Improving Digital Photography 401
Milican writes "'It's easy to have a complicated idea," Carver Mead used to tell his students at Caltech. "It's very, very hard to have a simple idea...And now one of Mead's simplest ideas--a digital camera should see color the way the human eye does--is poised to change everything about photography. Its first embodiment is a sensor - called the X3 - that produces images as good as or better than what can be achieved with film.'" We had a previous story about Foveon last February.
Pixel Noise (Score:3, Informative)
I hate pixel noise in my digital pictures. I have heard that since red color has to be detected at the deepest part of the silicon there is an abudance of noise in the reds.
Foveon won the PopSci Best of What's New for 2002 (Score:5, Informative)
Review of X3 Camera (Score:2, Informative)
This is hardly news... (Score:5, Informative)
http://www.dpreview.com/reviews/sigmasd9/ [dpreview.com]
It's like the eye because... (Score:5, Informative)
What you don't get is Moire patterns - at all!! That is what you probably hate when you say you hate "pixel noise" because it's totally obvious (due to the color changes), very distracting, and annoying to clean up after.
Re:Uh-oh, here come the digital bashers. (Score:2, Informative)
Plus, the speed of film is better. Digital cameras aren't very good for action photography.
So, uh, yeah. Digital is great for posed shots in good lighting. So I guess it is the best. Whatever.
Re:Pixel Noise (Score:5, Informative)
How is this at all like the way the human eye sees?
This foveon system is like the human eye inasmuchas the light photons penetrate multiple layers and register at more than one levels in the same spot. For example, take a look at this cross section of the human retina [eyedesignbook.com].
Current CCDs only collect one waveband of light at one area. To simulate colour, they collect three different wavebands in adjacent areas on the surface of the CCD. Hence the funky moire patterns you that you see in tightly patterned cloth on the sample piccies on the site.
I hate pixel noise in my digital pictures. I have heard that since red color has to be detected at the deepest part of the silicon there is an abudance of noise in the reds.
If the upper layers are completely transparent in the red, then your concerns don't apply. As long as the actual transparency of the upper layers is reasonable, then there is little cause to worry - traditionally CCDs are far more sensitive to the red end of the spectrum than the blue so even modest photon loss at the red end is unlikely to seriously degrade the pictures.
The other nice thing about this technology is that the spatial size of the light bins is approximately three times larger than that for the equivalent physical sized CCD - that means better signal-to-noise ratios for this new technology.
Anyway, the presentations look compelling. I await cameras with reasonable numbers of megapixels (say 4Mpixels +) and reviews...
Cheers,
Toby Haynes
Re:It's like the eye because... (Score:5, Informative)
It sees a real "color" instead of on red/green/blue (dispersed in fine pixels of course). It may not be able to see red quite as well as other colors, but it only means that the sensitivity at the red level is the limitation you have for the picture as whole.
I don't think I agree - it still looks like a standard red/green/blue pickup (and that is exactly like the human eye - we don't have different cones for, say, lime green and grass green). There is possible mileage in having more layers picking up wavebands spanning a smaller range of wavelengths (and there are humans with 4 types of cone rather than 3 - tetrachromatic vision) but it's not going to matter too much for our normal vision. Useful for simple spectroscopy (colour profiles etc.) though.
What you don't get is Moire patterns - at all!! That is what you probably hate when you say you hate "pixel noise" because it's totally obvious (due to the color changes), very distracting, and annoying to clean up after
It's pixelated still so you will still get Moire patterns as soon as the smallest details are finer than the resolving power of the X3 bins (think Nyquists theorem). However, the bizarre colours you get from a fine-grained black and white grid shouldn't be present to the same extent as all the measurements of colour intensity are done at the same point in the X3 layer, as opposed to the different spatial positions of the red green and blue bins in a colour CCD.
Cheers,
Toby Haynes
Re:Uh-oh, here come the digital bashers. (Score:5, Informative)
Remember, I said "please be sure you have used the gear".
The ISO 1600 and 3200 shots from the pro digitals are easily less grainy and have better dynamic range than their film counterparts. Try it. And my EOS-1D can do 1/16,000 shutter speed with zero lag. Is that fast enough for you?
Yet another person who is bashing without trying.
Re:Uh-oh, here come the digital bashers. (Score:2, Informative)
Most importantly, not many consumer level output devices can print photos as well as film. I have seen some really nice photo prints from digital but, on the average, still not as good as well developed film.
Re:Uh-oh, here come the digital bashers. (Score:1, Informative)
Re:Pixel Noise (Score:5, Informative)
The X3 actaully measures RGB at each pixel, giving much better quality, at a higher speed.
Re:Uh-oh, here come the digital bashers. (Score:2, Informative)
What do you think it is about low light situations that precludes digital cameras from working well?
As for speed.. yeah, my digital camera only goes up to ISO 1000. But you don't have to go to 1000 to take normal non-posed shots successfully (There's a lot of space between posed shots and extremely fast moving action shots.)
You forgot to add that you can't use UV or IR film in digital cameras.
Re:Need new screens, no? (Score:3, Informative)
RG
GB
The CCD device in a digital camera has one of these set up for every pixel the camera is to capture.
This new way will allow all 3 colors to be captured on one "pixel" instead of 4, so that will allow much higher resolution pictures to be taken. Hopefully this simplified explanation makes sense, and didn't totally confuse everyone
Re:Uh-oh, here come the digital bashers. (Score:3, Informative)
Besides, night scopes are digital, and they seem to work ok. You can even buy them at CostCo.
Nyquist free... (Score:5, Informative)
The bizzare colors (what I really hate about digital photos) are not just reduced - they are gone. If you read the review at DPReview.com you'll find that it has resolution right up to Nyquist is noise free and you get some detail beyond. Here's the relevant section (near the very end of the review, where they test against some resolution charts):
The SD9 is capable of delivering all nine individual lines of the horizontal or vertical resolution bars up to its maximum absolute resolution (sensor vertical pixel count) and slightly beyond. Note also that because the X3 sensor doesn't need a color filter array it doesn't suffer from color moiré.. Absolute resolution is just less than the Canon EOS-D60, Nikon D100 and Fujifilm S2 Pro (at 6 mp).
However, because the X3 sensor doesn't use a low pass (anti-alias) filter it is able to resolve detail all the way up to Nyquist. Beyond Nyquist the system will alias without any objectionable color moiré. Where a Bayer sensor camera would turn detail beyond Nyquist (such as distant grass texture) into a single plane of blurred color the SD9 will continue to reproduce some individual pixel detail (without color moiré).
Hubble? (Score:3, Informative)
Film still rules for taking pictures in low-light.
So that's why the shuttle keeps visiting the Hubble Space Telescope, to pick up the film!
The is also a company called SBIG [sbig.com] that makes a line of digital imagers for amatuer astronomers.
Steve M
The future of digital image sensors (Score:5, Informative)
To sense an RGB (Red, Green, Blue) pixel one can use a veriety of methods. At the center of this technology lies the ability to turn a stream of photons into an electric current. This photodetector is colorblind, it is only capable of measuring the _amount_ of light, not it's color. To recognize color the estheblished method used to be to put several photodetectors near each other and put color filters in front of them. The most widely used color filter array is known as the Bayer pattern and consists of 2 green photodetectors (diagonal from each other) a blue and a red detector in a 2x2 grid. These 2x2 blocks are then repeated over and over to create the full image sensor.
Specialized software or hardware needs to take these individual Red, Green or Blue pixels and recreate a single RGB pixel, this technique is known as demosaicing. The major advantage of this method is the simplicity of the photodiode (photodetector). It allows for the creation of very dense image sensors that are now passing the 10MegaPixel barrier while keeping the cost down (start seeing 5MegaPix sensors for less then $100 before the end of this year).
Foveon's approach is to layer these color filters vertically.
The good:
- idealy you get R,G,B at each pixel.
The bad:
- very complex layered photodiode technology, this makes the pixels significantly bigger. Currently the pixels are bigger then a 2x2 bayer image pixel. The complexity also adds to the manifacturing cost, these chips will not be cheap for the forseable future.
- Color bleeding. For example: Photons in the green wavelenght do not nescecarily stop in the green layer, but might be picked up by the underlying red layer. This means that specialized hardware needs to apply a non-trivial color correction for each pixel layer.
Foveon's idea is a very interesting approach. Since they nicely pattented their idea shut, we will have to patiently wait for this single company to provide the world with this technology.
Side fact: The human eye see's colors using pigments that respond differently to different wavelengths. In the simplest model we can say that we see Red Green and Blue with spatially seperated pigments that resemble a bayer image sensor closer then the foveon's sensor.
Re:May not be such a big deal... (Score:2, Informative)
Re:It's like the eye because... (Score:2, Informative)
You all are getting caught in marketing hype. Nothing more.
Re:It's like the eye because... (Score:2, Informative)
It sees a real "color" instead of on red/green/blue (dispersed in fine pixels of course).
Ahem. First of all, a Foveon sensor just stacks the red/green/blue sensors on top of each other instead of putting them side-by-side like conventional Bayerian sensors do. In the digital files that Foveon cameras output, each pixel is still represented by a R-G-B triple.
And second, I am not sure what do you mean by a "real" color. You mean all other digital cameras don't see real color? Our TVs and monitors do not produce real color? etc. etc.
What you don't get is Moire patterns - at all!!
Hate to break it to you, buddy, but moire patterns have nothing to do with digital cameras. They are easy enough to see with a human eye, anyway. You can get a moire pattern on a Foveon just as easily as on a regular sensor.
That is what you probably hate when you say you hate "pixel noise"
Umm... get a clue. Pixel noise has nothing to do with moire. It's NOISE -- more or less random fluctuation is the measurements of light that each sensor spot does. The amount of noise depends on a bunch of factors, such as temperature and the size of pixels on the sensor. That's why astrophotographers cool their CCDs in liquid nitrogen, and that's why professional digital cameras like Canon D1s, D60 and such have hugely lower noise than (relatively) cheapo consumer digicams.
And, of course, the Foveon sensor will also exhibit "pixel noise" -- that a fact of physics and kinda hard to get away from.
Sounds a bit like the Segway (Score:3, Informative)
Too much hype. All they did was stack pixel detectors rather than mosaic them. The mosaic was simpler and now cheaper, this thing costs $1800 in a camera, else I'm sure someone could've come up with it. The real accomplishment is creating those silicon layers precisely, not coming up with lets stack em
They say the resolution is like a 120mm film, and the color lattitude is big. So are CMOS sensors in Canon and Nikon's cameras. Checkout the awesome photos on photo.net [photo.net]. A lot of those have been shot by modern digital cameras with CCDs and they dont look bad. Mead has his own marketing to do to try and take Foveon to Intel and Microsofts level, so he has to push down CCD. Theres a reason why people are buying digital cameras with sensors smaller than fingernails and submitting their pictures on professional photography site. I think Mead has work to do.
Re:Digital Color Correctness (Score:4, Informative)
Be careful with fluorescent lighting as its spectrum is quite fragmented to a specific wavelenght and can be tricky. Watch out also for street lights, as they tend to either be sodium (orange), quartz ("halogen") and sometimes arc (very hot - blueish). Mixed lighting is hell for any cameraman (film - mostly slide - is lighting-specific too), hence the use of filters on lights or windows to balance the whole set to a consistant colour temperature.
Hey, I wonder how this X3 sensor deals with different coulour temps. Anybody care to enlighten me on this one? My CCD colorimetry/sensitometry knowledge is kinda rusty...
Cheers,
max
Re:Uh-oh, here come the digital bashers. (Score:3, Informative)
Yes, and don't forget the other end of the spectrum too, that these cameras can take wonderful long exposures as well. The D60 in particular can sit on Bulb for minute after minute without any major noise or pixel errors. Taking ten minute bulb exposures seems fairly "low-light situation" to me. I've had comparable results with the D100 has well. I also regularly take 10 to 15 second exposures with it, and never once have I had to contend with excess noise, boomy shadows, or any other difficulties.
Me thinks these people are playing with their friend's Kodak DC3400 or something.
Just to be clear... (Score:3, Informative)
RG
GB
blocks as one. That block is 4 pixels. Foveon-based cameras would have
(RGB) (RGB)
(RGB) (RGB)
which is still 4 pixels, but gives you more accurate color information at each pixel and reduces moire. So, while there will not be any more pixels per area with Foveon CCDs, the *effective* picture resolution will be much better.
I wish I had known this before I shopped for digicams-- it feels like false advertising to me, and I learned after I had made my purchase. Manufacturers ought to be required to state "4 single-color Megapixels" or "1 Megapixel effective with color" for 4MP cameras with traditional CCDs.
Re:It's like the eye because... (Score:5, Informative)
Hate to break it to you, buddy, but moire patterns have nothing to do with digital cameras. They are easy enough to see with a human eye, anyway. You can get a moire pattern on a Foveon just as easily as on a regular sensor.
If you're talking about color moire, you are just wrong. read the DP review and look at the resolution tests. The X3 has no color moire at all. That is not to say it does not experience noise (it does). Just not color moire artifacts.
Show me an image from a Foveon that has Moire if you are so certain. Remember that JPEG compression causes artifacts too...
Umm... get a clue. Pixel noise has nothing to do with moire. It's NOISE -- more or less random fluctuation is the measurements of light that each sensor spot does. The amount of noise depends on a bunch of factors, such as temperature and the size of pixels on the sensor. That's why astrophotographers cool their CCDs in liquid nitrogen, and that's why professional digital cameras like Canon D1s, D60 and such have hugely lower noise than (relatively) cheapo consumer digicams.
As C3PO would say, "How Rude!". Yes I understand the difference between noise and moire. But when I see digital pictures that look like they have issues, usually the most noticable aspect (to me) is moire and not so much the noise (which mimics film grain to some extent so our brain does not latch onto it as looking so artificial, being trained by looking at years of film images).
That is why I was saying that if he hated noise, it might really have been color moire he was seeing and hating, not the actual noise from the CCD. Thus I was not saying moire was noise, I was saying that might be what he meant by noise.
Hype, hype, hype... (Score:3, Informative)
HERE is where it gets interesting, and where I get to my point. Cones are what we use to see color. An individual cone cannot see red green and blue as this marketing hype would lead us to believe. "The cones come in three types: Red (60%), Green (30%) and Blue (10%). The red and green cones are randomly distributed in the center of the fovea and the blue cones form an annulus around the outside." So in effect this camera will actually surpass the human eye.
As a side note, the link goes to a very interesting document that states how "126 million photoreceptors must be transmitted to the brain via 1 million fibers in the optic nerve [while] [t]he overall compression ratio of 126:1 is not evenly distributed." Check it out.
Re:Whats the hold up? (Score:5, Informative)
X3 still displays some odd behaviors under certain conditions, and until these problems are resolved, the "big guys" aren't going to want to put it into a high end camera -- especially when their customers have grown to expect a certain level of all-around quality and attention to detail from them.
Re:Uh-oh, here come the digital bashers. (Score:3, Informative)
So, adding a $2,200 D60 wasn't a *huge* step, price-wise. I've had it around 6 months, and I've shot around 7,000 frames with it. Assuming for the moment that I'd have shot the same number of frames had I been using film, that averages out to $0.35/frame, which is in the same general range as film and processing (that's $10 for 36 exposures).
Assuming that I've got at least another couple years of functional use in the camera, the per-frame cost should drop down under a dime. Plus, I get instant feedback (nice when fiddling with lighting problems) and it's easier for me to sort, edit, and produce prints with digital then it is with film.
So, with six months of use, you can start to argue that it's paid for itself. Add another couple years of use, and it'll be hard to argue that it would have been cheaper to use film. So, even if it has no resale value in 3 years, it'll still have been a good move, financially speaking.
I suppose it all depends on how much you shoot.
Re:Comparison vs. Nikon F-Series (Score:1, Informative)
Re:Pixel Noise (Score:2, Informative)
For instance, a 15 megapixel bayer chip digital camera is better than 35MM color film in almost all metrics (except dmax).
An 8x10 piece of color film is likewise better in most aspects (except noise).
A 2mp fovean chip may look pretty good, but not better than 35mm.
See? It depends on the parameters of the technologies.
Re:The future of digital image sensors (Score:3, Informative)
Wrong. Said software or hardware takes two green pixels, a red pixel, and a blue pixel and recreates four RGB pixels. It conjures two thirds of its information out of thin air. (I've written software [duke.edu] to do this for the Color Quickcam.) The worst two effects of this hack are color moire and blurring. Color moire is when detailed B&W objects (detail above the Nyquist frequency) gets colorful edges. Blurring is the loss of detail that occurs when cameras use an anti-alias filter to reduce color moire.
dpreview.com has an excellent review [dpreview.com] of the Sigma SD9 in which they examine the pros and cons of the Foveon image sensor. It really does eliminate both color moire and blurring, but there new artifacts to be fixed.
Re:Whats the hold up? (Score:2, Informative)
Re:Uh-oh, here come the digital bashers. (Score:3, Informative)
In my opinion, the real key is the storage format. Consumer cameras generally store in 24-bit (8 per channel) compressed (i.e. JPEG) format and you lose a great deal of information that way -- the limitation is the storage format itself (JPEG), which isn't capable of holding all of the color and light information the camera captures -- the camera simply throws it away before storing the image. Of course in some low-end consumer cameras, the sensor is that poor to begin with.
With pro cameras you generally store the important shots in a raw format (12-bit per channel, 36-bit total) that discards nothing; you can then manipulate this in Photoshop as a 48-bit uncompressed image in a wide colorspace and get dynamic range and color reproduction very similar to what you can get with good quality film. If you happen to be on the road with your pro digital and need your images to stay as small as possible, many higher-end cameras will also allow you to shoot in JPEG format but using an enhanced colorspace (i.e. Adobe RGB rather than sRGB) to try to preserve this additional information while still gaining the benefits of compression. However, to use such JPEG images you must have software which supports these enhanced colorspaces (i.e. Photoshop does, GIMP does not).
Re:Uh-oh, here come the digital bashers. (Score:2, Informative)
But that's not all you're paying for. You're not counting the cost of storage and printing. DIY-enlargements work out to a couple bucks per 8x10 (about the same as having film enlarged to the same size). And what about ink longevity?
You have a lot of good points. I've spent about a hundred on film storage since I took up photography (negative/slide sheets, storage boxes, etc), and I've bought an extra 80 GB hard drive to store pictures from the digital camera. I have a 35mm film scanner, and I was using it to scan and print negatives before I got the D60, so a lot of the printing comparisons break down. Anyway, 90% of the time, I get prints by burning a CD and dropping it off at the local Costco with a Frontier, so the print cost and longevity are the same as film prints. I have a decent inkjet, and I use it occasionally, but you can't really compare it to conventional prints--I can print any size up to 13x19, and I can spend hours tweaking it to look the way I want it to look. 95% of the time, there's something wrong with machine prints, either from digital or film, but it's ususally too much of a pain to get it fixed. A DIY printing solution gives you a lot more control, at the cost of taking longer. So, I use both--if I'm not feeling super-critical, I let Costco handle it. Otherwise I do it myself.
Also by storage I don't just mean the cost of a single hard drive, regardless of size. You've got backups, transfers to other media, etc., to worry about. And 20 years from now my negatives will still be in the box on the bookshelf, available for prints and enlargements. Where will your photoshop files be?
Good points, and mostly looking for a good solution. Except I've had a hard time finding several pages of negatives for the last year, and I have a bunch of other negatives that I cheaped out and had Costco reprint (rather then scanning them and giving them the files), and now they're scratched. Film and Digital *both* have storage issues. They're just not the same issues :-).