Originally Posted by Kris Deering
Not many solutions that use metadata for tone mapping.
For this title specifically, the Lumagen will do an average between the two (MaxCLL and MaxDML) to achieve its peak for the tone map. For the Panasonic 820, it will use MaxDML if the MaxCLL is higher. If I had to chose how to handle it in this situation, I would chose the Panny.
What do you mean not many solutions use metadata? You just named two solutions that use metadata for tone mapping.
They have to look at metadata and possibly also measure for themselves if they are to adapt to content on a film by film, which the new JVC's do, or scene by scene basis, right? We need to test which takes precedence on the new models which I am very eager to do. I am just observing that, in either case, if the JVC is tone mapping for 4000 nits, or 10,000, then its going to be the darker of those two photos shown, hence the Sony is the bright one for sure. And in fact, this is not a good film to test metadata based Tone Mapping IMO since its pretty all over the place.
The old tone mapping of JVC, Sony, most
UHD TV's are just using a standard one solution tone map curve and you use contrast slider to clip.
What does the 820 do with the Masciola Suite? It has content at 10,000 nits, yet its metadata states Max CLL is 1000.
Interestingly, here are three cases where Mad Max absolutely blows through 4000 nits.
The problem is though, these will be so fast that dynamic tone mapping probably wont be able to actually adjust quick enough to these instances. Still interesting though.