is OCIO color management broken in the imager?

[TStrolia]

Perhaps I am configuring this incorrectly, however, something does seem wrong to me. I am currently using 2022.1.2-21.30 in Maya 2024, and I am attempting to viewport render. I have the color management for the 3D viewport panel set to Linear sRGB, and set my camera imager OCIO view to use AgX appearance punchy. When running the live view it seems it just gives it a 2.2 gamma curve, instead of using OCIO. Even manual changes to the camera imager seem to sometimes turn on and sometimes not. When I do get a color management process of some sort to work, once a change is made to a material or something it refreshes the viewport and breaks again. This used to work a few versions ago, ill be octane it's self was less stable in maya it seemed. Is there some other setting I need to be considering to fix this? What is especially weird about the whole thing is, Hypershade displays correct color-management. My scene simply doesn't, which is why I am curious if maybe it is a me issue here. I feel like maybe there is some other setting I am missing that would resolve this or something.

[TStrolia]

Just fixed it. It was in fact a me problem. I failed to set the render color space. Once I did that it works fine.

[TStrolia]

I was incorrect, it does visually look better but I did not manage to fix it. So, in the render view the whites blow out and the contrast is still wrong. It is subtle but still wrong. I have the viewport set to linear-srgb sso it should not be double color managing. I am not really sure what is going on. Maybe someone can help explain the correct way to color manage the view port with octane in the most recent release?

[elsksa]

Standalone user here. Unfortunately not much I can do beside mentioning that Octane is self color managed and doesn’t require the software host to handle that part, as you know. Perhaps this thread might help viewtopic.php?f=28&t=78193?

[BK]

I was incorrect, it does visually look better but I did not manage to fix it. So, in the render view the whites blow out and the contrast is still wrong. It is subtle but still wrong. I have the viewport set to linear-srgb sso it should not be double color managing. I am not really sure what is going on. Maybe someone can help explain the correct way to color manage the view port with octane in the most recent release?

Hi TStrolia,

Thanks for the post.

Yes, please always disable the host app's color management! After adding OctaneImager, OctaneOCIO view and Octane OCIOlook node.
Switch the Image format Output to **EXR** then set the OCIO color space *name* from the File Output Rollout.
You may need to Re-run the IPR.


cheers

[nospitters]

i run in the same issue, but with ACES...
-color management is turned off
-rendering space is set to acescg in the octane ocio render settings
-exr is the output format and set to acescg

the imager seems to have no effect....ocio-view and ocio-look are connected...changing views here have no effect...and rendering isnt look right...

when turning on color management again and set there the rendering space and view, then it kind of look better, but not quite right...

MAYA 2020.4, MAYA 2023.3, octane 2023.1.2

[BK]

i run in the same issue, but with ACES...
-color management is turned off
-rendering space is set to acescg in the octane ocio render settings
-exr is the output format and set to acescg

the imager seems to have no effect....ocio-view and ocio-look are connected...changing views here have no effect...and rendering isnt look right...

when turning on color management again and set there the rendering space and view, then it kind of look better, but not quite right...

MAYA 2020.4, MAYA 2023.3, octane 2023.1.2

Hi nospitters,

It looks like ACES is working correctly when we test.
Due to the native Image Maya buffer, we have to set the image format as EXR then the Octane View would work from Common tab>OCIO color space as the image below:

If the image format is PNG, the view works from Imager>view

Note, that you may need to Refresh the IPR if the image formats are changed!

cheers