Citazione:
Let’s consider the process necessary to get video off a shiny disc (or from a digital broadcast or cable channel). First the video needs to be converted from 4:2:0 to 4:2:2, then to 4:4:4, then to RGB, and finally it can be fed to the display controller. This is the same process no matter what display technology is being used, whether LCD, DLP, plasma, or CRT. It’s possible to shortcut the process slightly by going directly from 4:2:0 to 4:4:4, but in practice this isn’t used very often.
If you choose to output 4:2:2 from your player to the display, then the display will need to do the scaling of Cb and Cr to generate a 4:4:4 image and then convert that to RGB. If you output 4:4:4 to the display, the display will not need to do any scaling at all, but will need to do the conversion to RGB. If you output RGB to the display, then the display can avoid all conversion steps and send the signal right to the controller. No matter which you choose, the same conversion steps are still happening; all you are choosing is which device is performing the conversion.
There’s no specific reason that a display or a player would be the optimal place to do these conversion steps. In theory doing the conversion in the display minimizes the amount of data that has to flow across the HDMI link, but in practice HDMI is more than adequate to handle any format all the way up to 4:4:4 or RGB.
So the key to choosing the right color space to output is finding out which device does a better job of converting color spaces. This is not always easy to evaluate, and it’s quite possible for one device to do a better job in one area, like 4:2:2 to 4:4:4, but do worse in another area, like 4:4:4 to RGB.
You’d think that if a display handles a 4:2:2 input signal well, then feeding it an RGB signal would be no worse, but in fact some displays do extra work when they are fed RGB, because they convert the signal back to 4:4:4 or even 4:2:2! This happens because one or more of their internal processing chips is designed only for one color format. So for these displays, sending in any format other than the one it will use for internal processing will only add extra processing and potentially degrade the image.
The same logic applies to video processors. If the processor does all its work in 4:2:2, there’s no advantage to sending it RGB or 4:4:4, and in fact there may be a disadvantage.
Unfortunately device makers tend not to reveal the exact processing steps they use internally, or the algorithms they use to convert various color spaces to RGB. Some use different algorithms depending on which color space is fed in. The bottom line is to assume nothing, and test every combination.
