Quote:
Originally Posted by ppppenguin
AFAIK both PAL and NTSC use the same equations for generating Y, R-Y and B-Y
Y= 0.299R+0.587G+0.114B.
R-Y and B-Y then follow.
HD uses a different equation which is annoying. It's supposed to give a greater gamut.
|
Yes, NTSC and PAL use the same electrical equations for generating Y, R-Y and B-Y. However, the camera sensor responses are (or were, anyway) matrixed differently, NTSC for NTSC phosphors and PAL for EBU phosphors. In practice these days, it's hard to tell what is being done for NTSC, as the pictures are judged on modern displays, and the camera matrix may be fudged for the manufacturer's idea of "best" while receiver matrices were also fudged for the manufacturer's idea of pleasing results with modern phosphors. In the case of PAL and HDTV, there is a known (approximate, anyway) ideal matrix in the camera, and when there is a departure from it, the result can be known. Camera manufacturers still do as they please, but they are working to a standard display reference, which many physical displays do (or can) come fairly close to meeting (at least better than the large difference between NTSC P1 green and modern sulfide green).
HDTV went one step beyond PAL, using essentially the same phosphor primaries as PAL, but also recasting the electrical equations to match the actual luminance of the new phosphors when matching D65 white. This gives a minor improvement (decrease) in visibility of chroma noise/artifacts in desaturated colors, but doesn't make much difference in saturated colors, due to the gamma characteristic of displays. For this small improvement, we now have the nuisance of rematrixing when taking signals to/from HD/SD.