Wii: Component-Composite quality compared

[Hexman]

Hello,

for all you people out there asking if the 3-Component-Video cable makes your image quality look better than the one coming with the Wii (Composite):
I used a digital camera to take shots from the output my video beamer produces. One time with the 3-Component-Cable connected, one time with the included composite cable.



On the right side you can see what you get with the included cable (what "crispy picture" means). Also, a blue line on a yellow background turns into green. With a 3-Component-cable you figure out that you have 2 First-Aid-Packs, with the Composite (usually yellow video) cable you can only see that you've got 2 pieces of something red 

The bad thing is that the 3-Component-Cable has no Composite-out. Because of this I always have to connect the other cable when switching from Beamer to TV and vice versa (the old TV set does not support component in).

The shot is taken from the game "Alien Syndrome", and the part you can see is about 25% of the whole image width and height. Feel free to use this image anywhere else.

[blackevilweredragon]

Wait, is it RGB, or Component?  They are NOT the same.

[Hexman]

Right, you got me. What comes out of my european Wii is a 3-component-signal, not RGB.
I changed the title from RGB-Composite to Component-Composite (and some parts of the text).
It's easy to mix things up as the three cables are colored red, green and blue. But I also had to find that out the hard way when I finally found a menu entry on my DVD-Recorder (Video-Out) with its options RGB or Component.

Because I am new with television-standards, I also have to add:
I'm sitting in europe and my Beamer says that the signal is 576i with both cables.
For people in the 480i / 480p region, this picture might not tell you the truth about what would happen to your image quality with the 3-Component cable.

[blackevilweredragon]

No matter the resolution, Component will always look the same, it's not encoded like PAL or NTSC...

[NFG]

Component video IS encoded.  Of all the standards we discuss, the only one that's not encoded is RGB.  Component video not only has encoded parts, it can be manipulated with additional crap like Macrovision 'protection'.

Basically any CRT can display RGB, but component requires 'decoding' and some complex math to build a signal from the component stream.

[blackevilweredragon]

Component video IS encoded.  Of all the standards we discuss, the only one that's not encoded is RGB.  Component video not only has encoded parts, it can be manipulated with additional crap like Macrovision 'protection'.

Basically any CRT can display RGB, but component requires 'decoding' and some complex math to build a signal from the component stream.

I was comparing it to NTSC and PAL type encoding, in which it's not at all like them.

I was to vague, my bad.

[kyuusaku]

It's everything like NTSC and PAL since RGB -> YPbPr is the first step in encoding NTSC and PAL video! The Y in "component" video is the same as the Y in S-video.

[blackevilweredragon]

No, it's not.

Luma is luma, so what?

Pr and Pb are far different than the NTSC or PAL Chroma though...

[Endymion]

Hello,

for all you people out there asking if the 3-Component-Video cable makes your image quality look better than the one coming with the Wii (Composite)

Who are "all these people?" I am guessing they are probably not regular readers of this site, believe me we know that it is better and why.

And welcome to the board.

[Hexman]

Thanks,

yes, I made these pictures for people like me.
When I got my beamer I looked for 3-component outputs on every device I could connect. I never used something else than composite before, but had light horizontal and moving bars using a 45ft composite cable. For the Wii I found the wiki page and this forum, but there seems to be no site with a working picture for people to compare if it's worth buying the Wii-cable, a switchbox and additional cables per device, and 3 more 15m (45ft) cables to connect everything to the beamer this way.
Very happy about the blind investment I wanted to share my knowledge to others 
Please don't bring me down by posting links to hundreds of pictures I could not find, hehe.

I also found out: Watching a video does not make you recognize the quality change so much. As long as there are no sharp lines (like buttons in a dvd-player menu), there are of course no crispy lines.
But those sharp lines always seen in console games even lets one recognize the difference on a small tv.
That must also be the reason why there is a third-party Wii-to-scart cable sold in our stores while there is a composite-to-scart adapter coming with the Wii 

Go and buy cables, I cannot live without 3-component anymore 

[kyuusaku]

No, it's not.

Luma is luma, so what?

Pr and Pb are far different than the NTSC or PAL Chroma though...

Pr and Pb are QAM modulated to become chrominance!

Stop acting like you know what you're talking about.

[blackevilweredragon]

Give it up.

Pr is red minus Luminance, and Pb is blue minus luminance.

Chroma would have to be either UV or IQ, of which, is not Component Video.

All luma is, is the sum of R, G, and B.

QAM is digital btw.

[kyuusaku]

Give it up.

Give what up?

Pr is red minus Luminance, and Pb is blue minus luminance.

All luma is, is the sum of R, G, and B.

Your point? Can you say something that isn't irrelevant?

Chroma would have to be either UV or IQ, of which, is not Component Video.

QAM is digital btw.

I rest my case.

[blackevilweredragon]

In case I wasn't being clear enough (which wouldn't be the first time with my disability), let me try this again...

Ok, Y is one step to S-Video.  It is luma.

But Pr and Pb can't be directly converted to chroma, not without the help of Y.  Pr and Pb need Y, to even get the color green, as Pr and Pb do not have Green at all.

Sorry for the confusion.  I thought you were trying to say Pr/Pb could be directly turned to Chroma by itself, which obviously isn't true.  Nevermind then.

[kyuusaku]

Ok, Y is one step to S-Video.  It is luma.

But Pr and Pb can't be directly converted to chroma, not without the help of Y.  Pr and Pb need Y, to even get the color green, as Pr and Pb do not have Green at all.

Pr and Pb *ARE* directly converted to chroma, Y has nothing to do with it.

[blackevilweredragon]

Where does green come from then?  Pr and Pb sure don't have green.

[NFG]

Gentlemen, a gentle reminder to please keep things civil here.  So far so good, but watch your steps ok? 

[blackevilweredragon]

Lawrence, since your the master of Component video, could you explain the Y/Pr/Pb to NTSC transcoding process?  I figure Y would be requred to get chroma, because Pr and Pb are the difference of green, and Y after minus red and blue, is green, correct?

[viletim]

Lawrence, since your the master of Component video, could you explain the Y/Pr/Pb to NTSC transcoding process?  I figure Y would be requred to get chroma, because Pr and Pb are the difference of green, and Y after minus red and blue, is green, correct?

I can explain it. Y is the sum of of R, G, and B (a weighted sum, but i'll ignore that because it'll just compilicate the explanation). U(Pr) is made by subtracting Y from R and V(Pb) comes from subtracting B from Y. G can now be dispensed with. To get back to RGB the reverse is done:

R =  Y + U
B =  Y + V
G =  Y - U - V

Now we have YUV and want to make YC (S-video). Y is fine as it is, but U and V are modulated together in such a way that they don't interfere with each other (too much). C can be seperated into U and V through a demodulation process. To get from YC to "CVBS"(video) you simply add Y and C together.

Simply put, the reason that RGB is converted to YUV is that the modulation process that U and V are put through requires that the bandwidth for these signals be low, taking up as little space on the broadcast spectrum as possible.

[blackevilweredragon]

So, if I understand you correctly, even over S-Video, green is still never sent then?  That's what's throwing me off, is the mysterious green channel.

Also, I thought YUV was the wrong term for Component video, as per what Wikipedia states.

(Btw, thanks for your reply.  Your explanation is what I prefer, rather than "it just can".)

[viletim]

Don't rely on wikipedia too much...For technical info it's more confusing than useful.

YUV = Weighting for PAL
YIQ = Weighting for NTSC
component video (ie, the three phono connectors on the back of your TV) = Weighting for DVD

All three are different (oh so slightly) but as I ignored weighting in my explanation so it doesn't matter too much.

BTW, I've seen component video inputs labeled on the back of TVs as: Y Pr Pb, Y Cb Cr, Y U V, Y R-Y B-Y, and plain "component".

[blackevilweredragon]

My HDTV labels it as:  Y/ CrPr/CbPb

But I still don't get it, where does green come from?  It dissapears and is never sent.  I wish GamesX still had that page up.  I couldn't find it.

[kyuusaku]

Green *is* mathematically derived from Y, Pb and Pr.

These are the steps to encode RGBS into composite:
-Transcode RGB into YPbPr (YIQ/YUV)
-Sum Y with composite sync
-Delay Y, can't remember for how long...
-Mix Pb with subcarrier
-Mix Pr with subcarrier delayed 90 degrees
-Sum modulated Pb and Pr
-Sum Y and C

So to recap:

-You were wrong in thinking that chrominance was not UV/IQ was not Pb/Pr.
-You were also wrong in thinking QAM is "digital". QAM is two modulated signals of the same frequency 90 degrees out of phase.
-This thread contains empirical evidence that you truly don't know enough to argue coherently and you're so stubborn you somehow only realize it yourself when you're shown by people who I'm sure have put you in your place before.

[blackevilweredragon]

QAM:  Quadrature amplitude modulation

That IS used in digital TV tuners, so I was right, but I was thinking of the wrong QAM.
http://en.wikipedia.org/wiki/QAM_tuner

I'm slow, I said that, you just don't have to be a rude about it.  Just correct me with facts, and i'll read it.  I only bite back when you insult me at a personal level...

So, once again, sorry I'm slow, not my fault I was born this way.

[kyuusaku]

I was correcting you with facts; it's not my job to convince you of them.

[blackevilweredragon]

The only way I learn, is reading topics like these, because as mentioned, Wikipedia isn't the best place to learn.

I learn passively, unless I really want to learn about something specific then I'll read up hard.

Component video is one of the passively learned things about me.

Anyway, can't we just drop this?

[NFG]

OK, after reading all this I'm a bit confused.  viletim said

Now we have YUV and want to make YC (S-video). Y is fine as it is, but U and V are modulated together in such a way that they don't interfere with each other (too much). C can be seperated into U and V through a demodulation process. To get from YC to "CVBS"(video) you simply add Y and C together.

I was under the impression that Y/C video was Y + colourburst (chroma).  Does this mean, viletim, that the C in Y/C is just red and blue (PrPb, CrCb whatever) and the green still comes from the Y channel after they're mathematically combined?

The way I thought it worked was, working backwards, a colour composite (not component) signal was a black and white transmission with a low-bandwidth colour burst (chroma) at the start of every horizontal line.  From there I thought Y/C was these signals separated, but C still included green.  Is this wrong? 

[blackevilweredragon]

OK, after reading all this I'm a bit confused.  viletim said

Now we have YUV and want to make YC (S-video). Y is fine as it is, but U and V are modulated together in such a way that they don't interfere with each other (too much). C can be seperated into U and V through a demodulation process. To get from YC to "CVBS"(video) you simply add Y and C together.

I was under the impression that Y/C video was Y + colourburst (chroma).  Does this mean, viletim, that the C in Y/C is just red and blue (PrPb, CrCb whatever) and the green still comes from the Y channel after they're mathematically combined?

The way I thought it worked was, working backwards, a colour composite (not component) signal was a black and white transmission with a low-bandwidth colour burst (chroma) at the start of every horizontal line.  From there I thought Y/C was these signals separated, but C still included green.  Is this wrong? 

Lawrence, you said exactly what I was trying to get out of my mouth.

Everytime I look at this, it's that Chroma comes from R-Y and B-Y?  But how is that related to Pr and Pb?  I am far too confused.  I know how Component Video works, but it's the S-video and Composite I don't.
EDIT:  Ohhh I think I see where my mistake has been.  R-Y and B-Y ARE Pr and Pb..  Just different terminology.  Correct me if I'm wrong please.

[viletim]

OK, after reading all this I'm a bit confused.  viletim said

Now we have YUV and want to make YC (S-video). Y is fine as it is, but U and V are modulated together in such a way that they don't interfere with each other (too much). C can be seperated into U and V through a demodulation process. To get from YC to "CVBS"(video) you simply add Y and C together.

I was under the impression that Y/C video was Y + colourburst (chroma).  Does this mean, viletim, that the C in Y/C is just red and blue (PrPb, CrCb whatever) and the green still comes from the Y channel after they're mathematically combined?

The way I thought it worked was, working backwards, a colour composite (not component) signal was a black and white transmission with a low-bandwidth colour burst (chroma) at the start of every horizontal line.  From there I thought Y/C was these signals separated, but C still included green.  Is this wrong? 

It's wrong. The colourburst, those few cycles of 4.43Mhz (or whatever) sine wave wich comes right after a horizontal sync pulse, is there only to syncronise the TV's internal colour oscilator to the video signal. A "chroma sync" if you like. The chroma information is actually stored over the top of the luma. The best way to see this is to look at (on an oscilloscope)  the video signal of a colour bar test pattern then compare it with a greyscale bar test pattern.

The way the TV seperates the luma and chroma is with a notch filter. We know that the chroma amplitude modulated at 4.43Mhz so all bits of a composite video signal that are at that frequency, plus or minus so many Hz, must be chroma and the rest luma. This is why luma and chroma inevitably will intefere with each other - it's pretty hard to prevent luma from entering this frequency band when it is permitted above (maybe only with a direct connection, not sure about broadcast) and below it.

blackevilweredragon,

Yeah, it's all the same idea. The difference is in the weighting ie., the ratio of R, G, B which make up Y.

I've attached a couple of pages from Television Electronics by Kiver/Kaufman...

[blackevilweredragon]

Hmm..  So, if S-Video works like this, why the heck didn't the manufacturers just put Component video on their TV's?     Cheaper I guess...

I wonder how I can easily then turn Pr and Pb into chroma, because that means so many more possibilities..

[kyuusaku]

The same question could be said about S-video or composite or even 75 ohm coaxial. It just took a long time for the consumer to expect these inputs.

I personally wonder why RGsB never became the input of choice for all applications.

[blackevilweredragon]

The same question could be said about S-video or composite or even 75 ohm coaxial. It just took a long time for the consumer to expect these inputs.

I personally wonder why RGsB never became the input of choice for all applications.

Perhaps because since DVD's are encoded as Y/Cr/Cb they found it easier just to make an analog connection that was compatible.

Wasn't the reason of EDTV for DVD's anyway?  (though that failed, but HDTVs retain the 480p mode used)