HDR10+ Anchor Points and JSON Profiles

Been looking at HDR10+ and it seems that the anchor points are always greyed out? Analysed all clips, selected clips etc, always greyed out.

Also, noticed there is now a HDR10+ Profile A and Profile B json export. Cannot finf anything regarding this. If someone can provide information or point to a resource that would be most appreciated.

Hi,

yes same problem here. All ancher points are greyed out... I can not figure out why. Did you find any solution yet?

Best

[Peter Chamberlain]

All i can add to this conversation is to say the guidance we have is to make these view only parameters and if the automatic process is not getting good results please PM me the detail so i can feed it back to the designer.

Is this still the case ? So HDR10+ tab is for viewing only ? Or do i miss some config.
The dolby tab controls get activated after entering a license but HDR10+ni can not get active aparantly.

Would it be possible for BM to share an update to the road-map for expanding HDR10+ support in Resolve please?

Failing that, it would be very helpful to explain (in more detail than the manual) what exactly the current implementation of the 'Analyze clips' data is actually doing, and how does the output-nit value, and mastering luminance et al affect or guide the analysis.

It's all a bit black-box at the moment!

Many thanks.

**UPDATE**
After some poking around in the exported HDR10+ json files, and according to this line:

Code: Select all

"TargetedSystemDisplayMaximumLuminance": 400

... it looks like that for the moment, Resolve is hard-coded to targeting an arbitrary fixed display max luminance of 400nits for its 'Analyze clips' command. I believe in the SMPTE ST2094 standard, it allows for metadata to target many different display max luminance levels – like Dolby Vision. Would be nice to at least allow Resolve uses to select this target value themselves! Something for Resolve 18 maybe?

**UPDATE 2**

Further tests reveal that changing Output colour spaces results in the following metadata in the json files:

ST.2084 300nits = "TargetedSystemDisplayMaximumLuminance": 400"
ST.2084 1000nits = "TargetedSystemDisplayMaximumLuminance": 400"
ST.2084 2000nits = "TargetedSystemDisplayMaximumLuminance": 500"
ST.2084 4000nits = "TargetedSystemDisplayMaximumLuminance": 500"
ST.2084 (no limit) = "TargetedSystemDisplayMaximumLuminance": 500"
Rec.2100 = "TargetedSystemDisplayMaximumLuminance": 500"

Is this intended/expected HDR10+ behaviour? Can anyone explain the logic?

Thanks

I was also wondering why the JSON get limites to 400 nits. That's not logic.

Bedtime reading for HDR10+ metadata structure:

https://muygs2x2vhb2pjk6g160f1s8-wpengine.netdna-ssl.com/wp-content/uploads/2021/09/A341-2019-Amendment-No-2.pdf

…reveals how the various HDR10+ trim parameters (that we can't yet adjust in Resolve) feed into creating a basis OOTF: a hybrid-Lin/Log (Linear part + Bezier Curve) tone-map that the final presentation device (say TV) then interprets with its own engine, and according to manufacturer of HDR10+ supported device! Tricky to say the least.

My [ongoing] tests reveal that simply manually manipulating the ST.2094-40 data fields in the json file (and re-muxing the output file) produce little effect [on my HDR10+ Panasonic OLED.]

Until we know how the various data points in ST.2094-40 map to a corresponding interface in Resolve, we're all at sea with HDR10+.

**UPDATE**
After some poking around in the exported HDR10+ json files, and according to this line:

Code: Select all

"TargetedSystemDisplayMaximumLuminance": 400

I can confirm that I get the same result exporting Profile B json metadata. But when I choose Profile A things get even worse than that as I see

Code: Select all

"TargetedSystemDisplayMaximumLuminance": 0

and also

Code: Select all

"MaxScl": [0,0,0]

that doesn't look good (it is [10000,10000,10000] when exporting using Profile B).

Bugs?

Sergio