View Full Version : Kodak digital breakthrough
xayraa33
06-14-2007, 05:59
no need for fast lenses then?
http://www.cbc.ca/technology/story/2007/06/14/tech-filter.html
doitashimash1te
06-14-2007, 06:04
Sounds like a breakthrough indeed. I'm curious! Thanks for the post.
charjohncarter
06-14-2007, 06:05
Sounds good, but still no increase in dynamic range.
Interesting, perhaps down the road this could improve high ISO performance in the M9, etc.
An alternative to the Noctilux?
it's only one stop. I'm assuming the idea that this would eliminate the need for wide apertures is a joke :)
I don't use my flash. So, better to sell it now?
anselwannab
06-14-2007, 20:20
CNN (http://www.cnn.com/2007/TECH/ptech/06/14/low.light.photos.ap/index.html)
Doubles the light sensativity sounds good to the lay-person. I wonder if the quantum effieciency would be better for larger photosites of a APS-c sensor. Might be interesting for 4/3 sensors.
Mark
So it is only one stop and fails to address the dynamic range issue...
It is still a step forward, and to me we're not all that many steps from making digital my medium of choice.
May they keep coming!
What this actually means is that lenses can be designed a full stop crappier, and that the marketing buzzword for this filter allows cameras to be sold for twice the price..
Seems like the whole concept of shallow DOF is going down the drain, isn't it?
we dont all use fast lenses just because they see better in low light....selective depth of field is the main reason my lenses are all fast
and besides, ever time a new sensor technology comes out someone pops up and calls it a breakthrough or revolution....yawn....
I had the same thought as Mark ... the implications for cameras based on the 4/3 standard is a positive. The knock on the E1 was that its high ISO performance didn't put it in pro territory, even though the body and ZD lenses are certainly pro spec. I don't really care that much about high ISO performance, but I can't deny that the success of the E System is at least somewhat dependence on it gaining at least some market share in a pro market.
I also agree that any advance, whether it be quantum or incremental, is good, and benefits anyone who needs or uses digital capture.
In terms of RFs, I think this means that there is now a better chance (even if slight) for a digital RF with a 4/3 sensor. Given that one of the value propositions of 4/3 is the possibility of smaller bodies and lenses, I see this as an incremental step that supports the development of that. Yes, it excludes M mount and LTM lenses, but I think there is room for a dRF with another mount. Call me crazy.
zuikologist
06-15-2007, 03:52
Forgive my slow brain - does this breakthrough mean we will have a sensor that is half b&W (panchromatic, but presumably this means just records the "amount" of light) and half colour sensitive (ie the usual pixels we have now).
Is resolution the same or halved?
bmattock
06-15-2007, 04:11
This announcement is both more and less than it seems to some.
First, the big news here is that the Bayer mosaic is what is being replaced. This is what limits resolution in many cases:
http://en.wikipedia.org/wiki/Bayer_filter
This is what gets the Foveon folks so hot and bothered when people tell them that their new Sigma camera is NOT a 14mp camera, because the 14mp arrived at are stacked three deep, not side-by-side, and the Foveon folks reply that a 'normal' 14mp sensor isn't one either, because the total is arrived at by GUESSING. And before you argue with me, 'interpolate' means 'guess'. Look it up. It's a reasonably good guess, but it's a guess.
What Kodak has done, apparently, is replace the Bayer filter (which they also invented) with a new type of filter (undefined in public as yet) combined with a new type of light sensor that is more sensitive to all colors of light. Combined, these technologies will have two effects.
First, they will allow faster shutter speeds at lower light, which will reduce blurring of moving objects - sports photographers will like that.
Second, the technology will be patented and licensed - as the Bayer mosaic is now. Kodak desperately needs revenue streams.
And just as a side note. A personal axe to grind here, please forgive me. There is a person who likes to post from time to time "Kodak isn't innovative and hasn't produced anything of note since Tri-X." I would like to point out that Kodak files more patents in a year than 99% of all US companies and more than most other photo-related companies with the possible exception of Canon. They invented digital photography (and failed utterly to capitalize on it), they supply the image sensors for many of the digital cameras in use today in a number of high-end digital cameras such as Hasselblads and Leicas. I've had this person I'm speaking of challenge me on this, I proved they were wrong, but they never admitted it or apologized, and that person continues to state that Kodak does not make the high-end sensors for medium format digital cameras. So, that person is a liar. You can choose to believe him if you wish. Kodak innovates. They are quite inventive. It may be their only stock in trade at this point, due to amazing mismanagement and poor marketing in recent years. But they do innovate. This may well prove to be a nice breakthrough, affecting us all in time. End of rant.
zuikologist
06-15-2007, 04:33
Bill - is that you?
Not trying to start a flame war, just trying to understand. I can appreciate the Foveon argument, since the pixels are all colour - albeit stacked. This seems something of a hybrid - very welcome if it works well and it provides another alternative to photographers.
Regarding Kodak - couldn't agree more. I understand that with perhaps one exception, Kodak are the only volume manufacturer of mf sensors.
Welcome back btw.
bmattock
06-15-2007, 04:55
Yeah, it's me. As far as I can tell, anyway.
And I just hope Kodak survives. Call it nostalgia, but I like Kodak. I think they're trying really hard to make the right moves now, although they'd admittedly been pretty brain-dead for a long time.
Seems like the whole concept of shallow DOF is going down the drain, isn't it?
I really don't think that Joe Consumer (or Jo Consumer) has the slightest clue what DOF is, let alone cares about it when taking digicam snapshots.
oftheherd
06-15-2007, 05:42
Yeah, it's me. As far as I can tell, anyway.
And I just hope Kodak survives. Call it nostalgia, but I like Kodak. I think they're trying really hard to make the right moves now, although they'd admittedly been pretty brain-dead for a long time.
I have a certain brand loyalty myself. Although I like Ilford films' drying characteristics over Kodak, and like 25 ISO b/w film, Kodak was my first love in films. They gave me terrific museum shots in the late 50's and early 60's. They took care of police photographers without making a big deal of it. More of a public service kind of thing. And in Vietnam, they were a big help with my crime scene photo processing! Then they just sort of seemed to die in the marketplace. Hopefully they can not only keep their head above water, but grow to offer more to us photographers.
While the possibility isn't discussed, how would you all feel if a side effect of the new filter made the angle of light entry unimportant? By this I mean the new filter/ sensor design allowed RF lenses to perform at their best?
Bryce: I would be ecstatic if that were the case, but I don't know the technology well enough to guess if it's possible.
anselwannab
06-15-2007, 11:39
How about this sports fans, the SD14 has it so that you can remove the filter and make it IR capable.
What if you could remove the Bayer or this filter and make it just black and white? Wouldn't that improve the resolution?
I think that would be a cool way for Zeiss to get around the 'The sensor must be good enough for our lenses" arguement. Assuming that Zeiss is actually not just hiding behind that as they take time to develop their camera.
Make it so that you can use it as a color camera, but can also use it as an ultra-high resolution B&W camera; maybe with less vignetting issues.
To me for camera phones and such, the less noise is the big issue.
Seems to me the killer would be to take this new technology and mate it with the Fuji big/little site sensor.
Mark
Bill: I appreciate the good words about Kodak. Yes, absolutely they do innovate; those who don't think so are just plain ignorant of the facts or choose to ignore the facts out of some personal, emotional position.
The fact that they failed to immediately capitalize on the digital technology they invented was a result of corporate culture, not the lack of intelligence and effort of their scientists and engineers.
Another Rochester company invented the GUI, the mouse, the foundation of Ethernet, an integrated office application ... and also failed to capitalize on it.
bmattock
06-15-2007, 12:01
I know, Trius. I had an AT&T 3b2 computer running Unix System 7 and a (Xerox) GUI interface pre Windows, pre Mac. It had a mouse and 20 serial ports for dumb CRT workstations. It was cool.
bmattock
06-15-2007, 12:06
Seems to me the killer would be to take this new technology and mate it with the Fuji big/little site sensor.
Mark
To me, this is what is getting cool and interesting about digital camera technology. Up until very recently, digicams have tried hard to mimic what film cameras can do - and failed. Oh, they're more than close enough in many ways, I'm far from anti-digital, but they don't have (for example) the latitude that color print film does. They're not appreciably better at lower or higher ISO than film can be had at (say ISO 25 to 3200 more or less). And with the smaller sensors, a lot of creative control (selective focus) is given up that could be had with larger recording media (medium format and large format film, especially).
But when we get to an era where we can have digital sensors do things that film never could - ah, then it starts to get fun. When we have a digital sensor that can do ISO 100,000 and have it look like Tri-X with tight grain - ugga bugga. Me for that. Then we can start taking photography to a new level, not just trying to swap one axe for another and noticing that it is not quite as good, despite other conveniences it may have.
bmattock
06-15-2007, 12:35
I am a trifle confused as well, but it replaces the bayer filter, as I understand it, and it is coupled with a more sensitive sensor. Not sure how the two work together. Waiting for more details.
I'm not sure about ISO 100K ... 6400 or 12800 seem fast enough for me.
But I do think digital sensor technology is in its infancy.
Oh, and yeah, I remember the 3b2 ... never owned one, but played with one and was enthralled.
bmattock
06-15-2007, 13:01
You have to think outside the 'box' so to speak. ISO 100K would let us shoot stopped down in near darkness. And I'd love to see a digital sensor with latitude greater than that of the human eye. One that could internally filter IR as well as UV in or out as requested - no filters required.
Imagine a sensor that allowed you to selectively set sensitivity for different parts of the scene. ISO 50 here, ISO 100K there. The human eye does that - with help from the brain. No film or sensor can do that so far.
When liquid lenses become reality in production models, we may see some really interesting experiments in selective focus. Lens babies will die of envy, and imagine being able to mutate into any lens characteristic there ever was, including those which can be designed but cannot be built because glass won't do that?
anselwannab
06-15-2007, 13:11
I am a trifle confused as well, but it replaces the bayer filter, as I understand it, and it is coupled with a more sensitive sensor. Not sure how the two work together. Waiting for more details.
My understanding is that with out a colored lens, the sensor gets all the light that strikes it, not just the Red, green or blue.
Seems to me you could have big and small sensors like the FUJi and get better Dynamic range.
I wonder if that is why they say that this is for small sensors? Do they have another technology for larger sensors?
Mark
Equally likely is the possibility that this new transistor/filter technology will be used to make image sensors that are only incrementally better than those being manufactured today, but with reduced die size; meaning that the manufacturer gets to build more of them on the same sized silicon real estate, effectively reducing the per unit die cost. Translation: $$$.
How this could benefit small sensors is that for the same die size as current products they will have better performance, and thus can be sold for more $$$.
bmattock
06-15-2007, 14:08
Hence, the reason 110 film took over from 35mm.
Oh wait, you mean it didn't. Dang. I guess sometimes superior formats win even if they cost more.
Headlines are written either by marketing hacks whose job it is to pump up the company, or by newspaper hacks whose job it is to pump up circulation. Combine that with both groups not really understanding the technology, and you get headlines galore. Even Toyota does it. ;)
bmattock
06-16-2007, 20:28
You discuss the **possible** pros and cons, not facts.
True, we're speculating. However, there is one thing that Kodak has stated as a fact - this new technology replaces the Bayer Filter. That may end up being useful or not, but it is certainly a change, and a big one.
The headline had the buzz word: breakthrough. This is not a breakthrough at all, this is just a tiny step into another direction.
I'd say that since all CCD and CMOS sensors except for the Foveon use a Bayer Filter, this is a breakthrough. Semantics aside, one can call it what one wishes, of course.
A real breakthrough would be a storage medium you can hold up to the sun to see the images (like a slide) without any power supply. :D
Well, of course if only you are allowed to define what a breakthrough is, then anything you don't like isn't. Gee.
Anything else is a marketing hype, hot air, fraud.
Marketing hype is endemic, puffery is part and parcel of capitalism, like it or not (and I don't, actually), but it is far from fraud unless there is no such chip.
Are you suggesting that Kodak has not invented a new type of sensor that does not use the Bayer Filter? That everything they announced was a lie? Then, yes, that would be fraud.
Funny how much emotion is generated around the work "Kodak."
Sure. Maybe I should mention that I'm not affiliated with Toyota. I'm just driving a 20 year old 'bush taxi light' (=Toyota Landcruiser LJ70), and being a designer and photographer I chose 'toyotadesigner' as my alias. Sorry if this was misleading. :angel:
I knew that ... I was just playing on your ID to make my point. :angel:
clintock
06-17-2007, 06:06
Could a similar result be achieved by doubling the size of the sensor assembly for a given site count? Why the push for such density?
bmattock
06-17-2007, 06:14
> I'd say that since all CCD and CMOS sensors except for the Foveon use a Bayer Filter
The Bayer Filter is a specific pattern for the Mosaic filter used over the CCD/CMOS. It was patented. It uses twice as many green filters as red and blue, and the interleaving of the filters. The pattern used optimized the interpolation scheme for computing red-green-blue values for each pixel.
The Kodak DCS100, Nikon Coolpix 900 and 950 (and some others), and some others use different patterns in their Mosaic filters.
Beg pardon, I didn't know that. However, none of those are current cameras, so I would still guess that replacing the Beyer Filter is a reasonable 'big deal', yes? I mean, it is all down to semantics, but I'd call that a breakthrough.
come now people, lets be reasonable here. Take a 8 megapixle digital photo then print it at 12 inches, then take the same photo, resize it to 1/3rd its size then print it 12 inches. The simple fact is even 8 megapixles isnt enough to do a decent 300dpi magazine spread, which is why most places dont accept them. I know the foveon people will get on me for this but their so called 14 megapixle pixles just are not the same thing. Yes its 14 megapixle if you count total number of pixels, but if you print big one of the fuji or sigma sensors wont cut it against the sensor in my 8 mp canon.
Its not a who has the most layers of sensors game folks, if you have any experience printing digitally you know that.
Lets change the idea, I take a a film photo but I want maximum choice in everything so I stack bw, color slide, and infrared on top of each other, that doesn't mean my image is a 645 sized image, just means there is more data (if you could shoot through three pieces of film like that) which is one of the reason why 4x5 film gives you such clean fidelity, less enlarging, more over all data.
bmattock
06-17-2007, 06:31
This isn't about Foveon sensors, it is about Kodak's announcement that they have come up with a new replacement sensor that is more light sensitive and uses a different filter than the Bayer Filter used in traditional sensors. Foveon was mentioned just as to the fact that it is currently the only alternative to Bayer Filters.
But let's not have another Foveon-bashing session. You're wrong about Foveon technology, but that's for another thread. Like we haven't had that one about a dozen times.
vBulletin® v3.6.8, Copyright ©2000-2013, Jelsoft Enterprises Ltd.