Slashdot Mirror
New Sensor Has Real Per-Pixel RGB Sensitivity
jonr writes: "Well, the holy grail of digital photography is finally found. A company named Foveon have developed a sensor that captures RGB colours on each pixel. So what you say? Well, for the past 30 years (or since the CCD was invented) we have been using CCD with with red, green & blue sensors (or cyan/magneta/yellow) and then used software to figure out the real colour. But Foveon is the first company to deliver RGB-in-each-pixel sensor.
For those of you who are not into digital imaging, this makes a lot of difference, it's would be just as revelutionary if somebody would make a flatscreen with a real colour pixels, instead of the RGB dots. dpreview.com has the scoop.
(No, it won't mean the death of film, but I suspect we'll see dramatic improvement in quality)."
So when a digital camera is said to have 3 mega-pixels, does that mean that it only has 1 million pixels for each color??? Thus, the actual resolution isn't 3 mega but 1 mega???
High-end digital imaging devices (mostly digital and analog video cameras, but perhaps some still cameras) have been using 3 CCD chips for a long time to achieve RGB values for each pixel. It's usually done with a prism system that splits the incoming light into different colors which then are registered on different CCD chips.
In 1-chip devices, color is acheived through a matrix of filters which covers the CCD chip, allowing only certain wavelengths of light to reach each pixel on the CCD.
It seems to me that what this will really do is give us smaller, higher quality imaging devices. Let's hope X10 doesn't launch a while new popunder campaign...
Sorta... but now there are 3 full arrays to capture each color. Meaning it doesn't have to just downsample the color separation... It gives accurate color representation "in software".
A 400% increase in the amount of red and blue light accounted for and a 200% increase in the amount of green. (See figure 1.) A mean increase of 300%, but the overall image quality will be exponentially better because the true color balance will be maintained.
This is freakin' awesome, btw.
Guvegrra?
i don't think that's the problem Reality Master 101 was referring to. i forget my terms and exact figures, but the general idea is like this:
let's say pure black is light level 0 and pure white is level 10. now if traditional film can capture the range from 2-8 then digital film captures 3-7, so digital provides less shadow detail and less highlight detail than traditional film.
-f
www.blackant.net
Well the problem is mostly that there aren't enough bits coming out of the camera, so adjacent intensities get combined into one and you lose the shadow detail.
Plus a lot of cameras have a lot of noise in the sensor, which further screws up shadow detail.
Yes. At the University of Heidelberg, Germany, Physicists have developed a log-response CCD chip (covering 6 decades of intensity). They want to use it as the frontend sensor for their "tactile vision substitution system" (a machine enabling blind people to "see" with their fingers). The Log response is achieved by operating the sensor transistors in their non-linear range (very crude description, it's been a long time ago since I attended a talk about that project). Links to publications can be found following the above link.
This is very cool technology. So cool that the photodetector array must be cooled to "well below 1 degree Kelvin" in order to operate. This requires a liquid helium cryostat. So don't expect this to appear in pocket sized cameras any time soon. But it sounds great for astronomy.
Doug Moen
I have written a truly remarkable program which this sig is too small to contain.
This thing could also make one heck of a nice nightvision system, if used properly... so we could all have nice color pictures at night, just like the military folks have had for a long time. (Green screens are just for the media to consume).
--Mike--
The delta in focal length from the top to the bottom of the chip is going to be in the micron range - you have more focal length variation than that in a film camera just from the tolerances of the transport assembly.
www.eFax.com are spammers
Note that the EOS-D30 is not a "very high end" camera. It is very nice, but it's AF sucks, it has a fairly low frame rate and a small buffer (it's 3ish mega pixels on the other hand tend to crank out better images then all the 5 mega pixel $1000 cameras). The EOS-1D, Nikon D1h/D1x, and Kodak 760 are more like high end cameras (costing from $4000 to $8000), and medium format digital backs are even more expensive...
That's not to disparage the D30 though, it is a great camera, I own one, and enjoy it quite a bit. It is just far from "very high end"...
All the 5Mpixel stuff on the market, like the CoolPix 5000 mean 5 million sensors, not 5 million of each sensor. Really. Go check some of the spec pages on dpreview if you don't believe me.
It makes a lot less difference then you would think, but it does make some difference.
Yep, it's a good idea and it was around for some time.
See for example the "Creating Digital Dynamic Range Wider than Film's" chapter from the excellent "Mastering Nikon Compact Digital Cameras" book.
There are a couple of problems with logarithmic sensitivities in electronics- the little potential wells fill up too quickly. Make them too deep and they lose the low level light, make them too shallow and the electrons spill out.
Conventional AgX can capture around 14 stops of light (thats 2^14) - conventional paper can handle 8 stops or so... a typical scene has 2^11, give or take. Depends on the scene and the subject- obviously a shot of a barn with the door open in broad daylight is going to have a bit more range than a shot inside in a white room with light bouncing everywhere.
So, what you really want, is to have the SOFTWARE be cognizant of higher bitdepth images. When you have 8 bits to capture a 10 bit scene, information is lost. So you throw some out... and you end up with muddled highlights and muddled shadows, and something in the middle that looks decent.
Believe it or not, but alot of companies have spent alot of money trying to figure out the correct 'mental' representation of a greyscale- not even including colour. I'm partial to Kodak (I work there, but these views are mine).
I've worked with extended bit depth images quite a bit and know that there is none (read, big fat ZERO) ms support for anything over 8 bits.... in fact, ImageViewer simply locks up and crashes. So any sort of solution that gives you extended tonal rendering are going to have to be custom solutions... and that probably won't sit well with the average person- "what do you mean i have to process my pictures before I can view them?!?!? I'll just go buy another camera" etc etc. Even if the benefits are enormous, there is the simplicity factor that drives it.
I personally am interested in this sensor, but there seems to be the wrong website linked... which worries me...
Had you read the info on the site, you would know they aren't taking specifically about increasing the resolution.
They are increasing the amount of light captured per element.
Succinct overview: (Let's say we have a camera with 16 pixel (4x4 matrix)
In a typical mosaic CCD pixels would be the following:
gbgr
rgbg
grgb
bgrg
That's 8 green pixels, 4 red and 4 blue.
Thus 50% green, 25% red and 25% blue.
Now, after the image is captured, the camera has to do some serious image manipulation to translate the average color per pixel into an RGB value. Depending on the "correctness" of the algorithm used, you'll get all sorts of fun... low light noise, color shifts, purple fringing, moire patterns.. etc. and all that processing chews cpu time and battery life, and slows down the speed from picture to picture.
(Yes, some cameras have higher speed processors... just means yet more battery drain)
In their new method of capturing the layout is as follows:
aaaa
aaaa
aaaa
aaaa
In this case, all 16 pixels capture red green and blue. This means NO processing to be done to calculate color per pixel.
It also means a 2x increase in the amount of green light captured. And a 4x increase in the amount of both red and blue light captured.
More light being captured = more light and shadow detail.
Not having to process (interpolate) pixel colors = no more fringing, moire or other funkiness.
Now, that doesn't change the fact that digital sensors rely on converting a given set of photons into an electrical charge, and that's tough work... but having more light detected at each photo element is going to give you a LOT more light and shadow definition.
But, I digrees. I don't know of anyone pioneering new ways of doing ADC for photo elements.
Foveon has been making CMOS type chips for quite some time actually... just a special "flavor" of CMOS. They're actually partnered with NEC's semiconductor division, who actually manufactures their chips. Although YOU've never heard of them, the professional photography industry has.
They actually made cameras too in the past, but have stopped that in favor of focusing solely on the chip technology and leaving the rest to the "pros".
Their first camera used an "analog" CMOS chip (their words... better tonal reproduction, wider exposure lattitude)... actually three of them on a prism system, just like a 3-CCD video camera. It was/is stunning... achieving the same effect as their new X3 chip, with a little more complication/cost.
Although the new chip comes along with it's fair share of "buzz talk", they're definately a player, and have a proven track record of amazing quality cameras/chips on their side.
MadCow.
I used to have a sig, but I set it free and it never came back.
Probably not, as modern film emulsions are already broken into three (or more) color layers and work just fine with APO lenses. I wouldn't imagine that this new chip's light sensitive portion is much thicker than film...
-AP
Yes and no. Lower priced scanners have CCD sensors, one for each color component. This development will put drum scanner quality (think prices greater than $10,000) into scanners at much lower prices (less than $5,000 and dropping as innovation continues). Many current CCD image scanners often use separate CCD's and color filters to first scan each color channel, and then reassemble the composite image in software. This innovation will hopefully lead to much better color balancing as well as higher resolution.
vegan lifestyle : There isn't one really. It's a series of ongoing choices made by the individual. Vegan is just a quick label to tell/warn other people.
:)
As for the deer, they got along just fine for millions of years before people shot them. Looks like the shooters killed all the predators. What a mess! A difficult situation, my in the field decision would probably be different from the one I'm about to type. The population means nothing. Shooting an individual means everything to that individual. Deer, like most animals, self regulate their populations to food supply. Weak animals can't breed. The population will find a sustainable level given enough time.
Solution, do nothing. Except maybe take some food down once in a while.
Is it better to let nature take its course even if that's bad for the animals?
"the animals" don't exist. Individuals do. Nature will take it's course no matter what we do.
Can humans and their tools ever be seen as part of nature?
They already are, humans seem to think that they aren't animals. The tools at our disposal means we can produce more food with less animal products. You may have gathered it's the suffereing at the individual level that I feel is important. That may not be true of other vegans.
Is there ever a time when a human can kill an animal and not be "exploiting" it?
I try to think without the distinction between species. So replace the word animal with the word human and see how you feel about that. That's generally how I feel.
Is it acceptable to "exploit" an individual animal for the benefit of the group?
An eternal philisophical question. Humans often sacrifice themselves for the good of the group. My existence is exploited every day by those that have power over my life. Everywhere I turn profit is deemed king over people. Power & wealth isn't distributed anywhere near fairly.
If you stop considering animals and people distinct then your own values about fairness and cooperation can be applied.
Just because a cow can't talk doesn't mean it has no emotions. It's such biological arrogance to consider NO OTHER BEING except once branch of hairless monkeys sentient.
It's simple to see that a dog will feel hungry or cold. So why not lonely or happy?
I've worked on farms and been involved in the "animal rights" movement. I used to eat meat. I used to shoot rabbits and birds. I've seen first hand the disdain people have for the creatures around them. It's not difficult to witness the many terrible ways humans treat each other.
Here I am in my safe comfortable surroundings. I have no need to eat cows so why make them die for me? Mass murderers of humans are generally reviled and yet we kill millions and millions of cows, pigs & chickens. It's so wrong.
There's no real reason I shouldn't shoot you in the face and eat your tasty flesh. Out of courtesy, I choose not to and I extend that courtesy to other species. If it came down to it though and I wasn't feeling noble you might find yourself on the barby (unless you got me first of course
Carrion - nope, no reason why not (except digestion problems)
thanks for your curiosity, much better than the usual hostility people give me just for being nice!
It's all quite irrational but I have the luxury of choice and this is the choice I've made.
Incedentally and quite ironically I was diagnosed with Crohn's disease a couple of years ago. Recent research has made strong links between the consumption of cows milk and the onset of the disease. Just my luck eh! The cows got me.
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
Using CMOS sensors, it is possible to get both linear and logarithmic responses from pixels, depending on your biasing conditions.
For a linear sensor, the photosite is generally a floating N+ diffusion, that makes up one side of an NMOS transistor. At reset, the voltage here is set to VDD. As incident light generates electron-hole pairs, the electrons are collected in the diffusion, lowering the voltage in a linear fashion, dependent upon the parasitic capacitance of the photosite. When the integration time is up, this charge/voltage is sampled, and you have a linear sensor.
For logarithmic response, the reset level of the photosite is actually even with the biasing of the gate to that transistor (minus the Vt, of course). Incident light generates electrons, and the transistor operates in the sub-threshold region, making the voltage at the photosite vary as the logarithm of the current being generated and flowing through the gate region. Sample that voltage, and tah-dah, you've got a logarithmic response to light.
I admit, this is much easier to understand with diagrams of the diffusions, so if you want, here is a pdf of a paper discussing a sensor that has combined linear-logarithmic response:
CMOS Active Pixel Sensor With Combined Linear and Logarithmic Mode Operation
It may look like I'm doing nothing, but I'm actively waiting for my problems to go away.
--Scott Adams