|
That article on G-Y is mostly nonsense. While it's true that the G-Y demodulator needs only low gain to recover the G-Y signal, the G-Y and R-Y still contain only small proportions of B-Y and have to be amplified to the equivalant extent of a B-Y demodulator before matrixing to form the B-Y. Therefore, there is no net noise advantage. One real reason for G-Y demodulation was to try to avoid some patents on R-Y and B-Y demoduation circuits.
The way the demodulation and matrixing works is this:
R-Y is equivalent to a lot of negative G-Y and a little negative B-Y. G-Y is equivalent to a lot of negative R-Y and a little negative B-Y. Therefore, add a lot of negative R-Y to a lot of negative G-Y (in the right proportions), you cancel out a really large R-Y and/or G-Y component and leave a normal B-Y.
Edit: This could actually be a disadvantage, because the accuracy of the B-Y is now dependent on the matrix resistor values canceling the large R-Y and G-Y components.
|