Slashdot Mirror
Kodak Unveils Brighter CMOS Color Filters
brownsteve writes "Eastman Kodak Co. has unveiled what it says are 'next-generation color filter patterns' designed to more than double the light sensitivity of CMOS or CCD image sensors used in camera phones or digital still cameras. The new color filter system is a departure from the widely used standard Bayer pattern — an arrangement of red, green and blue pixels — also created by Kodak. While building on the Bayer pattern, the new technology adds a 'fourth pixel, which has no pigment on top,' said Michael DeLuca, market segment manager responsible for image sensor solutions at Eastman Kodak. Such 'transparent' pixels — sensitive to all visible wavelengths — are designed to absorb light. DeLuca claimed the invention is 'the next milestone' in digital photography, likening its significance to ISO 400 color film introduced in the mid-1980's."
Of course, you achieve this increased light sensitivity at the expense of losing 1/4 of your color resolution. Maybe if you want the increased sensitivity it might make more sense to pick up something like the Canon 1D Mk III, which, at least according to Ken Rockwell, gives great results all the way up to ISO 6400. I'd hate to lose 1/4 of my color resolution *all of the time* to get the added sensitivity that I only need for a small fraction of the shots I take.
It is hard to evaluate this from the press release. People have tried all sorts of variations, including ditching the whole pattern thing for true color (Carver Mead) and the results are about the same as other cameras.
too little too late from Kodak?
I dont know how this will play with the market leaders for DSLR (canon/nikon) or even for medium and large format back (phaseone) etc.
I cant really see Canon using Kodak sensors
Canon release the eos-30d equiv or eos-350d/400d equiv with this sensor within the next year? If so I'd wait to purchase :)
If your neighbours roof is flying past your window, you know it's cyclone season.
Kodak is going to patent this, and use it themselves and license it out to other companies (hard the story last night on NPR). For those who would abolish the patent system, why would this not be a "good" patent?
Please discuss.
"As God is my witness, I thought turkeys could fly." A. Carlson
The summary says the extra pixel is a 4th, but surely it is a 5th. 2*Green 1*Red 1*Blue and then the new one.
and color in the 70s.
I refer you to Tri-X b/w, and to Fujichrome 400 around 1972. a really nicely balanced and warm film. if you pushed it to 1200, you could peel the grains off the base and go bowling with them, but the picture held up remarkably well on the small screen. it was THE go-to magic film for 16mm newsfilm when it came out.
if that was a negative film, it would have been asa 800 with little more grain than the "fast" 125 color film of the time.
if this is supposed to be a new economy, how come they still want my old fashioned money?
I already feel that my digital rebels have remarkably low noise sensors and give me better results that shooting Velvia 50 and scanning. Still I usually carry a tripod and shoot at virtually never shoot at high ISO so it doesn't really affect me.
I expect this will have more value in cellphone cameras. Typically the noise floor goes up when the sensor shrinks, and increasing the brightness without increasing noise would be a massive boon for most cellphone photographers.
As you state, DSLRs already have fairly decent sensitivity, so this is not likely to be a good compromise for them.
Modern 'compact' digital cameras, however, which stuff 7-12 megapixels on 1/1.8" and 1/2.5" sensors (smaller than your fingernail) could benifit enormously from this. These sensors are already past the diffraction limit of most of the lenses, so a drop in color resolution may not be too damaging (the eye being less sensitive to color resolution, than luminance anyway). Kodak is claiming a 1-2 stop increase in sensitivity, which would be a great benefit to anyone using a compact inside, or in other poor light. (I have yet to own a camera that performs well above ISO 200)
As with all such tech announcements the proof is in the pudding, and until we can compare full size samples to conventional bayer sensors, its hard to tell if this is the next big thing or not.
That's a neat trick. I wonder how they can do that?
This is really not anything new to the image industry, just a new application. There is already the CMYK colorspace for printers, which is effectively an RGB + black to get deeper colors. I don't see this as really revolutionary, as much as "Can't believe this hasn't been done yet." Though, at least they admitted this too :)
My biggest hope for this is to reduce per pixel noise by being able to reference the fourth plane, but I doubt they will get there for a while, they still have to work out the color conversions.
Kodak has rediscovered what evolution found millions of years ago -- design a dual system such as the rods and cones of the biological eye. The average human eye has about 120 million sensitive, panchromatic rods and only 6 or 7 million color-sensitive cones (many in the central fovea). The brain merges the limited amounts of color information with the larger volume of B/W image data to paint color into the image that we think we see.
Two wrongs don't make a right, but three lefts do.
There was a story here a few days ago about them adding a "clear" pixel element to allow more light through. Sounds like the same premise.
"Sometimes a woman is a kind of religion, she can save your soul & set you free from all your sins" - Bad Examples
They divide each sensor of the regular bayer pattern to 4, half white, half color. This way one can also report a 4-fold increase in the number of pixels, without really increasing the resolution. (which actually will be a boon for digital photography, since no one needs the current resolution anyway, because the optics doesn't keep up, but a megapixel race is on...)
But does anyone know why sensors use RGB and not CMY? a Cyan filter would let green and blue through, but keep red out, instead of blocking two parts of the visible spectrum for each pixel. This way, by simply switching color space, the camera becomes twice as sensitive to light. I.e. instead of use or something like that. One could even combine the two methods, and use white pixels, to gain a slight further increase in light sensitivity (from 8/12 to 10/12). Is there any reason that current cameras use RGB?
Sure, "faster" sensors will be a boon to the consumer market, and will surely have some applications in the pro market as well -- existing light press photography come to mind.
For me, though, the problem is not so much speed as it is noise and dynamic range. That's because a lot of the time I still do fine-art level landscape and studio glamour photography -- neither of which are speed starved, but even the finest digitals could still use even less noise and wider dynamic ranges.
While DSLRs have a huge advantage over handhelds in this regard, it would still be nice to see improvements in s/n such that the darker zones maintained their clarity and detail. Even the finest Canon cameras suffer to a degree in this regard, at least for people with very high standards. Some of us have those standards because that is what our clients demand - and in some cases we still must use film to meet their criteria.
It's a virtual law that to obtain the best noise performance you need to use the lowest ISO speed that the camera can attain. So instead of bottoming out at 100, like most DSLRs, I'd like to see 25. Or better, 12.
For more info, visit http://www.normankoren.com/digital_tonality.html
They posted a full press release with images and sensor layout diagrams, additionally there is an excellent discussion in their news forum with a lot of good information. http://www.dpreview.com/news/0706/07061401kodakhig hsens.asp
One of the problems with DLP projection TVs with a "color wheel" was that since every color lets only 1/3 of the light through, the picture was dim. So they added a fourth element "clear" that lets out all the light to get every projected pixel a blast of light they need and the remaining portions of the color wheel adds only additional brightness for each color.
This technology seems to be kind of similar. The transparent sub pixel detects over all lumninosity and the remaining pixels "adjust" for color. Very close to what we have in our retina too. Almost all our cylindrical cells respond only to luminosity and the cones respond, to varying degrees, three colors. A poster was complaining about losing "color resolution". I think millions of years of evolution has shown us the balance. You need about 90% of the pixels responding to luminosity and just 10% to color. The same ratio in our retina.
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
This is so obvious - I've personally wondered why 1CCD sensors they don't have a fourth pixel group to carry brightness information only. There must be good reasons why this has not been done before now; I hope we get to find out why.
The patterns they suggested in the article were not as elegant as the Bayer filter (where each color formed an evenly spaced grid). They may be hiding the actual pattern for now or there may be some technical reason for those patterns that I don't understand, but I would suggest this pattern (C = Clear):
it keeps the same 4clear:2green:1red:1blue ratio but the different color pixels all form a regularly spaced grid.
Prov 9:8 Do not rebuke mockers or they will hate you; rebuke the wise and they will love you.
..the ISO 400 reolution was largely lost on me.
"The agriculture ministry is not in charge of Gundam" - Japanese ministry official.
This gets me wondering:
Does the clear array have a flat sensitivity level across the spectrum? Where it will give the same data value for the same number of photons striking it with a 700nm wavelength as it would for photons striking it that vibrate at 400nm?
If the sensor (for example here) was more sensitive to red, then this would skew the picture results significantly, especially if it picked up and added infrared light to the picture's data which isn't visible to the human eye.
They need every bit of light they can get because the sensors are so small. Resolution and color depth aren't really a problem in that space, but brightness really is.
Will this fourth sensor finally mean we can capture Gendale in all its beauty?
So Kodak's best cameras would then have acceptable noise at, say, ISO 400?
For most photography applications, it is a meaningful advance for which there is no downside.
The marketing hype surrounding resolution just keeps spinning further away from reality.
Digital photographic prints off the average production photo printer (my costco has them right on the floor) the lines per milimeter resolution is _way_ below what even a **really** good digital SLR with **great** optics can capture.
Also keep in mind the color gamut of the average digital camera is quite narrow, and unsophisticated compared to analog. There are a number of segments of photography where film still rules the day because the results are more "cinematic" than digital.
So throwing out 3/4 of the color resolution still leaves you with extra data that will be thrown out when the data hits the paper. I can think of one or two exceptions, but they are way, way out of the norm.
A related anecdote, I recall the photos from the Mars rover were taken with a 1.5MP sensor and they made *gigantic* beautiful images.
Got Trader Joe's? friendwich.com RSS feeds work now!
While I like Kodak's idea quite a bit, here are a couple of other ideas.
1) Sony was building cameras for a while with four color channels. There was the normal green, but also a different green they called "emerald" for one of the four Bayer pattern locations. Unfortunately, this was a solution in search of a problem, it never really caught on because there just wasn't any perceived benefit.
2) I do visual effects for films. For the last 50 years or so, people have been using bluescreen and greenscreen effects. The idea is to put a constant color background, and process the image so that any pixels of that color become transparent. Over the years, more and more lipstick has been applied to this pig -- so that you can now often extract shadows that fall on the greenscreen, pull transparent smoke from the greenscreen plate -- these things have become even more possible through digital processing.
Still, it sucks. Greenscreen photography forces so many compromises that I often recommend shooting without it and laboriously hand-rotoscoping the shots.
But -- say you had a fourth color filter, with a very narrow spectral band. Perhaps the yellow sodium color -- commercial lights that put out very narrow-band yellow are sometimes used for street lighting. If you had a very narrow-band sodium filter over 1/4 of the pixels, you could pull perfect mattes without 99% of the artifacts of traditional greenscreen and bluescreen photography. Finally (and this is killer!) you could make glasses that the director of photography and other lighting crew could wear that block just that frequency, so they could see the set as it really is -- without the sodium light pollution.
Still, Kudos to Kodak for thinking outside the box.
Thad Beier
I love Mondays. On a Monday, anything is possible.
It's done on TV all the time and nobody complains (chrominance is separated from luminance and often transmitted at much lower resolution). As has been pointed out below, your eyes are made up of rods (which see black and white) and cones (which see color), and only a fraction of those cones are devoted to each individual red, green, or blue spectrum. So your color resolution is already significantly lower than your luminance resolution. You can even see photos demonstrating this with a 9x decrease in color resolution (3x in each linear direction). You're most sensitive to green, which is why the Bayer sensors commonly used in digital cameras divide each 4 pixels into GRGB.
Does the clear array have a flat sensitivity level across the spectrum? Where it will give the same data value for the same number of photons striking it with a 700nm wavelength as it would for photons striking it that vibrate at 400nm?
Probably not... but the sensor will have some known characterization and the Bayer->RGB(->jpeg) conversion (that is done in-camera or on the computer if you handle RAW files) will account for this when it reconstructs the full RGB value for each pixel.
If the sensor (for example here) was more sensitive to red, then this would skew the picture results significantly, especially if it picked up and added infrared light to the picture's data which isn't visible to the human eye.
Most digital camera sensors have a infrared filter over them to prevent this. Canon sells a version of their 20D called the 20Da which is specialized for astrophotography. I believe the primary change compared to the standard 20D is the removal of the IR filter.
First against the wall when the revolution comes
I imagine that's part of the reason it hasn't been done yet. Finding the "true luminosity" from a nearby Red, Green, Blue, and Clear CCD is probably nontrivial. I imagine that IR sensitivity isn't as troublesome as you'd suggest, though, since most cameras now come with IR filters over the CCD array. Photographers interested in IR (and UV) photography sometimes have to have that filter removed outright.
"Evil company X is threatening to restrict our rights! Let's all get together to stop--OOOH! SHINEY!!!" -- AC
...and fancy lights?
My Hitachi CMOS works fine without any fancy color filters. Case mods are getting a little out of hand these days...
{/joke}
If you haven't been down-modded lately, you aren't trying.
Sacred cows make the best hamburger.
I would imagine that the camera is built to take into account the sensitivity of the sensor across the spectrum when converting the RGB + Luminance to RGB for output. It would be similar to the same calibration necessary to to get the colors right in the first place. You would have to figure out how the sensor reacts to the R, G, and B wavelength and apply a gamma transformation (or whatever, I'm not photography or light expert) to what the sensors detect to get a result that represents what the human eye would see in the first place. Adding the luminance channel makes it more complicated, but it's still the same kind of problem.
Actually, it still amazes me how complicated color really is.
You are in a maze of twisty little passages, all alike.
I wonder how this is going to compare to the Foveon sensors. They capture RGB data at all pixels - filtered based on depth rather than location. Now if only those babies cost less.
Sure I'm paranoid, but am I paranoid enough?
WhyTF is the Kodak press release showing a disc of random colors instead of the sensor layout?
Film really is nice, but changing film rolls is something i will never ever want to do again. I'm perfectly happy with a handful of 4GB compact flash cards and my canon eos-3D. Although i want a Mark II-1DS and Mark III ;)
Now as for losing color resolution, I think you won't lose much. The only place you are going to notice it is in dim light and it will be less than 1 bit of loss. Those would be shots you would nt have gotten anyhow because they would have been below the camera's ability.
Prior art? LCD projectors do this same trick to brighten the projectors for presentations. rgb+white on the color wheels. This is also why some projectors, designed for movie viewing, are a littel dimmer for the same wattage because they leave out the white on the wheel for better color saturation and higher wheel speed.
Some drink at the fountain of knowledge. Others just gargle.
If the sensor (for example here) was more sensitive to red, then this would skew the picture results significantly, especially if it picked up and added infrared light to the picture's data which isn't visible to the human eye.
I would flip that around and say that that behaviour might actually be advantageous. If you're in a low (visible) light situation, maybe you could use an IR flash to get luminance values and merge that with the dim visible colour data to get a halfway-decent colour image with no visible flash.
"...always new atoms but always doing the same dance, remembering what the dance was yesterday." -Richard Feynman
It'd be interesting to see the algorithm for that sensor. The color values of the adjacent pixels are going to have to be taken into account. A bare photoelectric sensor will provide a higher voltage for higher frequencies of light. To a "white" sensor, a blue photon looks brighter than a red one.
More complexity for RAW filters.
What are the alternatives to using filters? I have been wondering how feasible an approach based on spectrum analysis would be, i. e. if it would be possible to build a matrix of arrays of sensors, with each array having, say, a micro-prism on top of it?
The point she was trying to make was that when it became available everywhere (pharmacies and five and dimes) at a similar price point to ASA 200 then there was mass adoption, and most peoples' snapshots gained quality. Sure, they picked up some grain and lost some saturation, but most people care about non-blurry and better exposure.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
If you have read anything on the physical nature of the eye, you should come across three rather important things about how it sees. One being that there are (in the average non- colorblind/tetrachromic human eye) four light receptors: A red, blue, and green cone and a contrast sensitive rod. Second, if my memory serves me, the rod cell can detect down to a single photon while the cones are around 6 times less sensitive. Third that the rod is most sensitive to green and if this bias is reflected in the fourth pixel, it would help red and blue areas not look brighter in the image than the original subject. When you make a device to store an image, I would assume it should collect the same sort of information that the eye would.
In response to the previous post, however, the fourth, unfiltered pixel would decrease color resolution by 1/4 but it would be negligible because of the sensitivity of the rods. This is one of the ways they shrink the high definition movies. In that case they scrap 3/4 of the color resolution and with little perceived difference.
The human eye is awesome for what it's evolved to do. Photography, however, is a different task. The human eye is good at resolving things infront of it while catching movement to the sides and only turning if it's interested. A camera with a well resolved center section but lousy edge resolution except for movement is one of the last things we want.
So, yes, the human eye is the result of millions of years of evolution and is a very efficient means of achieving its task. Just don't confuse the task of a human eye that's trying to ensure our survival with that of a camera that's trying to capture all of the detail of a moment for posterity.
The 8 Mpixel color image that comes out of your camera is already a complicated guesswork; in terms of real color information, it's more like 2-3 Mpixel, since there really are only 2 million complete RGBG cells.
Making one of the RGBG cells into a "white" cell doesn't really change much of anything in terms of resolution: color resolution is still half what grayscale resolution is. And it does actually help with color accuracy, since having four different receptors lets cameras deal a lot better with fluorescent lights.
This will also increase dynamic range slightly, since the "white" receptors are more sensitive than the RGB receptors.
This looks like a very good thing overall, really.
Only low-end consumer gear doesn't put an IR filter in front of the sensor. Since the goal of a camera is to faithfully reproduce the color in the scene as visible to the human eye, not putting an IR filter in defeats that purpose.
Make me a friend and I'll mod you up
Researchers here at Rensselaer Polytechnic Institute recently came up with a super non-reflective coating -- it basically has nano-spikes that help absorb light from all angles and at all frequencies. Seems like it would be good to use for the dark pixel. http://news.rpi.edu/update.do?artcenterkey=1956
you're going to be blowing out quite a few of the clear pixels
In a production CMOS/CCD assembly this is not likely. In order to get a digital camera sensor to produce a pleasing image in many lighting conditions the CCD/CMOS assemblies already have controls for this.
The best example of proof is to try using a scanner head as a digital camera. You will find that the CCD assembly in a scanner is not designed to handle variable light, so most things outside a narrow range of brightness (luminance maybe?) are blown out.
Got Trader Joe's? friendwich.com RSS feeds work now!
What the heck? I posted this news yesterday and an admin deleted it. Thats bollix!! Give credit where credit is due!
Not true. 4K is next and is being used now for feature films. Peter Jackson thinks it is the future and many industry icons believe the same.
Once you have seen 4K with 4:4:4 colour space you may take that back. Some say it is pretty well 65mm resolution.
Is that article supposed to demonstrate that removing colour is not a big deal? I would never want something like that in my screensaver, and I'm not known to be picky. The writer also cheated by using an image that contains a lot less colour than a lot of real-world photos. And even then the results looked horrible.
Now, I know that percieved colour resolution is lower than percieved intensity resolution, but this is a fine balance between spatial and intensity resolution. By ignoring this, the article managed to make it seem as if the point the article makes isn't valid, when in fact it is (for the most part).
In printing technologies, at least in the early '90s they were using a technique called either "GCR" (gray color removal) or "UCR" (under color removal) which basically transfer almost all of the "light density" information from the cyan-magenta-yellow films of a color separation to the "K" film (black) -- because black ink is quite a bit cheaper than the alternatives. I have seen images printed with up to 90% of the density in the black that are virtually indistinguishable from images printed from a "normal" color separation by the naked eye, and sometimes if a high enough line screen value is used (+200 LPI) it is hard to tell that a print is a GCR'd image even with a magnifying glass.
So it stands to reason for me at least that if I devote more attention to capturing the "amount" of light with "one CCD eye" completely open, and the "quality" (hue and tint) of the light with my "other three CCD eyes" that are filtering for spectra, I should be able to do the same thing digitally that they have been doing optically in printing for yearsand still yield a superior result.
I'd love to hear a discussion about the best way to use the digital bits in a 32 bit "GCR" digital world by the way. For example, using 10 bits (1024 levels) for luma, 8 bits (256 hues and tints) for green, and 7 bits (128 hues and tints each) for red and blue, or whatever the optimal case could be
Thoughts?
...Open Source isn't the only answer -- but it's almost always a better value than the alternatives...
Are world class optics -- used by Hasselblad, Rollei, Yashica and now Sony .and unless things have changed radically since I last checked, have been the top lenses along with Nikon for years and years. I think Leica used Zeiss lenses also and Leica cameras WERE top notch back when. Not sure how far Canon, Minolta, etc. have made up the distance since about 1996, by the way.
...Open Source isn't the only answer -- but it's almost always a better value than the alternatives...
A choice quote: "It's almost inconceivable that nobody else thought of, or acted on this idea, until now." That sure sounds like they think this is obvious. Does that mean they'll skip getting a patent?
Cameras aren't "good enough" until I can shoot fast action at high magnification in near dark with a compact camera. Then I will be happy.
Then you will never be happy because it's intrinsically impossible to capture low noise images "in near dark" with a small area sensor because of photon noise.
We're significantly more sensitive to inconsistencies and errors in luma than we are in chroma, so perhaps it is time to try another option to see how well it can satisfy that sensitivity.
This has the promise of significantly increasing the dynamic range of the luma axis, as well as significantly reducing luma noise. Those sound like good improvements to me.
i'm a karmawhore, so i'm ACing this. but you'll still get an informed answer, heh. basically, you get what you pay for. Zeiss has the name, but i believe much of the low end consumer digicam stamped with ZEISS isn't particularly great. i mean, if Rolex made a $12 watch, how good would it be? probably better than other $12 watches, but certainly not a comparison for a $5k submariner.
that said, Zeiss makes some awesome high end lenses. The digiprimes (for 2/3" digital cinema cameras) are amazing, and run around $12k/each. The digizooms are like $60k, iirc, and also very good. I haven't use the zooms, but i've shot with the digiprimes and they are amazingly sharp.
my $.02
Not all cameras have the same goals. An X-Ray camera is designed to be sensitive to X-Rays, not visible light. Likewise, many photographers want to use the IR spectrum - so putting an IR filter in defeats that goal. Likewise, many photographers work in black-and-white, color reproduction is not an issue for them, but tonal reproduction is.
Why do you think that all cameras have the same objective?
... and then they built the supercollider.
The noise performance will be terrible. There is enough trouble matching and matrixing signal levels between 59% Green and the weak and noisy R and B channels. These guys want to match with a 100% clear channel, driving R and B more into the noise. Then they propose to recover G from clear by subtracting a more noisy R and B, ruining with their noise the G channel your eye is more sensitive to.
Marketing hype: back to the drawing boards and the successful optimized volume produced parts.
Because it's obvious to anybody who is vaguely in the field.
:)
There's "prior art" - human eye: cones = colour, rods = black/white.
What next? Camera people are going to put a reflective coating behind the sensors so light that goes through them will bounce and have a second chance to trigger the sensors?
See cat eyes - been there done that.
Now if they layered lots of semitransparent sensors, so that photons that aren't picked up by the first layer get a chance at the second, third etc, then MAYBE that's something different.
The way I understand it, for CMOS sensors, there is a significant non-light sensitive circuit area between sensor elements, so if photons hit that area they are usually lost. But I believe there are ways of making transparent transistors, so if the circuit area between the sensor elements is made transparent and you put another layer of sensors UNDERNEATH then you can capture _some_ photons that would have previously been lost.
But even so, how nonobvious is that? I'm not in the camera line and even I can think of crap like that
Yeah, and the Bayer-patterned sensors that this thread is about would *totally* work for any of those applications. Unless astrophotographers started using point-and-shoots when I wasn't looking, the cameras that these chips will be used in will be aimed at consumers taking snapshots.
Make me a friend and I'll mod you up
Their SuperCCD sensors have two photodiodes, one with a high sensitivity but relatively low dynamic range for the chroma component and one with a high dynamic range for the luma component. Seems like Kodak have just introduced a small variation on an existing approach.
Why wouldn't Bayer-patterned sensors work for infrared photography, black-and-white, or astrophotography?
Unless astrophotographers started using point-and-shoots when I wasn't looking,What makes you think that only point-and-shoot cameras use Bayer patterns? High-end DSLRs also use them, as do many specialized cameras.
... and then they built the supercollider.
I've personally seen the idea for a clear pixel idea 2x in 2 different sensor companies besides Kodak. It's not a revolution or a light bulb clicking idea. Christ, I bet 1/2 the grad students who take a class in pixel technology think they've solved the world's problems when they came up with this "original" idea.
... like we did... lol.
Both times we ran into the same problems.... color resolution is a predictable loss and not a deal breaker since the eye needs luma more then it needs chroma. You're taking pixels that you were getting color information from and taking out the color information... of course you're losing color resolution.
Here are the biggest 2 stumbling blocks with a clear color filter... or panawhateverpixel they're calling it.... it's a clear color filter.. jesus... self important sobs...
1) charge density - since you're not filtering out 2/3 of the color spectrum you're getting a lot more light coming into this pixel. That sounds good BUT the pixel saturates much much faster and then you get no information at all. A pixel can only hold so much charge before it maxes out. It's max charge is largely a function of the size of the pixel. When you max out the pixel all you're not getting any more info... so bottom line.. you're going to have to cut the amount of light that hits the pixel.. either by crippling the pixel relative to the other color filtered pixels or by decreasing integration time of the all the pixels. Crippling a pixel is just dumb so you decrease integration time. That means the pixels you need to get color information from are now getting even less light then you would have if you just stuck with a regular bayer pattern. So maybe your nighttime scenes look ok now since you don't care about color but your daytime scenes have either really poor color or huge areas of oversaturation.
2) cross talk - What happens is that light doesn't hit the silicon at a 90% angle (except in the middle). Most of the time it comes in at an angle. All manufacturers use shifted microlenses to redirect the light but that doesn't solve the problem.. it only helps. There's still lots of light that hits the silicon of pixel A after it's traveled through the color filter of pixel B. In the normal bayer pattern that's not a big problem because the color filter of pixel B will filter out every color except for say green... and when the light gets to pixel A... the color filter from pixel A will block green light because of the way the bayer pattern is layed out. Even if the light hits the silicon photodiode without crossing pixel A's color filter, you're getting a greatly reduced signal so the crosstalk isn't as bad as it could be. With a clear pixel you have a problem... pixel B has a clear pixel so it doesn't filter anything out. When the light hits pixel A it's going to falsely contribute a great deal to pixel A's signal..... so now you're colors aren't just lower resolution.. they're just plain wrong.
I haven't read their paper yet. I'm sure they've thought through these 2 problems and have probably addressed (2) by laying out the pattern differently and/or doing some whacky math to minimize the problem. However there's no chance that this is a revolutionary pixel that will change the face of cameras for the future. Hell.. it won't even matter as much as the Foveon pixel. Now that had the promise of a revolution... if it just didn't suck so bad...
Actually, chances are they didn't think throught the problem... they probably implemented it and then said... "aw damn... so that's why nobody does it..."