Color Values between sRGB and ACES in Houdini

I have ACES setup in Houdini, although when enabling OpenColorIO, whether I disable it, or not the color values are the same between renders of the same image.

Is there a way to inspect whether the colors in the image are that of sRGB or Linear sRGB values and ACEScg (Rec2020) ?

Hi Christopher,
I don’t usually work with Houdini and you also didn’t specify the renderer you use.
But in the simple case of rendering an object with a RGB color value of 0.5/0.0/0.0 will result in a EXR render with the same color value - no matter how you set your color-pipeline.
The view result will differ, because the view transform is different and the scene linear values actually mean something different in a different colorspace.

I was writing a while back about this topic.

I hope this helps.



The view transform is that of the hardware I’m viewing the ACES profile or rendering the ACES image ? In other words, if the view transform is sRGB then the rendered image will be rendered in ACES, but will fit the view transform of sRGB ?

Hi I made you two screenshots from Nuke. I hope this explains your question.

One is the from a standard Nuke reading a EXR file with a red color 0.5/0.0/0.0 viewed through the sRGB viewer.
The other is from Nuke with OCIO, loading the same file and “interpret” them in two different ways - then view them through an ACES/Rec.709 viewer. Please see the color values in the viewer.

If your 3D software is set to OCIO, the you load the EXR to ACEScg, if your render pipeline is set to the old way (which is now called in MAYA linear sRGB/Rec709), the your load the file as linear_sRGB.

Best regards


When I render an image in Houdini with OpenColorIO enabled with ACES, then I turn off OpenColorIO, I don’t see a difference as you show with Nuke; it doesn’t make sense, I should be seeing a difference.

you still didn’t tell which renderer you use - maybe you can show also some screenshots?



It’s been a while; do you know how to accurately get what you have setup in Nuke, within Fusion ?
Then as my original question was from two years ago ?