Simple question about component video

[maxhifi]

It looks like the "Y" output (green RCA socket) can be used as a source is monochrome video for old TVs. Am I correct? The constants are different for each of the three colours, and I am not sure if it will be a correctly balanced image.

[Electronic M]

Y is the monochrome channel. Pr and Pb are the chroma difference signals. The source has to be set to 480i or else you will have a hell of a time syncing the TV to it.

[maxhifi]

Originally Posted by Electronic M

Y is the monochrome channel. Pr and Pb are the chroma difference signals. The source has to be set to 480i or else you will have a hell of a time syncing the TV to it.

Yes, but Y = 0.2126 R + 0.7152 G + 0.0722 B.

My question (not asked properly!) was that does this mix the colours together in a way which looks natural in monochrome? I did a bit more reading and saw it referred to as a pan-chromatic signal, so most likely the answer is yes. I was expecting those coefficients to be closer to equal, for a balanced signal.

This is excellent news, if it does, because I have an HDMI to component adapter En route from China and am excited to see how the 23" new vista looks with a pure monochrome signal.

[Chip Chester]

Looks like you were already there...

Chip

[maxhifi]

Originally Posted by Chip Chester

Looks like you were already there...

Chip

Yeah, I did post a thread about adding a video input to an old tv, but the way I usually watch tv is through a box which outputs only HDMI, the analog converter I have now does only composite. S-video is great for the DVD player but I wanted the Xbmc box hooked up with a better picture

[Electronic M]

Originally Posted by maxhifi

Yes, but Y = 0.2126 R + 0.7152 G + 0.0722 B.

My question (not asked properly!) was that does this mix the colours together in a way which looks natural in monochrome? I did a bit more reading and saw it referred to as a pan-chromatic signal, so most likely the answer is yes. I was expecting those coefficients to be closer to equal, for a balanced signal.

This is excellent news, if it does, because I have an HDMI to component adapter En route from China and am excited to see how the 23" new vista looks with a pure monochrome signal.

I'm pretty sure that equation is what all 3 tube RGB color cameras use to mix the 3 tubes down to base-band monochrome. White light is NOT equal parts RGB IIRC...Or at least is not in color camera design.

[ppppenguin]

For both PAL and NTSC and Rec 601 digital:

Y= 0.299R + 0.589G + 0.114B

For HD they changed the coeffts slightly. https://en.wikipedia.org/wiki/Rec._709 This was a somewhat controversial decision. It gave a larger colour gamut but left an incompatibility.

For use with vintage monochrome sets it's not at all critical. You could even use the green signal on its own and get fairly good results.

There a lot more background to this. For one thing the RGB> YCbCr matrixing is specified on gamma corrected signals. This results in some Y travelling in colour difference and vice versa. This matters because the chroma channels have reduced bandwidth. The technical name for this is failure of constant luminance.

Some early colour cameras had a separate Y tube. For example the EMI 2001 and Marconi Mk VII. This reduced registration problems and gave better compatibility with monochrome receivers. Most later cameras were 3 tube but with "highs out of green". In other words all high frequency Y was taken from the green tube only.

Further reading on constant luminance:

http://downloads.bbc.co.uk/rd/pubs/reports/1972-29.pdf
https://books.google.co.uk/books?id=...nance.&f=false
http://www.bbc.co.uk/rd/publications/rdreport_1990_02

[maxhifi]

Originally Posted by ppppenguin

For both PAL and NTSC and Rec 601 digital:

Y= 0.299R + 0.589G + 0.114B

For HD they changed the coeffts slightly. https://en.wikipedia.org/wiki/Rec._709 This was a somewhat controversial decision. It gave a larger colour gamut but left an incompatibility.

For use with vintage monochrome sets it's not at all critical. You could even use the green signal on its own and get fairly good results.

There a lot more background to this. For one thing the RGB> YCbCr matrixing is specified on gamma corrected signals. This results in some Y travelling in colour difference and vice versa. This matters because the chroma channels have reduced bandwidth. The technical name for this is failure of constant luminance.

Some early colour cameras had a separate Y tube. For example the EMI 2001 and Marconi Mk VII. This reduced registration problems and gave better compatibility with monochrome receivers. Most later cameras were 3 tube but with "highs out of green". In other words all high frequency Y was taken from the green tube only.

Further reading on constant luminance:

http://downloads.bbc.co.uk/rd/pubs/reports/1972-29.pdf
https://books.google.co.uk/books?id=...nance.&f=false
http://www.bbc.co.uk/rd/publications/rdreport_1990_02

Thank you for this post, it's exactly the info I was after. I am going to have a look at those links. I would be curious to see an actual side by side comparison of monochrome derived from both the old and new formulae, out of interest's sake, but for the purposes of giving a signal to the RCA I can be confident that the Y output is adequate.

[ppppenguin]

For deriving monochrome from RGB it really won't matter what coefficients are used within reason. You'll be hard pressed to tell the difference between Rec601 and Rec709. Not even worth doing an experiment.