This is one of those problems that shouldn't exist in theory, but reality doesn't care. Self-taught videographers who may have a background in still-photography using camera-RAW are used to a workflow that captures (mostly) in linear space. They were limited by ISO noise but, with modern sensors, other than that they had so much data in the shadows that the only thing they had to fear was blown-out highlights. Hence the tendency to underexpose and just pull out the details in lightroom or whatever. They treat log as some kind of camera-RAW for video. Some camera manuals (Sony?) will tell you to "overexpose" log by one or two stops due to where the log-curve places midlle/18% gray, and that is not what feels "normal" to those users in their viewfinders/displays... If it is not a well-managed production, all other sorts of metadata problems compound. Source material may just be delivered as generic MP4 files that don't necessarily carry all the information people are used to from the still/RAW world, where your "digital negative" has all the information you could ever wish for in one neat place...
> They were limited by ISO noise but, with modern sensors, other than that they had so much data in the shadows that the only thing they had to fear was blown-out highlights.
Photography has leaned towards towards overexposure dating back to (at least) Ansel Adams:
* https://en.wikipedia.org/wiki/Zone_System