RGB, YCbCr and RED

[Eugene]

I read about Apple's Color program in MacWorld. Color doesn't edit in YUV, only RGB. Now I read that Red Alert will edit in RGB. Red will record in RGB also. I have a dumb question for Gramme. I don't know much about RGB or YCbCr, but is RGB better and should I be worried that Apple's Color and Red Alert don't support YCbCr. Is YCbCr obsolete, or will Red and Apple support YCbCr in the future as far as 2K and 4k are concerned. Don't be harsh on my lack of knowledge.

 

[Evan Owen]

Is YCbCr obsolete, or will Red and Apple support YCbCr in the future as far as 2K and 4k are concerned. Don't be harsh on my lack of knowledge.

No problem, it can be very confusing sometimes. YCbCr is the digital version of YUV, and yes, both are now essentially obsolete. RGB (if uncompressed) contains all of the original recorded color information. YUV, at least the NTSC version, doesn't.

Just found this on Wikipedia:

Another advantage of YUV is that some of the information can be discarded in order to reduce bandwidth. The human eye has fairly little color sensitivity: the accuracy of the brightness information of the luminance channel has far more impact on the image discerned than that of the other two. Understanding this human shortcoming, standards such as NTSC reduce the amount of data consumed by the chrominance channels considerably, leaving the eye to extrapolate much of the color. NTSC saves only 11% of the original blue and 30% of the red. The green information is usually preserved in the Y channel. Therefore, the resulting U and V signals can be substantially compressed.

However, this color space conversion is lossy. When the NTSC standard was created in the 1950s this was not a real concern since the quality of the image was limited by the monitor equipment, not the compressed signal being received. However today's modern television is capable of displaying more information than is contained in these lossy signals. To keep pace with the abilities of new technology, attempts have been made to preserve more of the YUV signal while recording images, such as S-Video on VCRs.

[EDIT] According to the article, another advantage of the YUV colorspace was backward compatibility with black and white TVs. The Y luminance channel could be displayed and the U and V chrominance channels simply discarded. :umm:

 

[Graeme Nattress]

YUV is bad terminology. YUV is a step in the process of converting composite analogue video to composite.

The correct terminology for a luma / chroma component system is YCbCr for digital video and YPbPr for analogue video.

Normal YCbCr is associate with chroma sampling such as 4:2:2, 4:2:0, 3:1:1 etc.

As well as seperating luma from chroma, YCbCr also de-correlates the signal making it more compressable. Even if you don't use chroma subsampling, YCbCr and other similar transforms can be used to get more compression efficiency, even in lossless codecs.

Given enough precision, you can transform between RGB and YCbCr and YCbCr and RGB all you like without bother. They're just two different ways of looking at the same data.

We use RGB and YCbCr with our codec - some apps want RGB, so we give them RGB and some want YCbCr, so that's what we give them.

So, YCbCr itself is not obsolete, but the associated chroma sampling probably is at the end of it's natural life span.

Graeme

 

[Evan Owen]

... We use RGB and YCbCr with our codec - some apps want RGB, so we give them RGB and some want YCbCr, so that's what we give them.

So, YCbCr itself is not obsolete, but the associated chroma sampling probably is at the end of it's natural life span.

Graeme

Thanks Graeme, I probably should've just waited for your response to this. It makes even more sense now.

Back to part of his question though, is it generally true that editing in the RGB colorspace is more desirable (accurate, etc.)?

 

[Graeme Nattress]

Well, there are as many RGB colour spaces as you've had hot dinners.

You're probably going to be moving the data around a number of spaces, so it's the accuracy of that conversion that matters more than anything.

Graeme

 

[vanguy]

Graeme, I was under the impression that YUV, R-Y/B-Y/Y, and Y/pB/pR were the same thing, and have actually hooked them up and gotten a reasonable image.

U=B-Y=pB
V=R-Y=pR
(I think)

Am I wrong? Or am I almost right? It seems to work...

 

[Graeme Nattress]

Nope, they're not the same thing. They are often the same just because people name things wrongly, but that's their error.

YUV is very bad terminology as it's hardly ever seen. I guess people can't say YCbCr so they say YUV as it's easier. Unless you're converting analogue component to analogue composite video for broadcast, you won't have anything to do with YUV.

Cb / Pb is similar to B-Y, but not quite. B-Y is what you're doing to make Cb, but Cb is more complex than that as you have to scale it to get the levels right, so B-Y is not strictly correct. Similarly for Cr / Pr.

Graeme