"If blue, red, and yellow are primary colors, why do color TVs use blue, red, and green?"

A very recent thread on the same topic. Timewinder in this thread and CalMeacham in the other give similar responses that give the best explanation I’ve heard for how it might work, although Sharp is rather quiet on what the technology exactly does.

Unfortunately, even if you had a photoreceptor that reaches down into UV enough, your lens block a lot of UV light. There is a solution, though! Just get some cataract surgery, and you may be able to see into the UV range. Just be aware of all the health complications and vision impairments which result.

That’s true about opponent processing, but I don’t know if it’s completely relevant to the TV. The blue-yellow channel compares the output of the blue cone to the sum of the red and green [S - (M+L)]. As you mentioned upthread, yellow is red+green, thus seeing blue is lots of blue light, little red or green, and seeing yellow is lots of red and green*, little blue.

In other words, the yellow LCD phosphor wouldn’t add anything that the red+green would. In addition, the “blue-yellow” opponent channel really responds best to a purple vs. a pukey-green color. The yellow pixels wouldn’t stimulate this necessarily.

*Technically, yellow is little red OR green. If the red-green L-M channel is balanced instead of biased, then that proceeds to a yellow percept instead of red or green.

Except that the response curves of the three types of human cone are not at all neat little parabolas or sines or normal-distribution. They’re asymmetrical and not even all single-peaked. So, for all I know, this RGBY with such-and-such LEDs may match actual human perception better than RGB with so-and-so LEDs. Until someone does the actual math with the actual numbers, we’re groping in the dark.

And, once again, this is not even taking into account tetrachromatism.

You linked to this thread. :slight_smile:

Told you they were similar. Maybe this works better. At least it didn’t point to embarrassing porn.

Interesting concept. I’ve always known that our color perception was less keen than our light-dark reception, and wondered if any technology took advantage of that in compression.

John K. hit on the answer. Everything we use to produce color images, print, video or film, has a limited color gamut. In other words, it can only produce a subset of all possible colors we can perceive. If you plotted all producible colors for a given video monitor, it would be an odd shaped blob in the middle of what humans can perceive. As John stated, this is due to the fact that the phosphors and dies used in video monitors only approximate the red, blue, and green of human visual perception.

The red and green colors in video monitors are slightly deficient around the color yellow. Simply modifying the red and green colors produced by the monitor wouldn’t work because they are based on the colors perceived by the video camera. Changing these colors would most likely make the color gamut worse. Instead, they synthesize a yellow signal from the red and green to enhance the yellow gamut of the monitor using a yellow pixel color. It should, in theory, improve the color gamut of the monitor.

How well it works, and what it’s worth is another story.

Now that we have digital video, adding additional color channels to improve the color gamut of our cameras and monitors is the next step. I don’t know too much about how they encode the new digital video, but if they went to an RGB encoding they made a mistake. If images are encoded using chrominance(color) and luminance(brightness), then any number of color channels can be synthesized from the signal.

I believe the system (which, remember, was established something like 20 years ago) is RGB, but as long as the cameras are RGB, shifting to any other system can only lose data. Chrominance/luminance was used on NTSC for the sake of compatibility with the original 1930s all-luminance technology.

Well, if you consider a triangle an odd shape…

=)
Powers &8^]

Or, maybe RGBY looks better in Sharp advertising:

“We have 25% more color pixels than any other TV!”

I’m afraid we’ll see this multiply now until we have ROYGBIV TVs. It’s sort of like the megapixel war in digital cameras and the number of razor blades in razors.

I remember when Gillette first came out with the two blade razor, Saturday Night Live did a parody on a three blade razor. Ha ha! That was so funny! Of course it’s ridiculous. Who’s going to put THREE blades in a razor? I think we’re now up to five.

Five? We’re up to six, now. They even come with shaving cream in the handle! (Here’s a review, I’m at a loss for words.)

And yet again I have to point out that all this “Yes it is”, “No it isn’t”, isn’t achieving one damned thing. We have presented arguments showing that it is not a simple given that RGB cannot be improved upon by more complex systems. Now, either the actual math of it or an actual side-by-side test is necessary to move forward here.

Same thing with camera pixels, by the way. Take it back to before digital cameras: do you think professional photographers continued to use medium- and large-format cameras when they could have used half-frame 35mm, instead, just for fun? An 8x10" photograph printed by a typical modern inkjet printer has nearly 500 megapixels.

While we’re at it, we can’t perceive fine detail as well in the blue channel as in the other two, which is why yellow text on a white background, or blue on black, is hard to read. This, too, is exploited in many compressed formats, such as JPEG and things based on it.

I wonder if this explanation helps. It’s nothing official, but seems to be what Sharp are trying to do.

Just wanted to comment that just because something could have theoretical benefit, does not mean every implementation is worthwhile. Sure large and medium format have advantages over 35mm and 10 megapixels can give a better picture than 5, but putting 10 megapixels in a tiny point and shoot with 1/4" lens is pure marketing. You aren’t really getting enough light to make the image projected on the CCD sharp enough to take advantage of those pixels. And even if this RGBY has a more complete gamut, will the picture be better or worse?

The amount of light reaching the film or sensor is a function of the f-stop, which is not and cannot be measured in quarter inches.

As far as I can tell, the answer to that is entirely dependent on the algorithm used to calculate the Y channel, since it doesn’t exist in the original source.
Powers &8^]

I think this has been brought up before on SD, but basically if your lens were transparent to UV light, or you had your lens removed, you could see UV light but it would appear white as UV can trigger all our photoreceptors.

So, alas, it’s not like getting a fourth primary colour.

It sucks…I wanna see Avatar II in mantis-shrimp-o-vision…

This question doens’t have a straight answer because it is based on a faulty premise. The faulty premise is that, “The primary colours are Blue, Red and Yellow”.

It may come as a surprise to you that these are NOT the primary colours but when you are finished reading this, I hope you will agree.

The question posed is not the only place where the faulty premise is adopted. The same faulty premise is adopted by some of the other who have posted in this forum.

The explanation for why the faulty premise is so commonly adopted, I believe, may go beyond the following, but for the moment I want to address what appears to be a common confusion among the colours in the two most common colour pallettes that are the “Primary” colours. Those pallettes are, RGB and CMY (aka CMYK which is just CMY with Black added to increase contrast because the CMY pigments are not perfectly manufatured).

As has been noted here, the RGB colours are “additive” and since televisions are luminescent (unlike your Spiderman comic book), your television relies on RGB.

Your Spiderman comic, on the other hand, book reflects ambient light (sunlight – the full spectrum) which means that Spidey relies on CMY instead of RGB.

I think there is little surprising here, so far. However, here is something that I think will be a surprise to some people. Certain colours in these two systems are sometimes spoken of as if they are interchangeable! They are not! That is, Cyan (subtractive) is confused with Blue (additive) and Magenta (Subtractive) is confused with Red (Additive). Nobody confuses Yellow (Substractive) with Green (additive).

The failure to distinguish C/B and R/M is understandable, but this error is not merely a failure to distinguish two colours that have a similar appearance (which is understandable) – but more significantly, it is also a failure to understand the distinction between Additive and Subtractive colours.

So let’s look at the distinction between the Additive and Subtractive Primary colour palettes.

First and foremost, think of the Additive colours as a light source (like a television). The Additive colours work quite well and at their best when we are viewing them in the dark – ie in the absense of any other ambient light. Meanwhile, think of the Subtractive colours as light absorbing pigments which of course will work only where there is ambient full spectrum white light (like sunlight).

It is only in ambient full spectrum light that the subtractive pigments in the CMY system can function. Afterall, the CMY pigments can function only if there is something from which they can subtract colours.

Consider how Cyan pigment functions: You add Cyan pigmented ink to a page and allow white light to strike it. Not all of the white light is refelcted by the Cyan pigment – only Cyan coloured light is reflected. Now point a video camera at this Cyan light – Of the 3 (RGB) pixels that can be activated in the viewfinder, two of them will be activated and the third one will remain dark. Can you guess which are which? (I’m probably about to make a mistake – somebody correct me in a follow up post)… B+G are activated while R remains dark. Another way to think of the light reflected by Cyan pigment is Blue light and Green light – while Red light is absorbed by the pigment (“absorbed”=destroyed in this context – converted to heat). This is why CMK is called “Subtractive” – C is not so much Cyan but rather it is Red-absorbing (and BG-reflecting)

So please do not think of Cyan as “Blue” – rather, think of Cyan as “Red-subtracting pigment” (equivalently, “Cyan pigment is Blue and green reflecting”)

Okay, now what if you discard the Cyan Pigmented sample and let’s get a Magenta sample and subject it to ambient white light. Again, the white light strikes the pigment and some of it is reflected. Think of reflected and absorbed colours in RGB terms as in the video camera example – can you guess which of RGB is reflected and which is absorbed? Before you guess, I’ll point out something you might already know perfectly well: two of RGB are reflected and only 1 of RGB is absorbed. So which two are reflected? Which one is absorbed? Answer: R and B are reflected while G is absorbed (ie Green is effectively destroyed – it is “subtracted”).

So please do not think of Magenta as “Red” – Magenta is a Green-absorbing pigment (equivalently, it is Red and Blue reflecting).

To summarize,
Ambient light is RGB
Cyan – As a pigment, Cyan subtracts R from RGB and reflects GB
Magenta – As a picgment, Magneta subtracts G from RBG and relfects RB
Yellow – As a pigment, Yellow subtracts B from RGB and reflects RG

Now let’s look at it the other way. You can figure the following out by examining the preceding table and calculating the “subtractions” accordingly:

When using paint to depict a colourful scene of flowers or when depicting the same scene in ink on a magazine page, you may have only CMY pigments at your disposal, but you can use them to produce each of R, G and B by using subtraction as implied by the preceding information – in just this way:
To produce B, you must subtract both G and R from ambient RGB, therefore mix:
Cyan (reflect only G and B – not R)
Magenta (relfect only R and B – not G)
This leaves ONLY B reflected (verify the subtraction yourself – have I made an error?)

To produce G, you must subtract both B and R from ambient RGB, therefore mix:
Cyan (reflect only G and B – absorb R)
Yellow (relfect only R and G – subtract B)
This leaves ONLY G reflected (verify the subtraction yourself – have I made an error?)

[NOTE: The preceding is why it is commonly said – including in this forum – that Blue and Yellow make Green. This is how people commonly confuse Blue with Cyan. As you can now see, that is not exactly accurate – Now you know that it is more accurate to say that a mixture of (subtractive-palette) Cyan and (Subtractive-palette) Yellow pigments exposed to full spectrum ambient light will reflect (Additive-palette) Green light. But nobody wants to say THAT! : ) ]

Finally, to produce R, you must subtract both B and G from ambient RGB, therefore mix:
Magenta (reflect only R and B – absorb G)
Yellow (relfect only R and G – subtract B)
This leaves ONLY R reflected (verify the subtraction yourself – have I made an error?)

I hope this helps makes it clear why it is no small error to confuse Cyan with Blue (and Magenta with Red) and more importantly, I hope this makes clear the distinction (and relationship) between and the two commonly understood primary colour palettes: RGB and CMY

How does this answer the question? It doesn’t answer your question directly, but I hope that you now see that there is no direct answer to your question since it is apparently based on a faulty premise.

That wasn’t the OP’s question. The thread title is (as you can see by the quotation marks) a quotation from a Straight Dope Classic column, linked in the OP. That question was indeed subject to the misconception you state, but Unca Cecil covered the same ground quite sufficiently in that prior column.

The OP’s questions were “Is this a marketing gimmick or is it the real deal? How do you even measure the number of colors a particular palette will yield?”
Powers &8^]

Alas, try CMYK with water colors or wax crayons.