Question about HDR10 BluRays with high maxcll values due to one single frame

@Markswift2003 sorry to ping you but I think you may know the answer to this.

I watched HDTVTest's analysis of the mad max fury road Blu-ray. They show that there is like one single reflection that spikes to 10,000 nits. The guitar fire dude hits to like 1000 nits with a few spikes to 4000, and the rest of the film is lower.

In the Moon (2009) blu ray maxcll shows 3500 nits. I used windows wcg app to detect maxcll and i couldn't find any scene that went above like 1000 nits. So i assume it's something similar here where one object or something hits 3500 nits.

Is this normal? Or is this a case of early HDR10 blu ray mastering not being very good? I'm asking because in moon, I think a 1000 or 2000 nit tone curve would probably be best, but my LG will use a 4000 nit tone curve based on the metadata.

Just wanted to hear anyone's opinion!

 

A bit of both, but in theory it's completely normal - MaxCLL is defined as the luminance level of the brightest pixel in the presentation, so if one pixel hits 4000nits, MaxCLL is 4000.

This is why you have MaxFALL which is the maximum frame average light level - the overall average luminance of a frame. So you might get MaxCLL of 4000 and MaxFALL of 500 as an example.

Tone mapping should take into account both values to work properly but in practice it doesn't always and in fact some sets only use MaxDML which is really dumb because that's just the spec of the mastering monitor used to grade!

But yeah, a lot of the early content has some needlessly wild figures partly because I think at that time people used to think HDR was supposed to be brighter than SDR so they racked it up a bit!

 

Do any devices use MaxFALL? I was under the impression devices always use MaxCLL and/or MaxDML but I've never heard of one using maxFALL.

 

Do new blu rays try to limit maxcll peaks to avoid jacking up the tone curve based on like one single frame?