Calibration Settings for Resolve on a MacBook Pro

[sssammm]

I know elements of this question/process have been spoken about on here but I’m still a bit overwhelmed with the process. My goal is to get the most accurate image I can from my MacBook Pro  (using Resolve to grade video).

I thought I had calibrated my monitor using my i1 Display Pro (and the software that came with it) but when I rendered a ProRes 4444 file, it looked significantly brighter & less saturated than inside Resolve. I’ve come to learn since about variances between Quicktime & VLC, as well as the ability to create a 3D Lut for inside of Resolve to make the GUI match the the rendered version. I also started to wonder if I correctly calibrated my monitor in the first place and after a few experiments using DisplayCal it does seem like there’s less of a difference than I first noticed — after just a basic sRGB calibration.

I’m still quite unsure of the best process for trying to have accurate (as possible on a MacBook) calibration that also matches (as close as possible) to what I’m seeing inside Resolve. What colorspace, tone curves (in the calibration & 3D Lut), etc should be used? How do Rec 709 2.2 / 2.4 relate to Rec. 1886, and how does that relate to sRGB?

I’m sure I’m probably missing basic info here but after trying to read through all of the available info I find myself unclear on the whole thing and would really appreciate help from someone more knowledgeable than me!

Calibrite Display Pro HL on Amazon  
Disclosure: As an Amazon Associate I earn from qualifying purchases.

[Vincent]

DisplayCAL calibration, like other calibration solutions relying on graphics card calibration, are just meant to fix white, grey neutrality and gamma, and after that create a profile in order that color managed apps know the displays capabilities (gamut coverage, gamma and such).

If you want to edit or view a Rec709 video that is meant to be viewed in a reference Rec 709 2.4 display you need to use a player or editor that supports some kind of color management.
For editing you have Resolve, create a LUT3D that transforms Rec709 content RGB values encoded with your reference gamam (like 2.4) to equivalent color RGB values in your display.
For vieweing you can use a player that supports LUT3D (not in OSX AFAIK, you’ll need Windows and madVR compatible layer like MPC-HC) or a player that supports ICM/ICC color management, like MPV. I’m not an expert in MPV but AFAIK its configuration relies heavily in comand line or configuration text files. Take a look on its documentation.
IDNK what color managemend capabilities are supported in VLC or Quicktime. I think that your bestchance in OSX is to use MPV even if its configuration learning curve is high.

[troy-s]

DisplayCAL calibration, like other calibration solutions relying on graphics card calibration, are just meant to fix white, grey neutrality and gamma, and after that create a profile in order that color managed apps know the displays capabilities (gamut coverage, gamma and such).

Uh… what? It can calibrate and profile fine, and can skip the VCGT in an ICC / ICM entirely.

I know elements of this question/process have been spoken about on here but I’m still a bit overwhelmed with the process. My goal is to get the most accurate image I can from my MacBook Pro  (using Resolve to grade video).

If it is a MacBook from late 2015 onwards, your colour primaries are different and it will require that your application honour the Apple DCI-P3 primaries.

I thought I had calibrated my monitor using my i1 Display Pro (and the software that came with it) but when I rendered a ProRes 4444 file, it looked significantly brighter & less saturated than inside Resolve.

Colorimeters are typically unsuitable for profiling displays beyond REC.709 based lights, and will deliver problems without correction files. More ideally would be to use a spectrophotometer, such as the entry level i1 Studio or the previous incarnations.

I’ve come to learn since about variances between Quicktime & VLC, as well as the ability to create a 3D Lut for inside of Resolve to make the GUI match the the rendered version. I also started to wonder if I correctly calibrated my monitor in the first place and after a few experiments using DisplayCal it does seem like there’s less of a difference than I first noticed — after just a basic sRGB calibration.

What year / model is your MacBook Pro?

I’m still quite unsure of the best process for trying to have accurate (as possible on a MacBook) calibration that also matches (as close as possible) to what I’m seeing inside Resolve. What colorspace, tone curves (in the calibration & 3D Lut), etc should be used? How do Rec 709 2.2 / 2.4 relate to Rec. 1886, and how does that relate to sRGB?

The ISO glossary definition for an RGB additive colour space is helpful here. An RGB additive light colour space must include three facets. The three facets are:

• A concrete, absolute colour science description of the three basis light primaries.
• A concrete, absolute colour science description of the “white point”, or achromatic colour when R=G=B.
• A well defined transfer function, or a relationship of intensity of encoded values and a ground truth of linear ratios of light.

REC.709 defines only the encoding aspect of code values. That is, the values encoded to the REC.709 standard should only be used as encoded, not for display. The display aspect is handled by BT.1886, which at risk of greatly oversimplifying things. a pure 2.4 power based transfer function at the display. REC.709 also defines the basis light primaries, and achromatic colour.

The base primary lights can be thought of as the colours of the three little lights if you were to zoom in on your display with a droplet of water or a magnifying glass. That is, on a typical sRGB display, they are of a certain colour. The achromatic white point when R=G=B, will also be another colour. That is, the basis lights in a traditional REC.709 display are different from all MacBook Pros and Apple devices from around late 2015.

In summary:

• The basis primaries for REC.709 are colour science defined absolute “colours” according to colour science.
• The basis primaries for sRGB are identical to the lights defined in REC.709.
• The white point, or achromatic colour when R=G=B, is well defined for REC.709 as “D65”, which is simply another colour science absolute colour.
• The white point, or achromatic colour when R=G=B, is well defined for sRGB matches REC.709’s precisely.
• The only difference between REC.709 and sRGB is essentially the transfer function.
• The transfer function for REC.709, also known as an Opto Electrical Transfer Function or OETF, is strictly for encoding.
• The decoding transfer function, or intensity mapping for viewing, is defined in BT.1886 and defines a 2.4 power function, and dim setting for viewing.
• You have to be very careful with all Apple products post-2015 as the colours of the little lights in the RGB display are different from REC.709 / sRGB. This means that if you think of RGB encoded values as indicating merely intensity of light, it communicates nothing about the colour of the lights in question. Sending encoded values designed for REC.709 with BT.1886 light output to a post-2015 Apple MacBook Pro set of lights will result in completely wrong light mixtures. This is identical to mixing paint using precise measurements, but with three different paint colours.

Hope this helps to get your started. There is plenty of misinformation out there, and plenty more folks willing to spread misinformation.

[Author]

Posts