Talk:Hue

From Wikipedia, the free encyclopedia

You have new messages (last change).
This article is supported by the Color WikiProject.

This project provides a central approach to Color-related subjects on Wikipedia.
Please participate by editing the article, and help us assess and improve articles to good and 1.0 standards, or visit the wikiproject page for more details.
Stub This article has been rated as stub-Class on the quality scale.
High This article has been rated as high-importance on the importance scale.

Article Grading:
The article has been rated for quality and/or importance but has no comments yet. If appropriate, please review the article and then leave comments here to identify the strengths and weaknesses of the article and what work it will need.
Please add new discussion topics at the bottom

Contents

[edit] Old stuff not in any section

I think it should be stated that humans (usually if not Tetrachromats) only can "see" the mixture of the colors red, green and blue, and the definition and calculation of hue is based on this fact.

Also: I think the given values are pretty much wrong:

R = \mu + \sigma \sqrt{2} \cos \phi,
G = \mu + \sigma \sqrt{2} \cos \left( \phi + {4 \pi \over 3} \right),
B = \mu + \sigma \sqrt{2} \cos \left( \phi + {2 \pi \over 3} \right).

I would say, this is correct:

R = \mu + \sigma \sqrt{2} \cos \phi,
G = \mu + \sigma \sqrt{2} \cos \left( \phi + {2 \pi \over 3} \right),
B = \mu + \sigma \sqrt{2} \cos \left( \phi + {4 \pi \over 3} \right).

(flipped 2 and 4)

HSV color space as a conical object
HSV color space as a conical object

Just take a look at the pictures on the page HSV color space: Hue cycles from red to green to blue.

--Abdull 17:29, 20 Mar 2005 (UTC)

I think there's something more wrong than that with these definitions: both the referenced definitions of brightness and saturation are symmetric in ( R, G, B ) and the definition of hue breaks the symmetry only with respect to R. Hence a transform swapping G and B produces unchanged values of hue, saturation and brightness, which is clearly wrong.

Pak21 15:50, 23 July 2005 (UTC)

What is "gradation" mean in this context?

Also, can someone please fix the formulas or remove them, since they are clearly wrong. dave 18:37, 5 August 2005 (UTC)

I'm not yet convinced the RGB formulae are wrong as opposed to the referenced definitions of brightness and saturation. The definitions of both brightness and saturation are symmetric in RGB. One extra variable (hue) can only introduce one degree of freedom, so we we've got only two degrees of freedom in this system, which isn't enough to reproduce RGB. Much more than this will need someone with more knowledge of the subject than me.
--Pak21 14:17, 8 August 2005 (UTC)

[edit] I think the "hue" equation is much too simple

in rgb, colors of constant hue are not on a single circle: red, yellow, and green are all on the same face of the color cube. Consider variable names red = r, yellow = y, green = g, cyan = c, blue = b, and magenta = m. the color gray shall be G. The colors r, y, g, c, b, m have 100% saturation, but they are not on the same plane in rgb space. I think that the six colors are flatened to a haxagon, with grey in the center. hue is computed specially for each triangle (G,r,y), (G, y, b) etc.

COnsider the equations on the HSL space cited above: they use 6 spacial cases to convert from hue to rgb —The preceding unsigned comment was added by 149.169.178.128 (talk • contribs) .

See my comments above; I believe there's something fundamentally wrong with the definitions we're using here, which goes beyond problems with just the hue equation. Don't have a clue what it is, though :-( Cheers --Pak21 15:45, 29 April 2006 (UTC)

[edit] Equation mismatch

The equations listed in the Hue article do not match those in HSV color space. Discussion about this is occurring at Talk:HSV_color_space#Mismatch_with_Hue. --Orborde 21:20, 4 July 2006 (UTC)

[edit] Sources needed

Can anyone point me to a web site describing the sinusoidal HSV->RGB transformations listed on this page? --Orborde 21:22, 4 July 2006 (UTC)

[edit] Formula mismatch

There is a major error in Hue formula \phi=\arccos\left( \frac{R-\mu}{\sigma\sqrt2} \right). The thing is the image of arccos function is an interval [0,π], so for pure green and pure blue we'll be given by the same value. Moreover you can never obtain φ = 240° [4π/3c] using such an equation. The formula should consider a ratios between R, G and B values (you should consider which value is the largest to obtain an addition for value obtained by given formula).

—The preceding unsigned comment was added by Blpell (talk • contribs) 08:32, 22 September 2006 (UTC)

You're right. The formula is wrong. The conversions from hue, saturation and value to RGB is also quite strange. The formulas here look better but are not completely correct (see my comment here). I've added the disputed template on the article, so nobody uses the formula without noticing that something could be wrong here. I could develop the right formulas, but wikipedia is not the place for original research, so we need sources first. I couldn't find anything with google but maybe I just hadn't the right keywords for it. (note: please sign your posts with --~~~~, it will automatically turn into a signature like the following one)

--Oxygene123 16:43, 15 October 2006 (UTC)

[edit] Lovibond's comments moved from top and put in order

Hue is a perceptual aspect of color; this should be covered in an introduction. Further, only one type of hue, that in the HSB/HLS encodings of RGB, are covered. There are many other mathematical definitions, some more useful.

Much of the discussion regarding a "spectrum" is potentially confusing, as Magentas and Purples are not part of the spectrum of visible light. Lovibond 02:06, 19 December 2006 (UTC)

Performed an edit which addresses some concerns, but not all. Citations added. Still need to work out correct formulae used in HSB and HLS color spaces. Is article ready to be upgraded to "B" or "G"? Lovibond 03:25, 16 January 2007 (UTC)

Removed references to saturation as one of the main perceptual attributes of color, per CIE Publication 15.2 (and later); Chroma is not Munsell-specific (ibid), so intimations to this effect were deleted. Lovibond 19:01, 30 January 2007 (UTC)

Please respond to my note below. It appears to me that Hue/Saturation is a more common system of chromaticity than Hue/Chroma; CIE Pub 15 is not the only game in town. Can we expand the viewpoint a bit? Dicklyon 02:59, 31 January 2007 (UTC)

Agreed that CIE Publication 15 is not the only thing which has something to say on the matter, but the definition of saturation in HSB/HVS is essentially in accord with CIE's definition (colorfulness relative to lightness). In HSB, saturation is, in essence, the complement of the quotient of the smallest of R, G, or B to the largest of the three. Mathematically, its easily to define, but it is not fundamentally perceptual. Take an orange (0xFF9900) and a brown (0x553300). Both have the same HSB hue, both have the same HSB saturation, as both are fully saturated. Yet one looks more colorful than the other. Show people patches of each side by side and ask which they think is a more saturated color and I am confident most will pick the orange -- overwhelmingly so. Saturation, as defined both by CIE and in the HSB/HLS encodings of RGB, is decidedly not fundamentally perceptual.

I'm not suggesting that HSB or HLS be scrapped; I am pointing out that the saturation coordinate is not fundamentally perceptual, so saturation should not be enumerated as one of the three principal perceptual attributes of color. Remember, Chroma and Saturation are different. Saturation is Chroma relative to lightness. It is even defined as such in both the CIE 1976 L*, u*, v* (CIELUV) color space (with a scaling by 13), as well in the newer CIECAM02 color appearance space. Chroma is perceptual difference from the closest gray; as such it is computed as distance in perceptually (relatively) uniform color spaces; saturation is physical rather than perceptual and is a distance on a chromaticity diagram.

Lovibond 23:12, 31 January 2007 (UTC)

I see. I had missed that point, that chroma is more perceptual and saturation more physical, and that our lead sentence mentions perception: "Hue is one of the three main attributes of perceived color". However, since Hue is also one of three dimensions of many physical color measurement systems, where saturation is another, doesn't this lead then need to be expanded to cover those as well? Dicklyon 04:02, 1 February 2007 (UTC)

Which physical color measurement systems use Hue? HLS and HSB aren't really color measurement systems, but re-encodings of RGB. They use coordinates which are more intuitive than RGB, but are also designed to have simple relationships with RGB. "Lightness" in HSL and "Brightness" in HSB are neither according to their commonly accepted definitions in the color community (see, e.g., Mark D Fairchild, Color Appearance Models, Wiley, 2005)

I have removed once again the inclusion of "saturation" as a "main perceptual attribute of color," as it clearly is not.

129.21.57.211 23:54, 14 February 2007 (UTC)

OK; I added another lead sentence for the use of Hue in HSV etc. color spaces. Dicklyon 02:30, 15 February 2007 (UTC)

[edit] Chroma

Lovibond, thanks for pointing out that my conceptions of chroma were out of sync with CIE Pub 15. Where can I find a copy of that doc? The problem in the article now, which prompted me to hack it in the first place, is that chroma links to a disambig page where none of the choices support the use in the article. The Munsell chroma was the closest thing, and I was more used to the video chroma, a 2D concept more like chromaticity. So, we probably need an article on CIE chroma, or not link the chroma article, or something. Dicklyon 17:44, 28 January 2007 (UTC)

I should check those links -- thanks!

Lovibond 23:12, 31 January 2007 (UTC)

[edit] Flushed bad equations

Following up on several years of complaints about the wrong equations, I removed them and stubbed out a section based on what I could find in a book. The book has code, not equations, but perhaps we should construct the equations. Or maybe they're what we have already in one of the other articles. Someone want to help? Dicklyon 00:03, 18 February 2007 (UTC)

The computations for hue for the Preucil Color Circle may be the same as those used in computer graphics. I'm not a computer graphics person, so I can't say for sure. Would someone be so kind as to check this? Lovibond 02:10, 22 March 2007 (UTC)
Retrieved from "http://en.wikipedia.org../../../h/u/e/Talk%7EHue_ea35.html"