what about green component? at DVinfo.net

Steve Witt · April 10th, 2005, 01:02 PM

Question: Can anyone explain to me in simple terms why when RGB signals are converted into Luminance (Y) and color difference
(R-Y)and (B-Y), There is no (G-Y)? Why is it done this way?

Andre De Clercq · April 10th, 2005, 01:24 PM

If you know that the Y signal is made out of (roughly) .3R+.6G+.1B and you know R-Y and B-Y you can calculte G-Y by simple substitution. In more practical terms: if you know what the R and B levels are (out of Y and R-Y and B-Y ), and you know the total luminance Y, you know how much of the G signal is needed to get that luminance.

Steve Witt · April 10th, 2005, 07:02 PM

Ok, but I thought Y=luminance. Sorry, still confused about this whole thing.

Andre De Clercq · April 11th, 2005, 02:21 AM

Indeed Steve, Y is luminance. In color camera's, incoming signals (also pure b&w scenes) are split out into 3 colors R, G and B. Each color represents a certain amount of luminance, and the total luminance signal(Y) is made out of a combination of 30% of the red signal, 59% of the green signal and 11% of the bleu signal. Bleu is e.g. a "dark" color for the human eye and so it only counts for 11% in the Y signal build up. In the above situation where we know the RGB values and the ratio in which they are present in the Y signal we get:

R-Y= .70R - .59G - .11B
G-Y= -.30R + .41G - .11B
B-Y= -.30R - .59B +.89B

Out of this we get:

G-Y = -.51 (R-Y) - .19 (B-Y)

This last expression means that G-Y is deductable from the R-Y and, B-Y values and so G-Y doesn't need to be transmitted.

Steve Witt · April 11th, 2005, 08:45 AM

thanks for the explanation Andre. It makes a little more sense now. I guess I need to brush up on my algebra.

A. J. deLange · April 11th, 2005, 03:55 PM

It is quite possible to encode B-Y, R-Y and G-Y but if you are going to do that you might as well just use R, G, B. The reason it's done this way is so that brightness information can be separated from color information. The Y signal carries no information about the hue or saturation of the scene being transmitted - only its brightness. The R-Y and B-Y convey no information about brightness (ideally, in fact they do - a phenomenon called "luminance leakage"). There are several advantages to separating color and brightness. One is that a B&W picture can be transmitted or stored by using the Y component alone. In the NTSC system, for example, if only the Y component is demodulated you have a nicely toned black and white picture and this is the basis for compatibility. Another is that we perceive detail in the Y signal. This makes it possible to transmit/store the B-Y and R-Y signals with less bandwidth than the Y signal (in DV these channels are sampled at 1/4 the rate of the Y channel). A third is that it enables color saturation and hue adjustment without affecting brightness thus making color correction easy.

In fact in many color TV sets the G-Y signal is generated (or at least it was done this way in the old days in some designs). Then R-Y, B-Y and G-Y can be applied to the cathodes and Y to the grids resulting in beam currents proportional to R, G and B.

Steve Witt · April 11th, 2005, 07:24 PM

Thanks AJ, I appreciate the info.

April 10th, 2005, 01:02 PM	#1
Steve Witt Major Player Join Date: Feb 2005 Location: kentucky, USA Posts: 429	what about green component? Question: Can anyone explain to me in simple terms why when RGB signals are converted into Luminance (Y) and color difference (R-Y)and (B-Y), There is no (G-Y)? Why is it done this way?

April 10th, 2005, 01:24 PM	#2
Andre De Clercq Major Player Join Date: Sep 2002 Location: Belgium Posts: 804	If you know that the Y signal is made out of (roughly) .3R+.6G+.1B and you know R-Y and B-Y you can calculte G-Y by simple substitution. In more practical terms: if you know what the R and B levels are (out of Y and R-Y and B-Y ), and you know the total luminance Y, you know how much of the G signal is needed to get that luminance.

April 10th, 2005, 07:02 PM	#3
Steve Witt Major Player Join Date: Feb 2005 Location: kentucky, USA Posts: 429	Ok, but I thought Y=luminance. Sorry, still confused about this whole thing.

April 11th, 2005, 02:21 AM	#4
Andre De Clercq Major Player Join Date: Sep 2002 Location: Belgium Posts: 804	Indeed Steve, Y is luminance. In color camera's, incoming signals (also pure b&w scenes) are split out into 3 colors R, G and B. Each color represents a certain amount of luminance, and the total luminance signal(Y) is made out of a combination of 30% of the red signal, 59% of the green signal and 11% of the bleu signal. Bleu is e.g. a "dark" color for the human eye and so it only counts for 11% in the Y signal build up. In the above situation where we know the RGB values and the ratio in which they are present in the Y signal we get: R-Y= .70R - .59G - .11B G-Y= -.30R + .41G - .11B B-Y= -.30R - .59B +.89B Out of this we get: G-Y = -.51 (R-Y) - .19 (B-Y) This last expression means that G-Y is deductable from the R-Y and, B-Y values and so G-Y doesn't need to be transmitted.

April 11th, 2005, 08:45 AM	#5
Steve Witt Major Player Join Date: Feb 2005 Location: kentucky, USA Posts: 429	thanks for the explanation Andre. It makes a little more sense now. I guess I need to brush up on my algebra.

April 11th, 2005, 03:55 PM	#6
A. J. deLange Major Player Join Date: Jun 2004 Location: McLean, VA United States Posts: 749	It is quite possible to encode B-Y, R-Y and G-Y but if you are going to do that you might as well just use R, G, B. The reason it's done this way is so that brightness information can be separated from color information. The Y signal carries no information about the hue or saturation of the scene being transmitted - only its brightness. The R-Y and B-Y convey no information about brightness (ideally, in fact they do - a phenomenon called "luminance leakage"). There are several advantages to separating color and brightness. One is that a B&W picture can be transmitted or stored by using the Y component alone. In the NTSC system, for example, if only the Y component is demodulated you have a nicely toned black and white picture and this is the basis for compatibility. Another is that we perceive detail in the Y signal. This makes it possible to transmit/store the B-Y and R-Y signals with less bandwidth than the Y signal (in DV these channels are sampled at 1/4 the rate of the Y channel). A third is that it enables color saturation and hue adjustment without affecting brightness thus making color correction easy. In fact in many color TV sets the G-Y signal is generated (or at least it was done this way in the old days in some designs). Then R-Y, B-Y and G-Y can be applied to the cathodes and Y to the grids resulting in beam currents proportional to R, G and B.

April 11th, 2005, 07:24 PM	#7
Steve Witt Major Player Join Date: Feb 2005 Location: kentucky, USA Posts: 429	Thanks AJ, I appreciate the info.