We have discussed this at some length over the years. Perhaps you could give a reference to the standard display gamma for BT.709? We have previously concluded that there is no right answer.

Sure.

ITU-R BT.1886: https://www.itu.int/dms_pubrec/itu-r/rec/bt/r-rec-bt.1886-0-201103-i!!pdf-e.pdf
In this document, you’ll find that the reference gamma is defined as 2.40. However, the actual EOTF specified by BT.1886 also depends on the display’s black level (as shown in the document). For displays with excellent black performance (such as OLED or Mini-LED), the gamma matches 2.4.

Appendix 2 includes the older BT.709. When referred to as a gamma curve, it uses a hybrid function: a linear segment near black followed by a power function (with γ = 0.45). The equivalent pure gamma is approximately 1.96, and it’s primarily used by cameras as an OETF.

Gamma 2.2 is a simplified approximation of sRGB gamma, commonly used for consumer monitors, web graphics, and still images. While some people now apply it to video, it is not a broadcast standard.

In DaVinci Resolve, when rendering with RCM set to Rec.709 gamma 2.4 or simply without RCM, the exported video gamma is 2.4. That’s why I suggest changing DoM’s default setting accordingly.

Thanks for the link.

As I understand it, the ITU doc says nothing about the Rec.709 EOTF gamma. I get your point about Resolve, but I'm not sure we should change a over 10-year default in DCP-o-matic to match just one other software package (even if it is a popular one).

Also, as far as I can see the ITU doc only talks about a different (i.e. not 2.4) Rec.1886 gamma if you are trying to match a CRT. DoM's use of 2.4 for Rec.1886 doesn't seem unreasonable to me.

I could see a justification for renaming the current Rec.709 setting Rec.709 (gamma 2.2) or something, and renaming what we currently call Rec. 1886 to Rec.709 (gamma 2.4) / Rec.1886 (or possibly not mentioning Rec.1886 at all, as you suggest).

Suddenly switching Rec.709 to 2.4 gamma after all these years seems like it would surprise a lot of people.

Thoughts welcome!

In the Opto-Electronic Conversion Recommendation ITU-R BT.709, it states:

Overall opto-electronic transfer characteristics at source:
V = 1.099 × L^0.45 – 0.099 for 1 ≥ L ≥ 0.018
V = 4.500 × L for 0.018 > L ≥ 0

So for 1 ≥ L ≥ 0.018, gamma=1/0.45=2.22
However, there's a a linear segment V = 4.500 × L. So overall it's closer to gamma 1.96.

As for the EOTF, BT.1886 recommends using a gamma of 2.4, as I mentioned earlier.

It’s okay that DoM uses gamma 2.4 in the ‘Rec.1886’ preset, it’s just this is not strictly accurate from a technical standpoint. I apologize for any confusion caused.

However, I believe using gamma 2.2 in the ‘Rec.709’ preset is more questionable, since the standard EOTF for broadcast video is clearly defined as 2.4. I understand your concerns.

It'd be a good starting point to relabel the preset as ‘Rec.709 (gamma 2.2)’. (and the other as Rec.709 gamma 2.4). That might help users like me who were confused by the naming.

Thank you again for your response—and for creating DoM. It’s truly the most powerful yet user-friendly DCP tool I’ve ever worked with.