Originally Posted by gadgtfreek
PM Rich B, he can explain it well and it was very helpful to me when considering a 17 or A1E. Sony chose to map some, and then clip the rest, while LG chose to try and map more which means dimming the panel down to hit those higher nit highlights so they did not clip.
Evidently, one complaint about the 2016's were that they did not show as much of the range, so I think LG changed in 17 to cater to these reviewers. I prefer the stay bright, and just clip what you can't show method to be honest.
I think he went from a B6 to a C7.
The LG 7 series implements tone-mapping that uses the static metadata to select a tone-mapping algorithm.
LG chose to preserve highlight detail but at the cost of lowering the overall picture level.
You can see these two curves in Vincent Teoh's review here:
The curve used metadata max luminance of 1000 nits
The curve used metadata max luminance of 4000 nits
Attached is an image of Batman V Superman: Left Sony, Middle LG, Right Panasonic.
The LG is the darkest because it darkens the image below 100 nits.
LG's Active HDR mitigates the problem by dynamically analyzing scenes and restore PQ EOTF tracking and it helps a lot.
If you have a Sony/Columbia UHD BD you can access test patterns my typing 7669 at the menu.
These patterns have static metadata of 10000 nits which is even more aggressive. The 100 nit bar pattern is clearly dark and nothing close the 100 nits because there is a ramp on the bottom that includes 10000 nits. In this case, the LG appears to fall-back to the aggressive tone-mapping and Active HDR does not change that.
Sony ignores the static metadata when dynamic tone-mapping (as it should). Static metadata is not consistent from title to title and there is no argument for two disks with the same scene displayed at different average picture level because they have different static metadata with dynamic tone-mapping engaged.