ACES ODT Candidates for ACES 2.0

[rand thompson]

Link to Github ACES 2.0 ODT Candidates
GitHub - ampas/ACES_ODT_Candidates

IN DAVINCI RESOLVE 18.0.2

Current ACES REC709 ODT

ACES 2.0 CANDIDATE A REC709 ODT

ACES 2.0 CANDIDATE B REC709 ODT

ACES 2.0 CANDIDATE C REC709 ODT

Current ACES REC709 ODT

ACES 2.0 CANDIDATE A REC709 ODT

ACES 2.0 CANDIDATE B REC709 ODT

ACES 2.0 CANDIDATE C REC709 ODT

 

[Blair S. Paulsen]

IMHO, based on the images posted, viewed on an M1 MBP, Candidate B seems like the Goldilocks version.

Cheers - #19

 

[rand thompson]

IMHO, based on the images posted, viewed on an M1 MBP, Candidate B seems like the Goldilocks version.

Cheers - #19

Blair,

I liked Candidate A better, IMHO Candidate B adds an extra bit of Saturation that A doesn't, I'd like to add my own saturation level later . Plus A reduces that somewhat "Yellow" appearance in the Skintones and in the overall image inherent with the current ACES REC709 ODT.

 

[Blair S. Paulsen]

Fair enough Rand. I just felt like the gamma characteristic in A was bit too steep and C too dim. I liked that the subtle luminance transitions in B, particularly on skin tones, looked very natural. A little saturation push doesn't bother me. That said, I'd need to see the candidates in a proper viewing scenario to judge whether it's too much.

Cheers - #19

 

[rand thompson]

Blair ,

You make an excellent Point about the luminance transition. My thing was how would that all hold up with additional contrast, and as you stated, with additional saturation. They may have additional Aces 2.0 revision candidates since these are rev009 so there may be other potential candidates coming in the future. Also the results above were with the current ACES 1.3 version, who knows how these ODTs and potential IDTs will actually work in ACES 2.0

 

[rand thompson]

I did a test with one of the ALEXA 35 "Encounter" clips with a high RED saturated light. The ACES 2.0 Candidate A, along with "ACES REFERENCE GAMUT COMPRESSION" enabled, did the best to control the "REDS" in the image, Then C and then B


Current ACES REC709 ODT

ACES 2.0 CANDIDATE A REC709 ODT

ACES 2.0 CANDIDATE B REC709 ODT

ACES 2.0 CANDIDATE C REC709 ODT

ORIGINAL

CANDIDATE A

CANDIDATE B

CANDIDATE C

 

[rand thompson]

Here's a link from ACESCENTRAL to give your input on the Three candidates and to download a stills EXR sequence to test the different candidates with.


LINK
ACES Output Transform Evaluation and Feedback - 3 Candidates - Discussions - ACESNext / VWG – Output Transforms - Community - ACESCentral





Project Settings

Some Stills from the EXR sequence with VECTORSCOPE

 

[rand thompson]

People and skin tones

Project

 

[Phil Holland]

Took a look at the candidates in my testing bed. Overall, good improvements. The curve is better. Color-wise, needs some work in a few areas, but I imagine that's something I can whip up. Hot red color rendering could be tamed a bit. Blue and violet are off, but I'll investigate as to why. The hard areas of those hues. I have a cyan/blue light going magenta/violet and that's odd. But will dig in a bit later tonight.

As ACES is used a bit across the industry might just release some ACES dedicated LUT's to tidy up some stuff. philmColor as is works within the ACES workflow, which is good, but colorimetry is different very slightly to make a difference in some cases. As well as tonal related things also in some cases.

 

[rand thompson]

Took a look at the candidates in my testing bed. Overall, good improvements. The curve is better. Color-wise, needs some work in a few areas, but I imagine that's something I can whip up. Hot red color rendering could be tamed a bit. Blue and violet are off, but I'll investigate as to why. The hard areas of those hues. I have a cyan/blue light going magenta/violet and that's odd. But will dig in a bit later tonight.

As ACES is used a bit across the industry might just release some ACES dedicated LUT's to tidy up some stuff. philmColor as is works within the ACES workflow, which is good, but colorimetry is different very slightly to make a difference in some cases. As well as tonal related things also in some cases.

Phil,

Which Candidate are you preferring out of the Three? Yeah I think on the ACESCENTRAL website they are leaning towards Candidate C with Further Modifications. Maybe lend them a hand on what you think in what areas they can improve further. And Like I said before, If you do decide to make some dedicated ACES LUTs I'll be one of your first customers.:thumbsup:.

 

[Phil Holland]

To expand a bit, Candidate A holds the most potential in terms of accurate and perceptually accurate rendering. Highlight saturation in B and C are fairly wild on higher gamut colors.

 

[rand thompson]

Here's an interesting discussion on the BLACKMAGIC DESIGN forum about the "ETA ACES 2.0?"

Blackmagic Forum • View topic - ETA ACES 2.0? (blackmagicdesign.com)

 

[rand thompson]

To expand a bit, Candidate A holds the most potential in terms of accurate and perceptually accurate rendering. Highlight saturation in B and C are fairly wild on higher gamut colors.

Yeah, I thought so as well about Candidate A. Candidate A atleast seems to have tamed the Reds, Yellows and Magentas Ranges. However Cyans and Greens still seem to need a little more work.

 

[Phil Holland]

I'll reach out to them the best I can. Likely through the ASC MITC side. But here's one example that I feel A is much more correct. B isn't too bad, but still is loud. C isn't even close.

I'll underline that color at the end of the day does get a bit subjective when looking at accuracy and perceptual accuracy. Often there's a choice that needs to be made between what's right, what's natural, and what looks good.

 

[rand thompson]

I'll reach out to them the best I can. Likely through the ASC MITC side. But here's one example that I feel A is much more correct. B isn't too bad, but still is loud. C isn't even close.

I'll underline that color at the end of the day does get a bit subjective when looking at accuracy and perceptual accuracy. Often there's a choice that needs to be made between what's right, what's natural, and what looks good.

Phil this is a great example of the three Candidates! I like A in this example as well.

 

[Phil Holland]

Phil this is a great example of the three Candidates!

They really need to investigate what's going on with cyan through magenta. Finding color artifacts on tricky through not-so-tricky gradients. Visible in their test images as well.

 

[rand thompson]

They really need to investigate what's going on with cyan through magenta. Finding color artifacts on tricky through not-so-tricky gradients. Visible in their test images as well.

I'm sure they're still working on what will be the final ODT's for ACES 2.0 so hopefully any remaining color gamut issues will be addressed . This also could be a side effect of using these ACES 2.0 candidate ODTs in ACES 1.3 as well but who knows.

 

[Blair S. Paulsen]

Still prefer B, even with the bit of over saturation in the red in Phil's car shot. C seems to knock down the lower mid-tones too quickly for my taste.

A is OK, but IMHO desaturates lighter hues a bit too much. I get that over saturation is a tougher thing to tame, but thinning out the pastels too much tends to allow the strong colors to overwhelm them - even when you don't want them to.

Cheers - #19

 

[Phil Holland]

prefer

Keeping on the car for a moment Blair, how's this?

You can open that in a new tab for a closer look. Didn't include C because it's just too out of whack. Only providing two of my options here because I provide a few things. This actually is some new stuff that I'd be curious to get feedback on. I made a custom highlight roll-off to the ACES curve to somewhat get the car and cone to render better.

I've now had a full day to look at everything on the ACES side. It to me is still a work in progress and I still can't recommend the previous workflow due to color artifacts. Hue related ones.

I can't stress enough. RED, Graeme really, put a great deal of time into doing this pretty right.

 

[rand thompson]

Still prefer B, even with the bit of over saturation in the red in Phil's car shot. C seems to knock down the lower mid-tones too quickly for my taste.

A is OK, but IMHO desaturates lighter hues a bit too much. I get that over saturation is a tougher thing to tame, but thinning out the pastels too much tends to allow the strong colors to overwhelm them - even when you don't want them to.

Cheers - #19

Blair,

Who knows what ACES 2.0 ODTs and IDTs may actually be included by the ACES Virtual Working Group but they have solved alot problems since when I first started working with ACES. After I tried Aces against RED's IPP2 with a R3D file with very saturated out of gamut colors I said F.T.S. and gave up on it. Then after an ACES Virtual Working Group ,don't know if was the same one as now, which involved NICK SHAW from ACESCENTRAL, ILM, NETFLIX and BLACKMAGIC DESIGN came up with the " ACES GAMUT COMPRESSION" add-in ACES version 1.3 it's the only thing I want to use now. Plus now that I am seeing what the ACES 2.0 Candidates ODTs are doing to make it even better, I am more sure than ever that I will stay with ACES over other workflows.

I think alot of individuals who use ACES were hoping for the AP1 primary Color Space to be 'Enlarged" because it is still smaller than most of the Camera Manufacturers Color Spaces which will then still have to have those Larger Color Gamuts brought back in-line with the ACES AP1 Primaries. Maybe they will but if not I think the ODTs in ACES 2.0 along with the " ACES GAMUT COMPRESSION" add-in will solve most if not all problems.


Before ACES Version 1.3 with the intergrated " ACES GAMUT COMPRESSION" add-in and the ACES 2.0 Candidate A REC709 ODT.

After ACES Version 1.3 with the intergrated " ACES GAMUT COMPRESSION" add-in and the ACES 2.0 Candidate A REC709 ODT.