[Tig] Looking for: Pix of Visible light with NTSC, YUV, & RGB Triangles
Fri Sep 30 04:07:02 BST 2005
If you're looking for a simple layman's explanation - the clearest one
I've found to explain it in non technical terms is to have people first
think about RGB. How do we make colors in RGB space? We have different
levels - the RGB signals must be different otherwise we have white,
black or shades of grey. Can you have full amplitude RGB signals or zero
level RGB signals and have color - no it's white or black. When you're
define colors in terms of luminance and color difference signals
(whether YUV, YPrPb, YCrCb or any variation) you disconnect the color
components from luminance (to some degree ;-). It's now possible to have
full white or black and still have a color. Take a look at SMPTE color
bars in analog and notice the U and V components then put that signal
through some RGB processing and watch the levels get clipped. Next take
a look at a monitor being fed the same signals - IF THE MONITOR AND
SIGNAL PATH IS PROPERLY CALIBRATED you should see no difference. The
reason is the signal being clipped is going negative - since the monitor
is an RGB device it can only display non-negative signals. As the
electron beams illuminate the RGB phosphors they emit light - if they
somehow worked with negative then instead of emitting light they would
absorb light - doesn't happen :-)
Again - this is intended to be a non-technical explanation for laymen.
For good discussions of the details I suggest Poynton's books.
Jan & Eileen wrote:
>Thanks to Bob Sliga and Cutters Inc. for supporting the TIG.
>I'm looking for a picture that I remember seeing in the PTV
>Color Analyzer manual, showing the visible light colors, with
>NTSC, YUV, & RGB Triangles superimposed on it. It's a good
>Visual explaination of colorspace issues between RGB & YUV/NTSC.
>I've found a pix of the CIE triangle, but not of all three...
>Anybody have one? Thanks.. Jan
50 East 42 Street
New York, NY 10017
(212) 986-3833 fax
More information about the Tig