Quote:
Originally Posted by ChrisW6ATV
I remember watching rock concert performances on NTSC displays, and noticing massive limitations in resolution of scenes that were mostly black and blue (such as a shot of the drummer when only the blue overhead lights were on), and perhaps other black-plus-color scenes as well. Would this type of programming look better on a wideband-color NTSC display?
|
The luminance of bright saturated reds is nominally 30% of white, but due to CRT gamma and gamma correction, while the luma SIGNAL is 30% of max, the luminance of the resulting gray is about 0.3 squared or 0.09. The luminance is restored to 0.30 when the chroma signal is turned on and sets the red to 100% (green and blue reduced to zero %). This means that the chroma signal provides 2/3 of the red contrast. When the chroma bandwidth is reduced, it means that details in a red rose, for example, have only 1/3 of the original contrast. Hence, such red objects appear to have reduced sharpness. [Edit - the wideband I channel restores most of this detail contrast.]
Saturated blue objects (with details varying from full blue to black) will have the same type of loss of contrast, with the original 11% of the full black-to-white range changing to only 1%. This may be less or more noticeable than the red problem depending on the over-all content of the scene. If the scene is entirely blue, it will be much more noticeable than if the scene contains full black to white contrast and some blue objects.
If the blue is deep violet-blue or purple, it will be carried in the narrow Q channel, so will not be improved by the wider I bandwidth.