You are quite right, but if VHS taught us anything, surely it's that 'it looks OK to me' is the usual quality assessment.
At which point the huge power reduction of LED backlights is a big plus.
We're not all trying to get Pantone matched TVs and watch them in a darkened room, to get the colours just so.
It does not require a continuous spectrum of light, however it must have certain wavelengths. It may require a trip to a textbook to understand it fully, but I will make an attempt.
In a TV, even if it is an LCD, they are using an additive color mixture. You basically need certain wavelengths of each primary color. This differs from subtractive colorimetry which is used for example in printers. In other words you can't make pure red out of orange and purple. The problem might be well described as color pollution, not much unlike impurities in a regular CRT.
Now whether we are talking about a phosphor or a color filter matters not. The color must be pure otherwise the -Y component will have to be accentualted for good color reproduction, and you never get it. Even notice on some CRT sets that some of them are better at reproducing very deep blue. Those are the RPTV CRT set in which you must defocus the blue because that phosphor is quite inefficient. With the blue in prefect focus you will not get the right color temperature without overdriving the blue no matter how new the set is. Other manufacturers intentionally pollute the blue to bring color temperature up to the proper level. They have the advanage of being able to display a sharp blue only part of a scene, but it simply is not AS blue.
Another example would be the low end NAPs of the past, the reds were orange and while they looked better under bright flourescent lights, once you got them home, in time you would fall in hate with them. And there is no fixing this in the color circuitry.
Now I am fully aware that there have been seven color printers, but such enhancements are simply not practical for display technology. That's the rules of the game, I didn't make them. Actually I haven't seen a seven color printer for quite some time, they may have abandoned the technique. I suspect it may just be too expensive, and think of what something like that would do to the cost of a TV set or monitor.
However, I suspect there is some actual white enhancement going on in some LCD units, and as many are aware, the color wheel in a DLP frequently has more than the three primary colors. I say this because after observing the display on one of those Zeniths which had a bad (and removed) blue polarizing filter, it could still reproduce white. Problem is it could only do it in the OSD. There was absolutely no modulation of the blue in the active video. So where did the white come from ?
Red green and blue are defined scientifically as primary colors. I don't know where those standards came from, nor do I care, but they are there. There is an inversion in the equation when you go from supplying the light to using incident light. A printer makes red by mixing yellow and magenta, but in this, it simply doesn't work. In the subtractive mode, by the same token, the yellow, magenta and cyan have to be pretty close to the defined complimentary colors or rendition will suffer.
However there is still the subjective aspect. Some people would find more pleasing to watch a DLP, and swear that the picture looks better, and that could be attributed to it using more than three colors. If so are they watching a more accurate picture, or is it something they simply prefer, like speakers with alot of bass or something like that ?
JURB
I agree, but the data is already split into RGB components before it gets to the monitor. The monitor can't make up the colours inbetween; it doesn't get given that information, so there's no point the light generating it.
Be careful. A lot of backlighting technology being developed is not based on traditional LEDS but on variations of OLED. The emission spectra can be radically different. Not saying tis so here, but it might be.
I was briefly involved with an OLED company trying to do this sort of thing: There are many ways, including UV-LED and phosphors..
Thats the point of it AIUI. And you can modulate each LED to the lowest output pixel that it illuminates (dont know if that tv does that). More contrast, but it messes with the ability to calibrate colours - which doesnt much matter for a consumer TV.
The flip side is that LED is a lot less efficient than CCFL. Which option consumes less I dont know.
I doubt any TV ever made has done that, nor is there any need to.
I once bought a radio that proudly proclaimed 'transistor' on the front. It did indeed have one transistor, in an otherwise valve set.
NT
I've just written a reply to you, then decided not to post it as they have better lawyers than me.
Andy
If you use narrow emission LEDs, then all you'll get is those colours.
Sod the efficiency - I want decent flesh tones. ;-)
LEDs are by nature narrow-emission -- the wavelength is determined by the band gap.
It wouldn't be horribly difficult to get the band gap set to come close to the desired primary hues.
In message , William Sommerwerck writes
Really ?
Do you really want an LED screen very high res has pixel pitch of 5mm so it does help to be on other side of football field for viewing.
Added sci.engr.lighting
Don`t think they do, very good fluro mebbe but don`t think cold cathode gets an A for efiiciency.
Someone needs to get a modern white LED, even out the front fixture efficiency is above 50 lW nowadays
Adam
You have explained nothing about the spectral character of the light that is used to create a display. You may need to consult a text yourself to get a better understanding of color science as applied to video. Poynton's text is a good place to start. Your discussion relates mostly to the choice of white points. This is one aspect of color reproduction. You are correct in stating that any color (within the gamut defined by the primaries of a device) can be made from combinations of three primaries. This is only part of the story. The way that you mix those primaries to get those intermediate colors and secondaries has everything to do with the spectrum that they can create, as well as the assumptions that are made when the source material is recorded and encoded. To get a good understanding of the matter, you need to understand the basics of color science, which really begins with the CIE standards and involves understanding the current standards for video production and display such as the ITU rec.709 standard for HD.
The color of white (gray) has specific colorimetry in the standards for video. Most manufacturers deviate greatly from these standards and start with OB settings that contain nearly twice as much blue as the standards suggest, and this is regardless of the technology. When we calibrate displays we bring them back to the standards to produce colors more accurately. In the case of your Zenith example, the filter on the blue was likely a polarizing filter or a UV filter. If you remove the UV protection, you will shorten the life of the blue panel dramatically, and if you remove the polarizer, you will compromise the brightness and black level control.
Some people do care, and those are the ones for whom a discussion of correct color reproduction matters. For the large majority of the public, there is no reason to buy one of these more expensive LCD sets when they would be perfectly happy with something priced half or a third the price of these high end units that have local dimming LED backlighting. For those that care, the jury is still out on the LCD sets. The are getting close, and are far better than many lower end PDP, DLP, and CRT based displays. There is much variation within and between technologies, as with brands.
Leonard
The display can and does make the colors in between from combinations of the primaries. It DOES get the information on the mix of those colors but that information is based on assumptions about how the display will produce the image. When an engineer designs a camera, they are trying to match the output of R,G,& B to the Standard Observer curves, not filtering it to narrow band output at a particular frequency. When the color matrix in the display recreates that RGB information, it contains the mix that will produce the intermediate colors. If the display only produces a narrow spectrum for each primary, the maker of that display has to account for that in the color decoder and map the colors to what the display can create. It can create the same colors but the mix to get any particular color may be very different than a standard matrix calculation would produce if the spectrum of the display primaries is very narrow compared to the CIE curve ( or CRT monitor response) upon which a camera is calibrated.
The bottom line is, there is more to the story than you are assuming. Manufacturers get it closer or not, it depends on the execution of their particular flavor of the technology.
Leonard
Ideally, you want the R, G & B spectrum of the backlight to match the R, G & B spectrum of the LCD pigments. That'll give you a spectrum with three big spikes in it.
No, it's standard on high end inkjet printers, & some use even more than
7 inks. For example, the Epson Stylus Pro 3800 uses 8 inks.It's certainly impractical for screens, yes.
There's no need for a display since it is theoretically possible to get all visible colours from RGB. Mixing dyes is a different matter.
Theory, remembered from many years ago, suggests that isn't quite true. I seem to recall my colour TV lecturer at college, spending a whole session on 'the chromaticity diagram', and then explaining that there were certain 'non-spectral' colours such as brown, which could not be created by an additive mix of R,G and B, and any brown that was seen on the screen was actually some kind of orange or red, which was *perceived* as brown because of the surrounding colours, and other visual cues. That might not be exactly it, as this was all learnt nearly 40 years ago, but something close, I think.
As to whether LEDs as backlights do a good job, I'm sure that they must be at least as good as CCFLs at colour rendering, otherwise, the manufacturers wouldn't be making such a thing about it. Flesh tones look perfectly fine on digital cameras which use LED backlit displays.
My whole issue with this, was that the LED 'angle' was being pushed by wording that *suggested* it was the main display technology rather than an LCD panel which it actually is, and which the great unwashed are now familiar with. That seemed to me to be a deliberate attempt to mislead people into believing that it was something new and revolutionary - as SED technology will be if it ever gets on the market, or OLED if they can get it big enough.
I don't have a problem with them claiming that this backlighting technique is revolutionary in TV sets - it is - and even claiming a reduction in power, if that's true, for a leg-up on the eco-bollox ladder, but I really think that they should be making that distinction, rather than trying to bamboozle prospective buyers with questionable use of terminology which punters are likely to have heard of, but won't actually understand.
On the power consumption issue, I still do not feel that this technology is likely to consume anything like as much as the 100 or so watts that CCFL backlighting does. The developments in the light output of narrow-angle LEDs over the last couple of years is staggering. Some of the 1 and 3 watt types could literally blind you. I believe that some cars are now starting to use LED headlamps. It would be interesting to see how they stack up against the
50 watt consumption of 'standard' headlamp bulbs.Arfa
Seeing most people don't know a liquid crystal from a light emitting diode I'd say the ad is pretty low on the deception meter.
Of which species?
Owain
Perhaps, but I think that the current generation might just be rather more savvy about this sort of thing than you give them credit for ...
Arfa
The reason that one might want to have more segment with more colors, as is done in many DLP color wheels (such as the Mitsubishis that use 6 colors, RGBCY&M) is that you can get spectral performance that is more efficient use of the available light source. You have to change your decoding to accommodate the relative spectral differences to take proper advantage of the new colors, however, and this may result in some improvement in some areas and poorer performance in others. The Mitsubishis seem to have this tweaked pretty well after several generations of doing so.
While it is true that you can get any color within the gamut defined by three primaries, RG&B, with combinations of those three, the amount of each required for a given color will vary with the spectrum available. The bottom line is that the performance depends on the execution of the system as much as the particulars of the spectrum of the available sources and the filters being used.
You simply cannot assume that all colors are going to be correctly reproduced without knowing more about the spectrum and the decoding matrix and/or LUTs.
Leonard
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.