Is there a config.ocio file we could download for 1.3 version and using OCIO 2 builtin features? I can’t find it here or on github repositories. I’ve seen is was available on latest Nuke and Maya release but some names are different and i would prefer to use the same config to match things easily.
Thanks!
Hello, you can find them here :
Oh, thank you a lot.
In the previous OCIO v1 ACES config, Output - sRGB/rec709 could be used as input transform to reverse tone mapped image. Is that no longer supported on this latest config? it was useful to view movie/jpeg while keep working on ACES context on movie player and use exposure compensation before RRT. Is there still a way to do that here?
Using sRGB - Texture with Un-tone-mapped view seems to be the way to go, but then using exposure compensation/fast grade is not working as expected.
OCIOv2 explicitly separates scene and display spaces, whereas in OCIOv1 they were all “just colour spaces.”
So to apply an inverse display transform you uses an OCIODisplayTransform with “inverse” selected, rather than OCIOColorSpace.
Some people have built configs with display spaces defined in the same way as the scene spaces, in order to mimic the behaviour of a v1 config. But this is not the way OCIOv2 is intended to be used.
Ok. Thank you for your answer. I will make some test with that in mind with some movie player we use (RV, Blender) and adapt the ocio file if needed.
I have one more question with this new config: what is the equivalent of the Output - DCDM that existed previously?
Sorry to ask again, but does someone know about the old Output - DCDM colorspace which is now not available on the latest aces 1.3 ocio config’s view transforms. Should I open a new topic for that?
I need it to integrate it into DCP.
I suppose the view transform has to be P3-DCI - Display device, but then, i don’t see XYZ view transform or something that matches render of older Output - DCDM.
The committee decided not to include the DCDM Output based on the recommendation of @sdyer. Here’s a relevant discussion as to why:
Thank you to have redirected me to this topic.
I’m trying to convert P3 output to XYZ with Nuke then, but don’t have the same result than what Output - DCDM did display.
What is wrong with this setup?
Is your EXR in ap0 or ap1? You should set the read input transform to that.
It’s AP0, I tried in OCIODisplay node to use ACES2065-1 as input and this is still not the same than old DCDM.
I’ve also compared with latest Davinci Resolve with ACES 1.3 setup and they keep DCDM output that has also the same result than old OCIO config.
You have the read node set to RAW. Try changing it to ACES2065-1. This will convert your AP0 file into Nuke’s working space of ACEScg.
Yes this what i was also trying.
But it’s not not matching DCDM.
(sorry for the pixelate render, not sure to be allowed to show more)
Here is sRGB with RRT:
Here is what nuke setup in the upper screenshot render:
And here is what old Output - DCDM from OCIO config of ACES 1.1 or Davinci with ACES 1.3 render and which is probably the good one I can’t reproduce with latest OCIO config:
Looking at the VWG meeting notes from 2020-12-01 there is this exchange:
- Kevin: Some controversy about DCDM ODT. - Doug: For people needing to make DCDM, isn't this there best option? - Kevin: Rely on downstream conversion to make tiff files. Rarely done in same application grading is done in. - Nick: Was ACES central discussion about this recently. Scott said recommended approach was P3 master and make DCDM from that. - Doug: Recommended in the sense that you get what you saw if grading on P3, but some people actually setup projectors in X'Y'Z'. - Kevin: And use native projector gamut as well. Just encoding XYZ and don't go near ACES configs. Is there intersection of someone who does this and who is using ACES as ACES?
I believe the ACES discussion that @nick refers to here is the one I linked to above.
Then on 21-01-19 it is said that there is a “Better way to do DCDM than the transforms.” and the committee agrees to omit DCDM.
Hopefully one of those folks can elaborate on what that better way is here. There was discussion of put this on the repo, but I’m not able to find it.
Thank you for finding it. I would really be curious of what’s the way to do it then.
We have XYZ projector in our studio and we feed it with DCP in DCDM for reviews, and that worked very well so if I want to use only the ACES 1.3 config, I’m a bit lost on how to do it right now.
For the moment, i will edit the config with the addition of the old DCDM colorspace use in the view transform. That would be enough even if we then lose benefit of increased precision that bring OCIO v2 with ACES transforms.
Any help on workflow to correctly manage RRT to XYZ are welcome!
Any other thoughts on this topic? the workaround using old DCDM !<ColorSpace> from older OCIO config into 1.3 one is actually not giving the exact same result. I still struggle get a perfect XYZ ready colors directly from OCIO 2.1 config.
Since OCIO config for ACES 2.0 is out, I saw that a DCDM-D65 display exists in the reference config. Can I use this for DCP JPEG2000 images ?
I thought DCDM was suppose to be D60, so I’m a but confused here.
Re-encoding a hero grade to DCDM isn’t really an ACES-specific operation. That’s why it doesn’t make sense for ACES to provide a dedicated transform for it. That said, it’s possible to create a transform to add to your chain to reproduce the same math, if we know exactly what color space you’re starting from and what you’re trying to convert to.
Right now it’s a bit difficult to understand your workflow. Some of the steps you described likely wouldn’t produce a technically correct result. However, that doesn’t necessarily mean they didn’t produce a picture you liked. If this is the case, then me simply explaining the “correct” method may not actually solve the problem you’re dealing with. So the first step is to clarify what you’re doing now and what outcome you’re expecting.
A few questions that would help - answer to the best of your ability:
Once those pieces are clear, it should be possible to suggest a transforms that reproduces the behavior you want.
