[stig olsen] "A regular TV can show black levels under 16. If I understand things correct, everyone working with normal monitoring limit their work. "
The difference is Y'CbCr vs. RGB mapped to SMPTE259M. In broadcast 0-100 IRE, the active portion of the gamut is 64-940, which allows for over- and under-shoot. AKA your "superwhites", where you are allowed to flirt with above-and-below 0-100 IRE... If you want to send 0-1023 as if it was 64-940, it will get re-mapped to those values, with resulting floating blacks and compressed whites. Its very simple math, but gets a little bit complicated as the gamma coefficient will also change the grey scale interpolation. In that case, a simple linear LUT won't fix it, even when you try to "tweak"-eyeball the values back to their unscaled rendition.
Everyone working with "normal monitoring" is not limiting their work, they are adhering to international SMPTE/EBU standards. This does not apply to digital cinema workflows involving RGB containers, for example, TIFF, TGA, CIN, DPX...
http://www.tri-sysdesigns.com/Articles/BlackisBlack.html
All of this is extremely obvious with even superficial meter-reading skills.... that is, using scopes of the outboard variety that are not being tainted by a software application's possibly erroneous assumptions.
Why this may be a problem for a lot of users is likely to be a training deficiency, as very few individuals are exposed to technical fundamentals anymore. Me, I pine for the days of Reed-Solomon, but that is all in the past. Back to the rocking chair...
jPo
You mean "Old Ben"? Ben Kenobi?
