: London
Contact:

FrankPooleFloating wrote:Ugh.. I was so excited to have displacement working right.. don't think I will bother to download this version... :cry: I'm sure they'll get it right soon though...

Hey, I love the dino without claws attached!.. Kinda like a Ticklesaurus Rex! ;)
:lol: :lol: :lol: :lol: :lol: :lol: :lol: hahahahahaha
Octane Render for Maya.
https://vimeo.com/jocg/videos
https://www.linkedin.com/in/jocgtd
http://www.hmxmedia.com/
--------------------
Join MAYA OCTANE USERS Skype discussion here :
https://join.skype.com/LXEQaqqfN15w
User avatar
wallace
OctaneRender Team
Posts: 188
Joined: Sun Dec 18, 2016 10:38 pm

divasoft wrote:"Fixed artifacts in the Ward BRDF."
It was a bad joke.
Hi.

This artifact is not produced by the Ward BRDF itself. The fix to Ward BRDF was to get rid of the dark rims, basically to improve the fall off of the original Ward BRDF.

The artifact you are seeing is caused by un-interpolated tangents and is causing issues across all anisotropic BRDFs. Hence this fix was unrelated, and we're still working on a fix for the tangents at the geometric level, rather than the BRDF level.

Thanks,
Wallace
User avatar
abstrax
OctaneRender Team
Posts: 5506
Joined: Tue May 18, 2010 11:01 am
Location: Auckland, New Zealand

Regarding the displacement changes: This is currently work in progress and the wording in the release post wasn't very precise and I fixed it. Apologies for that.

Fundamentally, there are two ways how you can do displacement mapping: Either displace a subdivided version of the geometry during geometry compilation (which includes some trickery to make the subdivided geometry manageable) or do the displacement mapping on-the-fly during render time on non-subdivided triangles. Octane currently does the latter. This has the advantage that the memory consumption is smaller for detailed displacement maps. The disadvantages are that it's mathematically/numerically a lot more complicated and every triangle is handled independently of its surrounding triangles so getting the displaced surface water tight is fairly tricky.

We have identified at least a handful of issues in the current displacement implementation and are working on improving/fixing them right now so this work isn't finished yet by any means. Our near-term goal is to make the current system as good as possible. Of course, there is the possibility that we will not be able to bring it into a good enough shape and we will have to resort to the other approach (subdivide geometry) eventually. We will decide on that when we get to this point.

So what is meant with:
  • Displacement of connected triangles are not torn apart anymore due to different vertex normals of shared vertices, e.g. displacing a cube doesn't create cracks anymore if the displacement map is continuous.
It's basically this problem: Imagine you've got a cube with this continuous texture, for visualization purposes I placed it into the diffuse channel:
marble texture.png
To check if it's really continuous you can render the diffuse filter pass (there should be no visible edges between adjacent faces):
marble texture is continuous.png
Now if you apply this texture as displacement map you get this in version 3.08:
marble texture displaced 3.08.png
while in version 3.07 you got this:
marble texture displaced 3.07.png
There was no way to get the displaced surface hole free without using shared vertex normals.

Again, there will be further improvements and we will see how far we can bring the current displacement implementation.
In theory there is no difference between theory and practice. In practice there is. - Yogi Berra
User avatar
abstrax
OctaneRender Team
Posts: 5506
Joined: Tue May 18, 2010 11:01 am
Location: Auckland, New Zealand

Jolbertoquini wrote:hum... same here really hard to get details on displacement, I need add bump to get right. but stills the gaps.

I attach renders and scenes... Please anything wrong I tested on Redshift, Iray, Arnold same setup all good out of the box.
this is a free scene from Truong CG Artist.

Cheers
JO
Thanks a lot for the scene and feedback. Yes, the normal calculation seems to break down in this version. We don't know yet what the problem is, but it seems to be related to the triangle size in UV space, but even if I disable subdivision surfaces, the problem persists. We will investigate and fix the issue as soon as possible.

The second problem are the gaps at the UV seams. These seem to be caused by different displacement heights for adjacent triangles:
dino foot displacement map.png
Looking at the displacement map itself, it seems to consist of some detailed map and some super-blurry layer underneath that causes these offsets.
In theory there is no difference between theory and practice. In practice there is. - Yogi Berra
User avatar
Jolbertoquini
Licensed Customer
Posts: 1067
Joined: Sun Aug 31, 2014 7:08 am
Location: London
Contact:

abstrax wrote:
Jolbertoquini wrote:hum... same here really hard to get details on displacement, I need add bump to get right. but stills the gaps.

I attach renders and scenes... Please anything wrong I tested on Redshift, Iray, Arnold same setup all good out of the box.
this is a free scene from Truong CG Artist.

Cheers
JO
Thanks a lot for the scene and feedback. Yes, the normal calculation seems to break down in this version. We don't know yet what the problem is, but it seems to be related to the triangle size in UV space, but even if I disable subdivision surfaces, the problem persists. We will investigate and fix the issue as soon as possible.

The second problem are the gaps at the UV seams. These seem to be caused by different displacement heights for adjacent triangles:
dino foot displacement map.png
Looking at the displacement map itself, it seems to consist of some detailed map and some super-blurry layer underneath that causes these offsets.
Hi Abstrax,

Thanks for the quick reply I will give either tests I time this model just for test I will check more the displacement texture,
And give some extra tests I may get better texture I need see what I have on the zip file is not my sculpt it was a quick test.

But strangely is working better in others engines, I will give others checks and I will post here my progress.

Is Great to know is working better now I need test the scale now for the small scale displacements aswell.

Thanks for the points.

Cheers,

Jo
Octane Render for Maya.
https://vimeo.com/jocg/videos
https://www.linkedin.com/in/jocgtd
http://www.hmxmedia.com/
--------------------
Join MAYA OCTANE USERS Skype discussion here :
https://join.skype.com/LXEQaqqfN15w
noisyboyuk
Licensed Customer
Posts: 103
Joined: Thu Apr 21, 2016 5:28 am

Really cool to see a proper acknowledgement of the displacement map issues and that it is being investigated! I really do hope that you manage to fix it using the current method as the speed it renders displacement vs other engines, the others don't even come close! It would be amazing if you manage to keep those speeds but if it comes down to it I'd rather have it working a lot slower on the geomety level (with subdivs) and it give me water tight renders.

Cheers!
Last edited by noisyboyuk on Sat Dec 09, 2017 2:08 am, edited 1 time in total.
WIPEXPO
Licensed Customer
Posts: 29
Joined: Tue May 23, 2017 12:37 pm

octane standalone crash after adding instance color node.
leehenshall
Licensed Customer
Posts: 184
Joined: Sun Oct 25, 2015 5:44 pm
Location: England
Contact:

I'm really interested in Octane toon shading. There a lot of ways to render toon out there but some I have tested rely on single threaded CPU post processing such as freestyle for blender. This creates a huge bottleneck for animation rendering and I need something that is fast and happens in the render or at least doesn't bottleneck animation rendering if it is created via post processing.

I had a quick play around with this build last night and have a few questions.

1) I manged to get the basics working such as controlling the diffuse and spec with a toon ramp and using camera based lighting but I couldn't find where to plugin the toon direct light. What's the correct workflow for this?

2) I felt like the outlines were quite limited and didn't work as I expected them to. Will there be an improvement to the outlines functionality to allow for more of a more precise and controllable outline similar to maya toon outlines or vray toon? Also it would be nice if there could be a way to make the outline colour adopt the colour behind or somehow blend over the top rather than just being a solid colour.

3) I couldn't figure out what the roughness slider did. I expected it to effect the specular hardness but didn't notice anything changing. I only had about 25 mins to play with this build so i could have just not setup thing up properly.

4) Also wasn't the toon shading going to be implemented into octane imager? my initial impressions were that this would be the case when I saw the early test images created with it.

5) Can outlines be separated into a separate pass for compositing?

I also had a quick go at using an OSL node I would love to use in Octane.....vrayflakes OSL (https://docs.chaosgroup.com/display/OSL ... normal+map) Unfortunately I couldn't get it working. Octane really needs some advanced flake and glitter capability and this would allow that functionality for me. Any ideas how to make it work?

Excited about the upcoming features...keep up the good work!

Cheers

Lee
calus
Licensed Customer
Posts: 1308
Joined: Sat May 22, 2010 9:31 am
Location: Paris

leehenshall wrote: I also had a quick go at using an OSL node I would love to use in Octane.....vrayflakes OSL (https://docs.chaosgroup.com/display/OSL ... normal+map) Unfortunately I couldn't get it working. Octane really needs some advanced flake and glitter capability and this would allow that functionality for me. Any ideas how to make it work?
viewtopic.php?f=33&t=63905#p325922
viewtopic.php?f=33&t=63905&start=10#p326013
Pascal ANDRE
leehenshall
Licensed Customer
Posts: 184
Joined: Sun Oct 25, 2015 5:44 pm
Location: England
Contact:

Thanks Calus :)
Post Reply

Return to “Development Build Releases”