Wide Gamut

I would not say that over-size color gamuts are the norm. In fact, I see many more 2012 displays that are fairly good vs. displays from 2 or 3 years ago.

If the colors aren't accurate and the controls can't fix the problem, all you can do is get the colors as close as possible. There's no "dealing" with anything. You get the colors as accurate as you can with the controls that are available. Primary and complimentary colors should both be as accurate as they can be made. Residual errors have to be tolerated. The only fix for oversaturated colors that the display can't fix is a good viedo processor. Right now the ones with the needed capabilities start in the $1200-$2000 range depending on which company you are looking at.

Color "Decoding" is taking the YCbCr data and turning into RGB.

CMS is the ability to remap the location of RGB.

And if you think back to the olden analog days, color and tint where originally amplitude and phase there to compensate for the effects of broadcasting at distance. The color and tint controls originally manipulated the signal before the color decoder.

Let's take a hypothetical case where all the colors but green are pretty close to where they are supposed to be and green is WAY over saturated. You're qiestion is... should you then locate cyan half-way between green and blue (which would displace cyan towards green) and should you displace yellow towards green putting it half-way between green and red.

The answer should be pretty obvious. Do you want sky blue that is contaminated with green? Do you want lemons looking yellow-green instead of yellow? Obviously NO. So if you can't fix green but you CAN move cyan and yellow around, the best place (in this example) for Cyan is at the Cyan reference coordinates for HD color space and you want Yellow to be on the reference coordinates for Yellow in HD color space to keep as many colors as accurate as possible.

There are no rules for what to do when something is broken. There are only rules for how they should work when they work correctly.

I know that's not the answer you want to hear, but it is the answer none the less.

So when you go to make a compromise, this is where personal preference plays as much as part as anything else. If you had a set of saturation swatches you could run those to see what would cause the smallest error in general.

would your answer be any different if all three primaries were wildly over saturated AND had significant hue errors?

Why would there be a different answer? Every color that is inaccurate is going to be inaccurate. If Red, Green, and Blue were way off, but yellow, cyan, and magenta were accurate, you'd have sky that looks fairly good, lemons that look fairly good, and magenta azaleas and rhododendrons that look pretty good and everything else would be pretty bad-looking.

If you then moved cyan, yellow, and magenta to points half-way between the bad locations of red, green and blue your sky, lemons and flowers would look terrible too. So there's no easing the misery by moving ANY color to a less accurate position in Rec. 709 space to try to "compensate" for some other color being inaccurate.

I was just trying to see if there was any scenario in which the standard Rec. 709 targets would not be used. Based on your posts, it seems there isn't.

And yet some say your eyes and brain will adapt to virtually any picture and they can not be trusted. Your description here is very apt. Errors in yellow, blue, and green have always been quite visible to me. Saturation is a different story, but hue for many of us is quite visible.

If you are calibrating for someone else, then you will be much more limited in the sophistication of concessions and multiple calibration experiments that can be done because of time constraints. So in this case, yes you just follow the hard rules to save time. If you are calibrating your own device, there are no rules, you can do unlimited number of calibrations and play around.

You're mixing two different statements here, Doug's point is that if you calculate your secondary targets from your measured primaries, secondaries hues will be off when watching any content with yellow, cyan, or magenta in it. In other words, secondary hues should be set with Rec. 709 targets, not custom ones based off of measured primaries.

What he isn't saying (but apparently you are suggesting) is that we can trust our eyes when setting the hues of the secondaries (and therefore don't need a meter to do it correctly). Yes, if something is way off the errors will be visible to our eyes, but saying that is not the same thing as saying we should set up our TVs without a meter OR that we can accurately. Having a reference point is the best way to spot a color error with real program material.

^ ^ ^

I said what I said. . . to Doug.

I said or implied nothing of the sort of what you the have regarding it, but just another example how you, and some others, like to impute and redefine what others do say. For reasons that just seem to be argumentative.

We have trouble (relative) detecting both saturation and luminance errors for color. It's not that we can't see them at all... we can see them fairly easily if we have a reference. But when the errors are in video or movies, we don't know exactly how something should look in those particular lighting situations surrounded by whatever else happens to be in the scene (there's also intentional manipulations by the director/cinematographer).

You WILL NOT adapt to every kind of error. Especially errors that occur in colors known as "memory colors" like sky, Coke or Target red, lemons, grass, oranges, apples, evergreens, and tons of other colors we see day in, day out throughout our lives. Things we adapt to are the less familiar content and white. White is probably the biggest adaptation and the easiest place to get fooled... our visual system latches-on to the brightest and bluest thing in a scene and makes that appear white. True white then appears to be too yellow. But our adaptation only goes so far. If images are too blue, we will see fleshtones go toward the zombie/vampire end of the spectrum, reds will never look right, yellows will look odd, etc. etc. So, yes, there is SOME adaptation, but it's not unlimited.

Yes, and it was a post on a discussion forum, not a private message. You like posting in a discussion forum but don't really want to have a discussion when someone disagrees with you or makes a valid point that conflicts with what you posted. Why?

Why not just focus on the topic of discussion instead of making such remarks?

Why? Because Doug makes valid points and, as I said( re-read it), you redefine and impute what myself and others have to say. So it isn't a discussion anymore. End of discussion for me.