I was wondering if someone could explain to me what the difference between YUV and YCbCr (or YPbPr) are?
I know that they are both derived from RGB, and I know that they are both luma-chroma-chroma. I also know that YUV is tied to PAL and that YCbCr is the modern digital standard.
However, Wikipedia has me all in a tizzy. From the article on YUV:
Then from the article on YCbCr:The YPbPr color model used in analog component video and its digital version YCbCr used in digital video are more or less derived from it, and are sometimes called Y′UV. (CB/PB and CR/PR are deviations from grey on blue–yellow and red–cyan axes, whereas U and V are blue–luminance and red–luminance differences respectively.)
They can't both be luma and then blue-difference + red-difference, or they'd be the same thing.Y′ is the luma component and CB and CR are the blue-difference and red-difference chroma components.
One of Wikipedia's sources led me to this article, which has this passage:
Also, the two Wikipedia articles give different examples of components signals for each encoding scheme.Y’CBCR scaling as defined by Rec. 601 is appropriate for component digital video. Y’CBCR chroma is almost always subsampled using one of three schemes: 4:2:2, or 4:2:0, or 4:1:1.
Y’UV scaling is properly used only as an intermediate step in the formation of composite NTSC or PAL video signals. Y’UV scaling is not appropriate when the components are kept separate. However, the Y’UV nomenclature is now used rather loosely, and sometimes – particularly in computing – it denotes any scaling of B’--Y’ and R’--Y’.
The YUV example:
And the YCbCr example:
Wikipedia does seem to offer a difference in the form of the mathematical conversion from RGB. The YUV article lists the conversion formula as:
While the YCbCr conversion is:
What gives, man!
+ Reply to Thread
Results 1 to 9 of 9
Last edited by CursedLemon; 9th Jan 2018 at 21:38.
You can for the moment skip the " ' " (prime) as that just denotes a modification that takes into account gamma or log adjustments...
Y is used for luma in all those varieties. Think of this as the Vertical (Z) axis.
The U and V can be thought of as the North/South (Y) and East/West (X) of the Horizontal axes.
They are derived functions based on differences between the pure numerical values of Green versus Red and Blue.
So, for example, U = xG - yR (where x and y are certain constant coefficients) and V = xG - zB. (I may have mistakenly swapped those names - didn't check - but it doesn't matter for the sake of example).
The U and V are generic/theoretical archtypes of those axes.
In practice, they are called Pb (the minus blue component) and Pr (the minus red component) when used in the analog domain. In the digital domain, they are Cb and Cr. They are the plus or minus values used along those axes.
In regular common usage, they can all be considered interchangeable (though it's good to be proper & consistent).
The axes mentioned above are only one way to create the color wheel/cube/diamond/etc. In old school analog NTSC world, U and V were not used, but rather I and Q (which were similar but the axes were together rotated a certain # of degrees away from U and V values in order to be more color stable as well as bandwidth efficient).
Each of R, G, and B tristimulus values could be considered luminance, but in practice it is only the gray (based primarily on green) which is truly luminance or more accurately luma component.
The others are chrominance or chroma components.
Hope that helps,
Last edited by Cornucopia; 9th Jan 2018 at 20:07.
https://www.itu.int/rec/R-REC-BT.1700-0-200502-I/en and they are:
[Attachment 44363 - Click to enlarge]
Proper equations for YCbCr are described in https://www.itu.int/rec/R-REC-BT.601-7-201103-I/en and in https://www.itu.int/rec/R-REC-BT.709/ and in https://www.itu.int/rec/R-REC-BT.2020/en
You may be also interested in https://www.itu.int/rec/R-REC-BT.1361/en
UV(IQ) are composite (with all implication related to modulation of chrominance carrier sinewave - they are used to control phase and level of carrier so at the end of colour encoding process UV are combined as one signal - this fundamentally different than in case of the CbCr or PbPr) when YCbCr (YPbPr) are component color difference signal representation.
Last edited by pandy; 10th Jan 2018 at 04:50. Reason: YPbPr is NOT YUV.
So does YUV actually have any relevance in the digital world at all? Do we use the actual RGB>YUV conversion standard for anything digital? For example, if I select "YUV2" as a compressor in VirtualDub, is that actually YUV or is it just YCbCr that is incorrectly named?
Conversely, has YUV ever been used for component analog video? Or has that always been YPbPr?
IMHO YUV2 describe sampling scheme and data organization not equation leading to color difference signal itself (YUV2 can describe signal obtained by using different than YUV or YCbCr equation for example YCgCo).
Didn't study Virtualdub code but i assume YCbCr accordingly to ITU-R BT.601 is used.
In ancient times YUV (or rather instead YUV a Y, R-Y, B-Y i.e. unscaled YUV) was used in pro broadcast equipment for example Sony Betacam. It was quite natural as Betacam stored Y and colour difference signal on tape separately. Check Sony Betacam VTR's (e.g. BVW-65, BVW-75).
Finally there is a few great papers on this written by Charles Poynton (Thank You very much Charles) - if anyone is interested then i highly recommend this lecture: http://poynton.ca/ColorFAQ.html http://poynton.ca/papers/YUV_and_luminance_harmful.html
If I can go forward in the world of digital video with the knowledge that everything labeled "YUV" is actually just YCbCr, then that'll put my mind at ease a touch about it all.
Read the Poynton paper.