Thanks for testing. At the moment you can only attach a static color to the absorption and scattering pins. We did this a long time ago for performance reasons. We want to allow normal texture evaluation, and we will look into it, but we can't guarantee that it will be fast enough to make the change. If it is significantly slower, then doing full texture evaluations will slow down scenes where you only need a static color instead of a texture. (In other words, if we do this change, it is possible Octane will become slower in general when there are volumes or media present).Hesekiel2517 wrote:Hey there,
2018.1 XB2 works really well. Once all GPU failed after moving the camera. But i couldn't recreate that. One thing i noticed is that if a noise is connected to the scatteringpin of the volumemedium node. The whole Volume turns red. Maybe this is not intended. Is there another way to add more detail to vdb volumes?
Thanks in advance
OctaneRender™ 2018.1 XB2
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.
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.
After significant changes in Octane we always redo our performance tuning, and this is still on our todo list.Zay wrote:On average it's about 5-6% slower than v4.0. This is mostly indoor scenes V4 is faster. Any reason why this might be the case?
I noticed just playing with the Volumetric lighting sample file, that it looks great from an angle, if the camera is in the cone, it is not scene, and if the camera is outside the cone but looking directly down the cone it gives some weird light issues.
______________
Nate Mac - Chicago, IL
Cinema 4D R26.107 | Octane 2021.1.6
Main: AMD ThreadRipper | Win10 | 32-core | 128GB | RTX A5000 & GTX 2070 SUPER
Slaves: 4x GTX 980 Ti | GTX 1080 | GTX 1070 | GTX 1070Ti
Nate Mac - Chicago, IL
Cinema 4D R26.107 | Octane 2021.1.6
Main: AMD ThreadRipper | Win10 | 32-core | 128GB | RTX A5000 & GTX 2070 SUPER
Slaves: 4x GTX 980 Ti | GTX 1080 | GTX 1070 | GTX 1070Ti
- PolderAnimation
- Posts: 375
- Joined: Mon Oct 10, 2011 10:23 am
- Location: Netherlands
- Contact:
Could the padding size in the bake camera be unlocked to 100 or 200 of something like that? Would help allot thanks.
Win 10 64bit | RTX 3090 | i9 7960X | 64GB
please add an option to lock dirt map after frame 1. That way any approaching objects wont' leave dirt marks like a car tire moving forward leaving dirt mark because of the close proximity. We asked many time for dirt map to be locked on SUB-OBJECT LEVEL so external objects have no influence
Hmmm....when working with Iray in Daz Studio, it seems it has an edge over Octane in making things look more photo-real.
It has a smoother look whereas, Octane has a slight grainier, but Toonish/flat look.
Now, I know it is up to each artist to achieve their vision of what they want, but there is something mechanically different in the feature set of Iray that creates this, and I think I can articulate what it is now, and make a recommendation for the developmental builds -
Iray has the following in Tonemapper, which help make the image's finer details 'pop' post-render:
1- Burn highlights - (adjusts only lighter colors of the spectrum up or down, independent of the darker ones), and slider can be driven + and - (2 directions)
2- Crush Blacks - (adjusts only the darker colors of the spectrum up or down, almost like a shadow increaser/decreaser, independent of the lighter ones), and slider can be driven + and - (2 directions)
Iray = Total of 4 directions
Now, the thing is is that these are then independent of Exposure and Gamma, which affect every pixel en masse.... so with Iray, you get more of a visual 'Parametric-EQ' to work with.
Almost like an LUT, but more independently finely controlled at the same time.
To compare, Octane's Imager has the following:
1 - Highlight Compression - reduces burned highlights, which in effect is like sliding the Iray Burn Highlights slider negative. But, Octane's Highlight Compression cannot itself be driven negative, so it gives you only 1 direction of control.
And, not just that, it darkens the darker pixels too, so it is not really giving you what Iray gives even for the 1 direction Octane does give of highlight control.
SO....to summarize, in Octane you have 1 of 4 directions when it comes to independent controls of highlights and darkening pixels.
In Iray, you have 4 of 4 directions when it comes to independent controls of highlights and darkening pixels.
IDEA:
In Octane's Imager node, create new
(1) Highlight slider - make it linked only to pixels on the lighter end of spectrum, and let the user slide it + and -, to adjust the lighter spectrum of pixels in the image
(2) Shadow Slider - make it linked only to pixels on the darker end of spectrum, and let the user slide it + and -, to adjust the darker spectrum of pixels in the image
This will give us the Iray look if we want it, just by adding in sliders!
Because, Iray can, with its Tonemapper sliders, give a render, having even level of gamma, the pop of light and dark pixels.
This nets to a softer image between bright and dark details that pop.
BTW - Iray is slow as F%^$....
THX!
It has a smoother look whereas, Octane has a slight grainier, but Toonish/flat look.
Now, I know it is up to each artist to achieve their vision of what they want, but there is something mechanically different in the feature set of Iray that creates this, and I think I can articulate what it is now, and make a recommendation for the developmental builds -
Iray has the following in Tonemapper, which help make the image's finer details 'pop' post-render:
1- Burn highlights - (adjusts only lighter colors of the spectrum up or down, independent of the darker ones), and slider can be driven + and - (2 directions)
2- Crush Blacks - (adjusts only the darker colors of the spectrum up or down, almost like a shadow increaser/decreaser, independent of the lighter ones), and slider can be driven + and - (2 directions)
Iray = Total of 4 directions
Now, the thing is is that these are then independent of Exposure and Gamma, which affect every pixel en masse.... so with Iray, you get more of a visual 'Parametric-EQ' to work with.
Almost like an LUT, but more independently finely controlled at the same time.
To compare, Octane's Imager has the following:
1 - Highlight Compression - reduces burned highlights, which in effect is like sliding the Iray Burn Highlights slider negative. But, Octane's Highlight Compression cannot itself be driven negative, so it gives you only 1 direction of control.
And, not just that, it darkens the darker pixels too, so it is not really giving you what Iray gives even for the 1 direction Octane does give of highlight control.
SO....to summarize, in Octane you have 1 of 4 directions when it comes to independent controls of highlights and darkening pixels.
In Iray, you have 4 of 4 directions when it comes to independent controls of highlights and darkening pixels.
IDEA:
In Octane's Imager node, create new
(1) Highlight slider - make it linked only to pixels on the lighter end of spectrum, and let the user slide it + and -, to adjust the lighter spectrum of pixels in the image
(2) Shadow Slider - make it linked only to pixels on the darker end of spectrum, and let the user slide it + and -, to adjust the darker spectrum of pixels in the image
This will give us the Iray look if we want it, just by adding in sliders!
Because, Iray can, with its Tonemapper sliders, give a render, having even level of gamma, the pop of light and dark pixels.
This nets to a softer image between bright and dark details that pop.
BTW - Iray is slow as F%^$....
THX!
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
Mobo: 1 Titan RTX, 1 Titan Xp
External: 6 Titan X Pascal, 2 GTX Titan X
Plugs: Enterprise
Nice release!
Do you have a sample scene regarding vertex color attribute?
I've made a test model (fbx) with Quill and imported it in Maya where I get the colors using the "_VertexColor" attribute.
But in Octane I can't manage to make it work :/
Thanks!
Edit: It's also working with blender if I read the Attribute "Col" for some reason..
But in Octane it still doesn't work if I use this name
Do you have a sample scene regarding vertex color attribute?
I've made a test model (fbx) with Quill and imported it in Maya where I get the colors using the "_VertexColor" attribute.
But in Octane I can't manage to make it work :/
Thanks!
Edit: It's also working with blender if I read the Attribute "Col" for some reason..
But in Octane it still doesn't work if I use this name
Last edited by Renart on Sun Dec 23, 2018 12:08 am, edited 1 time in total.
Two findings with this 2018.XB2 build
1 - if you alter scene while denoiser is active, it will show serious spikes in Delayed Procedure Calls (for those of us with USB 3.0 PCIE connected GPUs). If you have Denoiser off, spikes are not nearly as significant.
So, you run the risk of a crash if you tweak too much while denoiser is on, but if you turn off and tweak, no issue. So, what I do is only use denoiser at the end when I'm rendering, and no longer fiddling with anything in the scene.
2 - I found that if texture path and name exceeds a certain # of characters, when you try to load that texture into a material, it will crash Octane. If I then rename the material to a 1 or 2 character file name, it loads fine.
(I didn't yet identify what that character count is yet, but I am using Win 10.)
But, then also, if I bring the original full-named texture file to a higher level folder, it then also loads fine.
So it seems Octane doesn't like materials having texture names & paths that extend out too far with too many characters.
1 - if you alter scene while denoiser is active, it will show serious spikes in Delayed Procedure Calls (for those of us with USB 3.0 PCIE connected GPUs). If you have Denoiser off, spikes are not nearly as significant.
So, you run the risk of a crash if you tweak too much while denoiser is on, but if you turn off and tweak, no issue. So, what I do is only use denoiser at the end when I'm rendering, and no longer fiddling with anything in the scene.
2 - I found that if texture path and name exceeds a certain # of characters, when you try to load that texture into a material, it will crash Octane. If I then rename the material to a 1 or 2 character file name, it loads fine.
(I didn't yet identify what that character count is yet, but I am using Win 10.)
But, then also, if I bring the original full-named texture file to a higher level folder, it then also loads fine.
So it seems Octane doesn't like materials having texture names & paths that extend out too far with too many characters.
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
Mobo: 1 Titan RTX, 1 Titan Xp
External: 6 Titan X Pascal, 2 GTX Titan X
Plugs: Enterprise
14000 samples later in AO. Just noise. 20 x 1080ti - 5 min to render. Why do cpus calculate cleanly and gpus just cannot. I know CPUs are slower but they do the job.
Maybe GPUs just cannot handle a big varieties of colors, lights and objects at the same time. GPUs are very slow at rendering stuff like grass but fast at one uniformed color like sky.
And denoiser creates blotchy patches out of that blue noisy interior ceiling.
Can we get a bucket mode to do clean final renderings?
Maybe GPUs just cannot handle a big varieties of colors, lights and objects at the same time. GPUs are very slow at rendering stuff like grass but fast at one uniformed color like sky.
And denoiser creates blotchy patches out of that blue noisy interior ceiling.
Can we get a bucket mode to do clean final renderings?
In principle we could do it, but it would be quite slow to calculate. Why would you need such a large padding?PolderAnimation wrote:Could the padding size in the bake camera be unlocked to 100 or 200 of something like that? Would help allot thanks.
In theory there is no difference between theory and practice. In practice there is. - Yogi Berra