Aces vs non aces comparison

Hello guys!
A, probably dumb, question comes in my mind. While researching for a aces workflow I came across this issue.
I’ve got a simple 8bit jpg with srgb 2.1 color profile that I’d like to open in nuke in an Aces environment with srgb (aces) as viewer process.

So my question, based on the image that I attach, is: why the 2 images (left is a screenshot taken in PS while right is a screenshot take in nuke) are different? Is it normal that the same jpg imported in an aces environment differs from the same viewed in photoshop?

Sorry for this that probably is a noob question but I finding more and more holes in my knowledge.

Thank you very much!

It is the expected result that they do not look the same. There are a great many threads here on this very subject. The trouble is that you are going from a non-color managed workflow (Photoshop) into a color manged workflow.

The workflow that you are using in Nuke [ input: utility srgb texture, display sRGB (ACES) ] is the workflow for textures in a render. One would not generally want to create textures in Photoshop (which is non-color managed) and rather create them in a program that supports ACES so they can be viewed under the ACES display transform, then render them in a program that supports ACES, and comp them in a program that supports ACES. This ensures a consistent viewing environment throughout.

It is possible to have an sRGB image appear the same in Nuke as it does in Photoshop by reading it in with an inverse of the display transform. However it’s important to understand that this will result in white (1.0) mapping to a value of over 16, making it light-emitting.

Thank you Derek! it is now clear…
In an aces environment (nuke) if I use an image with an icc color profile but not made in aces I should expect some kind of difference, and it is normal.

So the IDT serves to “adapt” the image (the srgb 2.1) in the aces environment but not to have the same corrispondency of the image in the aces environment, because even if aces is set in srgb is different from the srgb of the icc… could this be, more or less, correct?

Thank you

Yes. If you did the same thing on the IDT and ODT it would look the same as it does in photoshop.

One might think that the IDT “Utility texture sRGB” and the ODT “sRGB (aces) “ are the same because they both have sRGB in the name, but they are not. The IDT is doing a simple sRGB conversion, but the ODT is doing that plus a bunch of other stuff including tonemapping.

1 Like

It might be helpful to think from the vantage of what we see. That is, the “picture” that is formed out of this soup.

sRGB defines the colourimetry of the medium something is being presented on. ACES forms a picture up out of the data, and presents it. As such, when trying to compare already formed picture stimulus to a picture that is being pushed back into the garburator, mulched up, and then formed back into a new picture, it is expected that the two pictures will be different.

In the end, no such mystical unicorn of “colour management” exists, because researchers are still trying to figure it all out.

TL;DR: It’s two wildly different pictures.


Thank you derek! much clearer now!

Troy! thank you, yes with this approach you suggested I have now a clear view of my question!
I am still full of question that I am trying to solve… one by one… thank you!

1 Like