Hello World: OctaneRender 4 is here
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.
- ff7darkcloud
- Posts: 73
- Joined: Sun Mar 18, 2012 11:20 pm
- Location: Fairfax, VA, USA
Will V4 XB2 be able to use network rendering?
( PC ) Win10 x64 | 2xRTX2080TI | ASUS Rampage VI Extreme | i9-7900x | 32GB | 1600W
( Laptop ) Win10 x64 | RTX2080Max-Q | Lenovo Legion Y740 | i7-9750H | 32GB - Octane Standalone Win 10/ Sketchup 2019
PORTFOLIO: https://renderspace.wixsite.com/portfolio
( Laptop ) Win10 x64 | RTX2080Max-Q | Lenovo Legion Y740 | i7-9750H | 32GB - Octane Standalone Win 10/ Sketchup 2019
PORTFOLIO: https://renderspace.wixsite.com/portfolio
Do I get to use network rendering with a single license? I know v4 licensing changed but couldn't find any info for this.haze wrote:Not yet, we are stabilising all the other features for the moment, and will get onto it soon.ff7darkcloud wrote:Will V4 XB2 be able to use network rendering?
The best would be for users to be able to train Octane AI on small pieces of an image and then use the denoiser to work on the image.vijay_thirukonda wrote:Currently, I'm collecting scenes where denoiser not able to denoise properly. I would like to test the scene and add it to training set if needed. Could you please send me the scene?.enricocerica wrote:I got this black splotch in an area of my denoised image after 500 samples. The problem doesn't occure if I use let's say only 50 samples.
Strangely using another camera for a different angle of view, I got the exactly same splotch pattern on a different area.
Edit : It occurs even with 50 or 20 samples
I find universally that when AI Denoising is active, the actual render runs slower, compared to when it is not active.
Is that because it is a separate render pass being computed?
Like, if I have the image completing in normal V4 (with Denoiser unchecked) in 10 seconds, for 100 samples at 3840x2160, and then I check AI Denoising on, and set it to have the Denoiser run at completion, the Denoise-active run takes 26 seconds, followed by a 5 second Denoise processing hold. So, the 30 second Denoised result looks better in the end, but why doesn't the render to 100 samples take 10 seconds in both instances?
If the actual render speed of the Denoise-run would match that of a non-Denoised run, this would be killer!
Is that because it is a separate render pass being computed?
Like, if I have the image completing in normal V4 (with Denoiser unchecked) in 10 seconds, for 100 samples at 3840x2160, and then I check AI Denoising on, and set it to have the Denoiser run at completion, the Denoise-active run takes 26 seconds, followed by a 5 second Denoise processing hold. So, the 30 second Denoised result looks better in the end, but why doesn't the render to 100 samples take 10 seconds in both instances?
If the actual render speed of the Denoise-run would match that of a non-Denoised run, this would be killer!
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
A lot of background data are collected for denoising. we do have plans for reducing it, by training model multiple models with lesser data. But first, our focus will be on stability, volumes, memory usage., etc.Notiusweb wrote:I find universally that when AI Denoising is active, the actual render runs slower, compared to when it is not active.
Is that because it is a separate render pass being computed?
Like, if I have the image completing in normal V4 (with Denoiser unchecked) in 10 seconds, for 100 samples at 3840x2160, and then I check AI Denoising on, and set it to have the Denoiser run at completion, the Denoise-active run takes 26 seconds, followed by a 5 second Denoise processing hold. So, the 30 second Denoised result looks better in the end, but why doesn't the render to 100 samples take 10 seconds in both instances?
If the actual render speed of the Denoise-run would match that of a non-Denoised run, this would be killer!
Thanks for the feedback. The denoiser is really fabulous, I very much enjoy using the blend feature as well. It becomes a practice in finding the sweet spot and avoiding diminishing returnsvijay_thirukonda wrote: A lot of background data are collected for denoising. we do have plans for reducing it, by training model multiple models with lesser data. But first, our focus will be on stability, volumes, memory usage., etc.
For example, I can get the render I discussed up to 800 samples in 30 seconds, vs 100 samples w/denoiser in 30 seconds. So a lot of times it winds up being a wash, or even a loss in certain circumstances.
Now, if Octane V4 could somehow cache some of the background data into memory at the beginning of a scene load somehow, and then use it, like a baking process, for the denoiser, that would be pretty cool.
But, okay looking forward to the next releases.
Appreciate it!
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
AI light works with textured emitters and black body emitters, in fact that's the main use case.ramone163 wrote:Will AI light work with Black body emission material? As I understand, now it is not working.
Can you elaborate on how it's not working? Is it just crashing for you as in your previous post?