RGB + Yellow? New Sharp Television "Coming Soon"

[DMark]

I am sure that. at least in the US, people have started seeing the new ad campaign for Sharp's new television - claiming a new color combination of RGB and Yellow.

Supposedly, the addition of yellow will make it possible to add colors (such as brass, gold and other similar metal/colors) that make things look better.

I am a bit dubious - but have yet to see what it looks like. Then again, the "coming soon" part of the ad seems to imply I cannot run to my local electronics store and see what it looks like to prove or disprove their claims.

This is the only link I could find from Sharp, but doesn't go into much detail.

Has anyone actually seen this television, and does the addition of Yellow really add to the quality of color on the screen?

[fuzzypickles]

Sounds like marketing bullshit. The human retina has only green, red, and blue receptors (cone cells) and current technology is sufficient to reproduce all 32,000,000 shades of color which the human eye can detect.

Maybe adding a yellow pixel makes the golds more golden, or the brasses more brassier, but on the surface it sounds like pure snake oil.

[Ellis Dee]

I'm highly dubious. Current technology can produce more colors than the eye can differentiate.

Though I'm glad to see George Takai getting work.

[GSV Consolation of Dreams]

Originally Posted by DMark

I am sure that. at least in the US, people have started seeing the new ad campaign for Sharp's new television - claiming a new color combination of RGB and Yellow.

Supposedly, the addition of yellow will make it possible to add colors (such as brass, gold and other similar metal/colors) that make things look better.

I don't get it... I have seen gold, silver and all sorts of metallic colours on regular old RGB. They look realistic enough to me.

[CalMeacham]

The colors you can see on a color TV are found in a polygon on the CIE diagram bounded by the sources. For a standard RGB screen, that means you see everything inside a triangle bounded by the chroimaticity coordinates of the three phosphors. These don't give all the colors you can see with natural white light sources, but it gives you most of the visible range.

In principle, you can extend this range by adding more sources. If they add a yellow source, that triangle can turn into a quadrilateral, and you'll be able to see a sliver more of real colors.

http://en.wikipedia.org/wiki/CIE_1931_color_space

This diagram shows the real color space, and the embedded triangle shows the space achievable with three phosphors:

http://en.wikipedia.org/wiki/File:CIE1931xy_CIERGB.svg


Yoellow would be off to the side around 570, and it doesn't look as if it would add much, based on the diagram. You'd think they add a new phosphor around, say, 520, and get that big chunk in the upper left. I should point out, however, that the diagram doesn't give you equal portions of energy or of visibility. Your eye is more sensitive in the green, and yellow is pretty close to that. Yellow might be the optimum choice in expanding obviously visible color values.



The biggest problem I see is that your TV signal only broadcasts for the RGB signal (and, I believe, black and white) -- it doesn't have the information needed to accurately add the portions for a fourth color. The only think I can see is that they're somehow interpolating the information, generating it from the broadcast tristimulus color. That means not merely adding some yellow to the mix, but also in changing the relative proportions of Red, Green, and Blue as well. I can see this going awry occasionally, and giving some goofy color mixes.


by the way, the fourth signal isn't "overkill" because we only have three receptors -- nature broadcasts in essentially infinite sources. The three phosphors don't coincide with the sensitivity of our three receptors, so it's not as if there's a one-to-one correspondence. Anything that gets us closer to reality is arguably a better picture. But, until everyone is broadcasting signals corresponding to the output of the TV, the system is going to be something of a kludge.

[Turek]

Originally Posted by Ellis Dee

I'm highly dubious. Current technology can produce more colors than the eye can differentiate.

Though I'm glad to see George Takai getting work.

He's the best part of that commercial. The "Oh my!" is great. Plus, he's giving a little upside down vulcan salute with his right hand!

[Lethal Babydoll]

Originally Posted by CalMeacham

The colors you can see on a color TV are found in a polygon on the CIE diagram bounded by the sources. For a standard RGB screen, that means you see everything inside a triangle bounded by the chroimaticity coordinates of the three phosphors. These don't give all the colors you can see with natural white light sources, but it gives you most of the visible range.

I concur.

[The Second Stone]

Originally Posted by Turek

He's the best part of that commercial. The "Oh my!" is great. Plus, he's giving a little upside down vulcan salute with his right hand!

I get a picture in my head of the filming of this commercial with the director and Takai having an argument: "George, we need you to gay it up a bit for the Star Trek fans." G: "If it gets any gayer, you'd have to add a pink receptor!"

Thanks for the good explanation Cal.

[thelurkinghorror]

That CIE diagram is an ideal RGB color space though. On most TVs the gamut will probably be smaller and be a different shape depending on brand.

I don't know how Sharp works their technology, but I'd almost guess that the new yellow phosphor is close to where the current green location is, and the "new" green is somewhere in the green/cyan range in the upper left. This would expand the gamut to a quadrangle.

[zwede]

I wonder how that fits into the old DLP TVs that used color wheels. My 6 year old Samsung DLP TV uses a color wheel with the following colors: Red, Blue, Green, Magenta, Yellow, Cyan and White (clear). So does that mean DLP TVs had superior color reproduction?

[Amblydoper]

Originally Posted by fuzzypickles, bolding mine

Sounds like marketing bullshit. The human retina has only green, red, and blue receptors (cone cells) and current technology is sufficient to reproduce all 32,000,000 shades of color which the human eye can detect....

This is not correct.

The human eye does indeed have three types of color receptors, but they do not absorb only certain colors, like RGB, and mix them like a television does to display an image. the three types of cones are each sensitive to a wide range of wavelengths. This image from Wikipedia shows the response from each cone at different wavelengths.

Because the human eye can not resolve the resolution of individual pixels from a television set, the addition of a yellow pixel will not change how our eyes interpret the colors, but will only increase the range of colors, as Cal has so excellently described.

[Annie-Xmas]

Originally Posted by Turek

He's the best part of that commercial. The "Oh my!" is great. Plus, he's giving a little upside down vulcan salute with his right hand!

He and the fish look remarkably alike.

[fiddlesticks]

These have four colors. -- Nigel Tufnel

[LeftyRodriguez]

Am I the only one that's bothered by the fact that they got an Asian actor to introduce the YELLOW feature to their TVs?

[ministryman]

Originally Posted by LeftyRodriguez

Am I the only one that's bothered by the fact that they got an Asian actor to introduce the YELLOW feature to their TVs?

I was....but I was laughing so hard at the "Whoooooo" and the "Oh My" to care.....

[Vinyl Turnip]

I was impressed. The image looked amazing on my RGB set!

[BigT]

Originally Posted by LeftyRodriguez

Am I the only one that's bothered by the fact that they got an Asian actor to introduce the YELLOW feature to their TVs?

It upsets me about as much as it upset me when I figured out why Trini was the yellow Power Ranger.

[Rhythmdvl]

There are four LEDs!

Wait, that was the wrong generation.

[Powers]

Originally Posted by BigT

It upsets me about as much as it upset me when I figured out why Trini was the yellow Power Ranger.

It's because the yellow ranger was a male in the original Japanese production, the costumed footage of which was reused for the American version. They wanted to add another female (besides pink) to the lineup for the American version, so they picked the smallest (and lightest-colored) of the remaining rangers and made "him" a girl.

At least, that's what I heard.


Powers &8^]

[GargoyleWB]

Originally Posted by thelurkinghorror

That CIE diagram is an ideal RGB color space though. On most TVs the gamut will probably be smaller and be a different shape depending on brand.

I don't know how Sharp works their technology, but I'd almost guess that the new yellow phosphor is close to where the current green location is, and the "new" green is somewhere in the green/cyan range in the upper left. This would expand the gamut to a quadrangle.

Lighting engineer here...

Agreed, the "sweet spots" for enlarging the color space are not the pure Red/Green/Blue/yellow points. The goal is to maximize the area of the enclosed space, and if one could develop a pure aqua or teal LED or phosphor, then that would have far greater effect than adding a yellow corner would. I'll also drop this additional link for the Macadam ellipses, which overlays ranges where a shift in color becomes human-perceptible.

The current color triangle works pretty well, though. The only colors left are the 1950's melamine and designer naugahyde furniture palette, which are probably best left unseen to the naked eye

[Revtim]

Originally Posted by LeftyRodriguez

Am I the only one that's bothered by the fact that they got an Asian actor to introduce the YELLOW feature to their TVs?

I'm not bothered by it, but I'm gonna think of that now every time I see that commercial.

[Ackerman02]

Sharp USA is a subsidiary of an Osaka based parent company. I didn't think skin tones for Japanese were that far from Caucasian. Now, if they ever add a magenta pixel and could get Patricia Quinn to do the commercial ... that would be something - and she's younger that Takei.

What I want to know is how they determine the driver function for the yellow pixels, since there is no specific signal for yellow. They must be deriving it from the RG portion of the data, but I don't see how that would significantly improve the image, except that you can now directly display the yellows without having to mix the color.

I can see getting an marginal increase in the yellows intensity, but it seems like a bit of a reach unless they are going to flesh out the set. I would assume that the next logical step would be to add the other two secondary colors in future TVs. Now, if someone will please develop a silver LED, we could really get metals to look real.

[Chronos]

Silver isn't a color, though. When you see something silver, you recognize it as silvery because of the way it reflects the colors in its vicinity. As it is, if we see something silver on TV, it's reflecting the colors of the other things on the set, just as we would expect. If, however, you made some sort of pixel that actually turned silvery, it would reflect the colors in your living room, and thus look out of place with the rest of the scene.