Primary Grading with ACES in Resolve

(Brett Rayner) #1

Hi All,

We have just started looking into using ACES in our grading suite, but have a few questions that I can’t seem to find answers to.

What tools are supposed to be used for primary grading in ACES? The primary controls seem to affect far too much of the image to be useful. We are working in ACEScct and even then, the controls seem like “rubber”.

If I adjust the lift control, the entire picture (highlights and all) seem to shift upward. If i counteract that with a gain control down, then the entire picture (shadow and all) seem to shift down. Basically the primary controls all counter themselves too hard, it doesn’t respond like traditional log footage in the slightest.

Log controls seem to work better, but they are somewhat cumbersome to use as primary grading tools.

If anyone has any insight into the correct tool in Resolve to use as primaries, the workflow or mentality one is supposed to use when working in ACES, I would be grateful.

This community has so far proven to be an invaluable resource in making the jump to ACES!

(Fabián Matas González) #2

I just opened a post with a question really close to yours “Order of operations” I got interesting answers from the Mentors.

You should use Log controls. Specially contrast(in davinci with the S curve disabled), pivot and the offset, and then fine tune with Shadows, midtones, hightlights playing with the ranges. In my opinion first past is faster than with LGG and fine tuning little bit more slower however you get more control of what are you doing and the result looks much more organic.

(Brett Rayner) #3

Hi Fabian,

Thank you very much for your response! This is very interesting information, I will go try out your suggestions.

Out of interest, you mention Contrast with the S-curve disabled? What exactly do you mean? Is there a setting to change the behaviour of the contrast control?

Thank you very much!

(Chris Conway) #4

Hi, I believe there is s checkbox to turn off s-curve in contrast in the preferences.

(Chris Conway) #5

Look in Davinci preferences for a checkbox to disable s-curve for contrast

(Brett Rayner) #6

Thanks for this, I found it! I have noticed it does not make much difference to the way the contrast control works, is it supposed to be like this? Its basically imperceivable

(Nick Shaw) #7

In ACES that will be the case. The s-curve setting only affects whether the top and bottom end roll off or clip. The mid range is unaffected. In ACEScct the bottom end of the 0-1 range represents negative values, and the top end represents highlights up to more than 10 stops above mid grey. So by the time they are passed through the RRT and ODT, the effect of roll off at the top and bottom end is negligible.

The s-curve contrast roll off is really designed for “video” style grading, where the altered pixel values are output to the display with no further transforms. You would find a similar lack of noticeable effect if you switched s-curves on and off when e.g. grading LogC “through” an ARRI LUT.

(Brett Rayner) #8

Thank you for this info Nick. Are there any other notable differences that this makes to the way certain tools behave?

I feel that if I understand what the space is doing compared to “video” style grading, i will understand how to make corrections that work with the ACES design, not against it.

(Nick Shaw) #9

It’s not what the space is doing as such. It is the fact that the effect of changes you make is not seen directly, but rather “through” the RRT and ODT, which broadly are similar to a LUT which adds contrast (with an s-curve) and saturation to a log image.

It’s not easy to describe what the effect is. You really need to just use it and get a feel for how it behaves. It can also be useful to try the process on a linear ramp, watching the result on a waveform.

(Brett Rayner) #10

Thanks Nick. This helps my understanding immensely in terms of why the controls behave the way they do. I played around with a gradient and found the curves and “bumps” that are introduced very strange.

It seems much more difficult to make changes that you are 100% understanding of what in the image will be affected. Especially at the lift end of things. I suppose its just a new way of thinking and muscle memory.