"Stretching crystals promises flexible colour displays"

[jharker]

Just ran into this article over at New Scientist. Some researchers at the University of Toronto have developed a new material that's a kind of flexible crystal. This material is perfect for color e-paper because it reflects ambient light and requires no backlighting.

The really new aspect of this material, though, is that it can change color to become any color in the world: all the way from infra-red to ultra-violet and all colors of the spectrum. It selectively reflects the colors in ambient light. This paves the way for true full-color displays that have never been seen before, anywhere. All existing display technology has a limited color gamut, and is incapable of producing all colors that the human eye can see. With these new crystals, e-paper may also become the first display technology to truly reproduce pictures of the world as we see it!

The article quotes the researchers as saying the technology could be on the market in only two years, but I imagine it'll take longer than that. Still, very exciting!

The New Scientist article is here.
The researchers' new start-up company, called Opalux, has a web page here.

[NatCh]

Excellent find, jharker, and a nice synopsis too -- I hope you don't mind that I've put it on the Front Page.

[alex_d]

hopefully it also reflects that light well.

now as for their claims that their display is the first to reproduce true color (everything else "tricks" the eye via Red-Green-Blue)... well i think that's kind of pointless. Our eyes can only see rgb, so it doesn't matter. The highest-quality prints still use this trick, and they look stunning.

What determines color gamut in a display is the brightness of individual colors. E.g. a color may be composed of a really bright red and a dark green (e.g. 200 R 10 G). However, a regular monitor will only reproduce it accurately if it has a medium total brightness. If the display tries to reproduce it brighter, it finds it can't make the red bright enough. Similarly, if it tries to make it darker, it can't scale the green.

The quality of this new display will be determined by boring old things. Its new ability to actually reflect precise frequencies may have many technical uses (inside various instruments, for example) but is useless (and impractical) for a display.

Still, this tech is cool. P.S. Their displays are NOT flexible, just the crystals.

I think the best-looking displays i ever saw were those old watch LCDs that didn't have a backlight but instead a nicely reflective background. This tech seems essentially the same, but in color. It's too bad, though, that adding a backlight messes up the reflectiveness. In fact, I think this tech has no future because people won't abandon backlights. eInk is a special case because they promise flexibility and it has a lot of excitement around it. Fact is, if Sony made the Reader with a high-res LCD with the same reflective stuff as in my casio watch, its screen would turn out much better and its battery life would be the same.

[NatCh]

Originally Posted by alex_d

In fact, I think this tech has no future because people won't abandon backlights.

You may be right, but they could move to a CCFL frontlighting approach, such as LightWedge uses, a lot of folks would never notice/figure out/care about the difference. I know the frontlighting thing is one of my favorite comments, but I just realized I hadn't made it several months.

[jharker]

Thanks, NatCh! I don't mind at all, thanks!

Alex_d, you raised a few interesting points that I'd like to go into a little more deeply...

First, I should clarify that I said that their display should be capable of real true color rendering. I inferred that based on the capabilities of their display. As far as I know, the researchers themselves haven't claimed this (although it seems fairly obvious).

In fact, their display will probably only be capable of close to true color, since from what I've read in their journal paper (available for free download from here), it doesn't actually produce pure monochromatic colors. It does produce fairly narrow slices of the spectrum, however.

The researchers do claim that it's capable of producing a flexible display: "7. Applicable on rigid or flexible substrates", from the list of features on this page.


The concept of colors and color gamut can be fairly confusing and it took me a week to figure it out, back when I first learned about it a few years ago. Without going into the details, though, I think a reasonable illustration would be this picture of the color gamut:

The original of this image may be found at the excellent site Where's Purple?, which has several pages giving a lengthy discussion of color and the conversion of real colors into video display colors, as well as what makes a color out-of-gamut and how to bring it into gamut most accurately. I've mirrored the picture at my googlepages site so as not to impose on their bandwidth.

Anyway, in that picture, you can see several features. The whole area of color, which is tongue-shaped, is the range of colors that the human eye is able to see. The round border edge of the "tongue" represents monochromatic colors, i.e. very pure colors of visible light. If you were to shine white light through a prism, and take a very narrow slice of one color of that rainbow, that would be a monochromatic color. The numbers represent the wavelength of the light in nanometers (nm), ranging from the extreme red (650 nm) to the extreme blue, sometimes called indigo (450 nm).

Lasers produce almost perfectly monochromatic colors. Red laser pointers usually operate at 650 nm or 670 nm.

All the colors in the world that we can see are combinations of the pure monochromatic colors. To represent combining those colors on the color gamut, you would have to move from the edge into the interior of the tongue. For example, to illustrate combining a red laser pointer at 650 nm, and a blue laser pointer at 470nm, you would draw a line from the 650 nm point to the 470 nm point. Any possible combination of those colors would lie along that line. Which color you get depends on the relative brightness of the two lasers: more red brings you closer to red, more blue brings you closer to blue. Here's an illustration:

Similarly, you can combine any two colors simply by drawing a line between them.

This brings us to the question of computer screens. Most computer screens (and TV screens, etc.) use three colors of pixels: Red, Green, and Blue. Unfortunately, the Red pixels aren't monochromatic red, they're a blend of red, so they're away from the edge of the "tongue". Similarly, the Green and Blue are also blends. Now, when you combine two colors, all the colors you can get are on a line between those two colors. To get all the colors a computer or TV screen can make, we just need to draw the Red, Green, and Blue points in the color gamut and connect them, making a triangle like this:

Every color a computer screen can make lies within this triangle. As you can see, there are plenty of colors outside it! Those are all colors that we can see, but that the computer (or TV) screen can't make. Newer HD televisions have slightly different colors of R, G, and B, and they can make a slightly larger triangle. But they are still pretty limited compared to what the human eye can see.

The interesting potential of this new display technology is that it might allow us to create pixels colors that are truly monochromatic, and not limited only to Red, Green, and Blue! We could make pixels of any monochromatic color. And by combining those colors, we might really make pictures with almost every color the human eye can see!

Some interesting links:

• The Where's Purple? page explains why purple is a creation of your brain. (So is brown!)
• The related Primary Colors page discusses why the primary colors of the eye are not red, green, and blue.
• Wikipedia has some excellent pages on the subject. The most relevant are the entry on the color space, and the entry on the RGB color model.
• Better color gamut illustrations from Wikipedia: the visible color gamut, and the RGB color gamut.

[Steven Lyle Jordan]

Well, detailed lesson in color perception notwithstanding, this is great news, as it brings us that much closer to "full color" ( quotations added in deference to jharker) displays and color e-readers.

Let's see: If I start scanning and digitizing my graphic novels, comics and manga now, in 2 years' time I might be just about ready...

[lexico]

Nice summary from jharker!

Also as for being the first to reproduce true colour, possibly the claimants hadn't head of the Lippmann process (circa 1891); this captured the wavelengths of the light directly. There's a very short/poor Wikipedia article about it at: http://en.wikipedia.org/wiki/Lippmann_plate

It's now getting renewed interest as an 'unforgeable' biometric image or 'signature' for passports, etc.; see: http://www.holographyforum.org/HoloW...pmann_Security

.... and could be a basis for e-book displays of the future...

[nekokami]

Thanks for the overview of color gamut, jharker! That was a very nice summary of a set of concepts that I vaguely understood, but would have had a lot of trouble explaining to anyone if asked.

I think a couple of other aspects of this display will factor into its eventual success (or lack thereof): pixel size, power consumption and refresh rate. Pixel size can only be helped by being able to set the color of each pixel without having to use 3-4 pixels to get color (as with LCD displays now). If the power consumption is still fairly low (and generally speaking, reflective screens use a lot less power than backlit screens, that holds some promise for ebooks. But speed of refresh is still a potential hurdle. This is an issue with e ink now-- probably existing e ink displays are about as slow as the market can possibly accept, and too slow for some purposes. I wonder how the refresh rate would be with this technology?

(Then there's manufacturing expense, of course....)

[alex_d]

man... those gamut graphs are 256-color dithered GIFs! that's really ironic. that site should fix it.

but thanks for explaining something I didn't consider before. Gamut has a lot to do with the purity of the red, green, and blue that the monitor can produce and now i see that this tech will help.

However, a real color graph would have to be three-dimensional (eg, this particular 2d projection has no concept of brightness). I don't know if many people realize this, but the color gamut is much more complicated than a triangle when considered in the full 3 dimensions. Ie, at high brightness the only thing left of the glorious color gamut of any monitor is a white spot in the middle.

[jharker]

Yeah, that's actually a really good point! While I was writing the above, I read an article that pointed out that color is also dependent on brightness. For example, brown doesn't appear on the color plots above. Brown is actually how we perceive a low-intensity red-orange, but most people wouldn't consider brown to be the same color as red-orange.

[Tuello]

Thanks for pointing out this very interesting new technology and for the explanation of color displays. While looking at the color gamut on my rgb monitor it struck me that I shouldn't be able to see any colors outside the triangle, but I do - what's up with that?

[alex_d]

it struck me that I shouldn't be able to see any colors outside the triangle, but I do - what's up with that?

its because the graph is all just a big, fat, stinking lie and noones got the guts to admit it. *spits on the floor*

actually, the graph is somewhat accurate for the triangle inside. Each corner is the proper 256 R or 256 G or 356 B. The center is (256, 256, 256) which is pure white in MS Paint's little color picker toy (a great thing to experiment with to learn about color*). Outside of the triangle, however, is where some artistic type got creative. Notice that colors outside the triangle seem kind of the same, just darker.

you're right, there's no way to show any colors outside of the triangle.


the 2nd-best bet is to perform a trick where the triangle itself doesn't actually use the monitor's purest red or brightest blue. You can approximate the relative relationships by remapping the colors so that the triangle is drawn with dull pastels and the outside uses more of the monitor's capabilities.

hence, you get a gamut graph that makes more sense.

in fact, similar forms of remapping are used by people to adjust photogoraphs in order to achieve maximum quality. however, parameters that give the best color fidelity also make the image, on average, more dull. That's a tradeoff they have to make, and unfortunately they don't usually please everyone.


*who said computers aren't educational? everyone, press WinKey+R and type mspaint (then enter). Then double-click one of the colored boxes at the bottom, followed by 'Define Custom Color!' Here, you can compare colors to their RGB compenents. Better yet, you get to pick colors out of the 3 alternate dimensions of Hue, Saturation, and Luminocity. (which are dumb words for color, intensity, and brigthness) Note there are many other 3-dimensional systems by which color can be specified (a famous one being YUV) but you can't play with them here.

in the color-picker edutaitional dialog box are evident the two fundamnetal problems of gamut (i'm telling you, this thing is deep):
1. the first row of the color-picker box doesn't really look anything like a rainbow
2. all colors wash out when you turn up their luminocity or look a bit the same when you turn it down.

[Alexander Turcic]

The Technology Review has an article on Opalux (whose website seems currently down), which quotes the company saying that their e-paper technology is superior to that of E Ink and others.

With P-Ink, it's a different story. "We can get 100 percent of the area to be red," Arsenault says. This is because each pixel can be tuned to create any color in the visible spectrum. "That's a three-times increase in the brightness of colors," he says. "It makes a huge difference.

"It's a spectacular innovation," says Edzer Huitema, chief technology officer of the Dutch firm Polymer Vision, based in Eindhoven. Even traditional screens, such as cathode-ray tubes, LCDs, and plasma displays, use three or even four differently colored pixels to generate color. "It's a major limitation for all color-display technologies," Huitema says. When the color of each pixel is controlled, not only does the color quality increase, but the resolution should also improve by a factor of three.

Arsenault predicts that Opalux will have the first products on the market within two years, probably in the form of advertising displays. But, he says, it will be a long while before P-Ink will be in a position to completely replace traditional displays. "The caveat is that we are not at video speeds," Arsenault says.

Currently, the P-Ink system can switch pixels in less than a second, which is on par with other e-paper displays. "We're still early in our development, and there's a lot of room for [improving] the material and optimising its performance," says Arsenault.

http://www.technologyreview.com/Infotech/19337/?a=f

[NatCh]

Sounds like a nearly viable e-magazine device already, but I'm having drool-inducing dreams of a video wall ....