Hi, I read about the denoiser for AOV nodes. I have a scene where I render only Render Layer 2. The output is a "denoised beauty" and a "Black Layer Shadow" pass.
I want to denoise the Black Layer Shadow pass. How do I do that?
Denoise AOV Nodes HOW?
Moderators: ChrisHekman, aoktar
Hi,
I have posted an example here, it is about denoising Light AOVs, but the process is the same for any type of AOV:
viewtopic.php?f=30&t=83664#p433908
ciao,
Beppe
I have posted an example here, it is about denoising Light AOVs, but the process is the same for any type of AOV:
viewtopic.php?f=30&t=83664#p433908
ciao,
Beppe
- sivertbakkeng
- Posts: 12
- Joined: Wed Mar 07, 2018 12:51 pm
Since denoising render passes is always the same procedure are there any plans of implementing it as a tick box or something more simple than the current setup? Seems redundant to make the same set-up repeatedly for every scene that uses it.
Also, denoised lightpasses need a gamma correction of 0.7 to be identical to the normal light pass, is that avoidable? Am I missing something?
Also, denoised lightpasses need a gamma correction of 0.7 to be identical to the normal light pass, is that avoidable? Am I missing something?
On point observation. Are content within a project files re-usable? Via copy-paste or some form of local library saving system.sivertbakkeng wrote:Since denoising render passes is always the same procedure are there any plans of implementing it as a tick box or something more simple than the current setup? Seems redundant to make the same set-up repeatedly for every scene that uses it.
It's commonly expected for end users to save project elements as presets, templates for efficiency to be used again and save time, whenever that option is available. It is straightforward to do so in Octane Standalone, via .ocs import, .orbx or OctaneDb (local).
All AOVs and composited "beauty" render exports are and should be linearly encoded (non-linear transfer function, mistakenly known as "Gamma".sivertbakkeng wrote: Also, denoised lightpasses need a gamma correction of 0.7 to be identical to the normal light pass, is that avoidable? Am I missing something?
It is the exponent value being implicitly referred to, thus the "Gamma" is a "no such thing".
That "0.7 correction" is done in post?
Or is the difference occurring within Octane C4D?
The answer to this will help more precisely pin point the reason of this occurence and suggest how to resolve it.
Octane resources
OCTANE POSTS (URLs have changed, which will break some links but all content remains available).
OCTANE POSTS (URLs have changed, which will break some links but all content remains available).
- sivertbakkeng
- Posts: 12
- Joined: Wed Mar 07, 2018 12:51 pm
It was all within Cinema, attached screenshot shows what I did to the denoised light pass to have it match the look of the original light pass in the Live Viewer:
- I´m testing in another project now and this brightness/contrast difference doesnt seem to be the case there, not sure why it was the case in the original test project.
- I´m testing in another project now and this brightness/contrast difference doesnt seem to be the case there, not sure why it was the case in the original test project.
Creating presets is definitely on the list.
In case of a checkbox for each AOV to get denoised, I think this might happen in the core side so we could simply re-use it on our side but have to check.
About the Gamma issue, we'll investigate this, thanks!
In case of a checkbox for each AOV to get denoised, I think this might happen in the core side so we could simply re-use it on our side but have to check.
About the Gamma issue, we'll investigate this, thanks!
Hi sivertbakkeng,
please check the Camera Imager.
Compositor works in RAW without tone-mapping, and it is expected to apply it on top of the Denoised AOVs.
If this is not your case, please share a link to the scene, even via Private Message, thanks.
ciao,
Beppe
please check the Camera Imager.
Compositor works in RAW without tone-mapping, and it is expected to apply it on top of the Denoised AOVs.
If this is not your case, please share a link to the scene, even via Private Message, thanks.
ciao,
Beppe
Good initiative.sivertbakkeng wrote:I´m testing in another project now and this brightness/contrast difference doesnt seem to be the case there, not sure why it was the case in the original test project.
Tests showed no such issue.
As Beppe pointed out (regardless of the proper wording), it is likely an inappropriate Imager or conflict with C4D picture viewer.
Octane resources
OCTANE POSTS (URLs have changed, which will break some links but all content remains available).
OCTANE POSTS (URLs have changed, which will break some links but all content remains available).