Talk:Chrominance

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Media (Rated Start-class, Low-importance)
WikiProject icon This article is within the scope of WikiProject Media, a collaborative effort to improve the coverage of Media on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Low  This article has been rated as Low-importance on the project's importance scale.
 
WikiProject Color (Rated Start-class, Low-importance)
WikiProject icon This article is supported by WikiProject Color, a project that provides a central approach to color-related subjects on Wikipedia. Help us improve articles to good and 1.0 standards; visit the wikiproject page for more details.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Low  This article has been rated as Low-importance on the project's importance scale.
 

UV colour plane, at a Y value of 0.5, mapped to RGB..png

I wonder how chrominance was defined. How did they come up with this particular color plane, as you can find it with YUV? Why is red in the upper edge, and purples on the red side? There must be some logic behind that. Thanks, --Abdull 13:16, 23 Mar 2005 (UTC)

Mistake?[edit]

It says that "Chrominance is represented by the U-V color plane in a PAL and SECAM video signals, and by the I-Q color plane in NTSC."

yet under Composite video, it says that "YUV is used in the PAL and NTSC systems of television broadcasting, which is the standard in much of the world."

These two don't go together, is the YUV or I-Q color plane used in NTSC? Hope that this helps.

Probably a mistake[edit]

The YIQ system was proposed for NTSC for a while, but it fell into disuse because it was too expensive to implement.

Charles Poynton's book "Digital video and HDTV" has some excellent information on video engineering information.

Terminology: Luminance should really be called luma.

For info on the difference between luminance and luma: http://www.poynton.com/papers/YUV_and_luminance_harmful.html

Chrominance should be called chroma?[edit]

I don't think SMPTE has said anything on this, but would it sort of make sense to call it chroma? For one, it rhymes with luma.

On the other hand, chroma is an ambiguous word since it means something else in the Munsell color system. Chrominance may mean something other than video chroma if you're talking about the chrominance in Lab color space. This unfortunate ambiguity aside, the terms luma and chroma do make more sense than luma and chrominance. In my opinion. Glennchan 05:24, 22 November 2006 (UTC)

It's often called chroma. But I'm not sure what the story is on that. It was called chrominance for a long time, as luma was called luminance. Did they change both? Or just luma? Beats me. Dicklyon 06:55, 22 November 2006 (UTC)
A book by Hazeltine laboratories staff (they were part of the original NTSC committee) uses the terms luminance. However, they recognized the difference between "luma" and luminance (the luma video page, which you've read, describes this difference). I think they use the term chrominance, although I may be wrong. I think they also used the symbol "Y' ", but again I can't remember.Glennchan 05:54, 24 November 2006 (UTC)

Chrominance not the same thing as the saturation described on the wikipedia saturation page[edit]

Article- see the difference between chrominance and chromaticity Glennchan 06:06, 10 December 2006 (UTC)

Your link is bad. But your point is true. However, though they are different in detail and not 1:1 related, they are both luminance (and hue) of a sort. Dicklyon 06:32, 10 December 2006 (UTC)

21:37, 6 August 2008 (UTC)21:37, 6 August 2008 (UTC) I believe Chrominance was coined simply to make a distinction and obviously is the combination of the words luminance and chromaticity. There is a simple rationale for this, since Chrominance is actually a brightness/luminance/magnitude value of a limited spectra of colours, i.e. ignoring any twiddling coeffs, R-Y = G+B, & similarly, B-Y =R+G, whereas chroma(ticity) is a proportional value that does not contain any information regarding brightness. i.e. again ignoring twiddling coeffs (R-Y)/Y and (B-Y)/Y. Indeed it is therefore theoretically possible to define the colour of a black picture!.

The latter fits the original ideal of colour TV system, long ago, where the extant B&W transmission could simply have an additional signal glued on that would define the colour character of an area of the picture. The human eye could tolerate a significant reduction in the detail resolved in this, by bandwidth reduction, and would make efficient use of its transmission channel in terms of its dynamic range. However, this would have been expensive in terms of electronic circuits of the day, a division and multiplication being required, as well as problems pertaining to the transmission channel itself (of high levels of chroma signal in dark areas of picture would be difficult with composite in band subcarrier coding).

A more practical approach was simply to use the magnitude signals R-Y and B-Y i.e. Chrominance, since addition and subtraction is relatively simple. Filtering of this to achieve bandwidth reduction is possible but will introduce transient errors that are conveniently ignored, (continues to be done with the processing technology we have today) Witness the black line down the middle of the colour bars between green & magenta, where the colour channel has to make a maximal change. This approach better fits in with, particularly, the NTSC/PAL systems, that effectively demand the correlation of signal levels in the colour subcarrier to main luminance channel to mitigate crosstalk. and is hence less efficient in its use of a transmission channel. (The situation with SECAM V with FM carriers is a little more complex!) See article by a Dr E L C White in Wireless World 1958 regarding “Non Constant Chromatictity and Luminance” (Check)21:37, 6 August 2008 (UTC)21:37, 6 August 2008 (UTC)21:37, 6 August 2008 (UTC)21:37, 6 August 2008 (UTC)93.97.145.169 (talk)

Possible inaccuracies?[edit]

often referred to as Cb and Cr or as U and V components

this might be a little misleading, since there are multiple schemes in use, all of them different. Y'UV, Y'PbPr, Y'CbCr, etc. The "often referred to" suggests that the schemes are the same.Glennchan 05:46, 11 December 2006 (UTC)

Good point. Cb and Cr are fairly generic terms used in a lot of digital colorspace representations; the term may even include reference to U and V, but I'm not sure. I'm not familiar with Pb and Pr. Dicklyon 06:16, 11 December 2006 (UTC)
Cb and Cr aren't really generic terms. They typically refer to Y'CbCr encoding, although there are variations on that format. For 8-bit studio video, luma ranges from 16-235 and chroma from 16-240. 10-bit is the same thing, but with everything mutliplied by 4. Full scale Y'CbCr (used in things like JPEG) have everything ranging from 0-255. Y'CbCr can also use the Rec. 601 or Rec. 709 luma co-efficients.
That's what I mean by generic; lots of variations in usage. For example, JPEG's Y'CbCr can be computed from arbitrary different R'G'B' spaces, so the space has no colorimetric definition of its own. Dicklyon 05:19, 12 December 2006 (UTC)
This is an issue of semantics, but I don't think that Y'CbCr should be considered a generic term. There are simply variations on the scheme. Some authors use notation to specify which form of Y'CbCr encoding they are talking about- i.e. whether it's JPEG/full scale, Rec. 601, Rec. 709, etc. Y'CbCr does not refer to other schemes like Y'PbPr or anything like that.Glennchan 04:57, 13 December 2006 (UTC)
Y'UV is for analog formats. U and V are scaled such that the excursion of the composite signal (Y' + C) is constrained to the range -1/3 to +4/3 of the unity excursion of luma. Glennchan 03:52, 12 December 2006 (UTC) Dicklyon 05:19, 12 December 2006 (UTC)
  • Not all video systems used luma + chroma. There are some R'G'B' interfaces, like dual-link HD-SDI. I think there is component R'G'B', but I'm not sure. Glennchan 05:46, 11 December 2006 (UTC)
True. Component video doesn't use luma/chroma. I don't think the opening sentence implies ALL, but you could change it to be more clear if that's your point. Dicklyon 06:16, 11 December 2006 (UTC)
I think most component video uses Y'PbPr, which is luma+chroma (not R'G'B'). Glennchan 03:52, 12 December 2006 (UTC)
Really? I thought component video meant RGB components. And then there's s-video, where luma and chrominance are on separate wires, but the chrominance is still on a color carrier; is that called component video? Does Y'PbPr use three cables? OK, I justed checked component video and sure enough these are all multi-wire and hence component video. But Pb and Pr are just rescaled Cb and Cr or U and V or I and Q; conceptually these are all the same, and the Cb Cr pair are the ones used more generically than the others, I think. Dicklyon 05:19, 12 December 2006 (UTC)
The component video article might be inaccurate??? I don't believe in S-video a color subcarrier is required; the color subcarrier is necessary on a single cable, since you have to somehow put luma + chroma onto that one wire. In S-video, Y' and C are separated. I don't think that most people refer to S-video as component; however, I'm not sure on that point. Y'PbPr does use three cables. Pb and Pr are analog; Cb and Cr are digital- so I think the distinction is important. Glennchan 04:53, 13 December 2006 (UTC)
Well, I don't know if it's generally called component video or not, but it does require a carrier to put both chrominance dimensions on one wire. Dicklyon 06:00, 13 December 2006 (UTC)
Oh wait you're probably right about that. A look in Poynton's book does show that Y'/C does have subcarrier.Glennchan 06:04, 13 December 2006 (UTC)

Color killer[edit]

Monocolor broadcasting may be the technology of the past. But still, the color killer should also be mentioned while discussing the color burst. Nedim Ardoğa (talk) 08:22, 5 February 2010 (UTC)