SonyAlphaRumors posted that image of a new sensor that will be launched in early 2015. It does get rid of the common Bayer RGB pattern because each single pixel is capable of capture the full color spectrum. It does it by using an electrified moving filter.
SAR writes that design will be used for a 1.5 inch Z4 smartphone first and than used on larger photocamera sensors too. We can expect following advantages:
– Bigger pixels compared to same resolution Bayer sensor. This means more electrons captured and therefore higher dynamic range and lower noise (crazy ISO possible…even crazier than those of the Sony A7s).
– No moire issues anymore. No Anti Aliasing filter needed (increases per pixel sharpness).
– That tech also allows to make crazy high megapixel sensor. For example if you keep an Sony’s APCS pixel size same as the current Sony A7r 36MP RGB pixel size sensor you could in theory make a much higher resolution sensor with same noise level.
– Less pixels to read means also faster processing and readout.
– You actually can have a “full-monochrome” sensor
As usual with those kind of new stuff, we have to see them for real to see how good it works! ….But it looks good on paper 🙂
12 Responses
Fuji has proven through all their RAW conversion issues that changing the CFA concept in the present sensor technology isn’t possible without introducing resolution and color killing artifacts. Time for something new, but also there, I remain a bit sceptical. The first wave of adopters is always solving the problems for the next generations. But it could indeed be a revolution. More than that fabulous but never landing organic sensor story.
Doesn’t Foveon sensors show this is possible?
Will this be the result of projects someone (from Sigma?) said both Sony and Canon were working on, following principles alike the Foveon’s sensors?
But offering small sensors does it mean they are not confident enough to apply the new tech to APS-C or FF cameras for the time being?
Nobody knows how the organic sensor will be yet, when it will show up and if it will be used only by Fuji or also by (at least) their project co-partner (Panasonic). But do you really think Fuji’s X-trans introduced resolution killing artifacts? If so, it seems to go against all known reviews of X sys cameras.
This is not a small sensor (it has a 23.8mm diagonal). It’s got a longer diagonal than a micro four-thirds camera sensor. It’s almost twice the diagonal length of the 1 inch sensor in the Sony RX100.
And no, this works in a completely different way to Foveon. That’s if it exists at all.
You’re right but what I said is that there are other sensors (e.g. Foveon) that do not use Bayer filter without introducing all the negative effects ejpb said to be inevitable.
I would never be the first to jump in as it will take 2-3 year for proper RAW support. At least with Fuji you have a half descent JPEG engine but I would not want to depend on Sony’s JPEG
Sony knocked-off Sigma’s Foveon design.
You clearly have no idea how Foveon works as it doesn’t use filters at all (active or passive). Foveon works due as the penetration power of photons varies according to wavelength. As such it has three depth layers which produce separate colour channels. Unfortunately the penetration profiles are not sharp, so there’s a lot of cross-contamination of signal information which is why Foveon is not good at high ISO (too much information is lost sure to these common mode errors),
In contrast, from the little information we have, the Sony sensor would use an electronically switchable filter at each site and sample the three colour channels separately. That’s a completely different technique which would lead to better colour separation albeit with loss of sensitivity and being prone to motion artifacts.
However, there’s good reason to believe this is not a real product as many of the features lack credibility, especially as it combines revolutionary technology with infeasible specifications (2K at 1600fps?).
It’s also difficult to believe that a sensor with a longer diameter than a micro four-thirds camera will end up in a phone as it will require proportionately longer focal length lenses to maintain field of view. It is twice the size of the largest sensor in a camera phone, the Nokia 808, itself a bulky device with a distinct “camera bulge”. This would be much worse an wholly inconsistent with Sony’s slimline smartphones (which is where it is rumoured to be destined).
It may be a sensor for super high speed cameras like the Phantom series.
Just a typo, 1.5 inches is not 23 mm, thats rather a cropped Aps-C on 3:2 sensors, 1.5 in diagonal would be a full frame 36 mm which is not. For all people whi seldom talks in terms of one company leaving others out of the game please think that is not the best for the industry, competition and richness of ideas has always been better on any field, either technollogical, biological or social systems.
1.5″ sensor doesn’t mean it has a 1.5″ diagonal. Google sensor size.
1,5″ is its “optical size”, the diameter of the image circle. A strange measurement but people are used to this lecacy so lets keep it. :p
“– No moire issues anymore. No Anti Aliasing filter needed (increases per pixel sharpness).”
Plain wrong. The kind of things one guy said once and then others repeat. Any digital sampling system correctly designed needs an antialiasing filter as long as the input signal can contain frequencies beyond the capabilities of the sampling system. Foveon sensors NEED antialiasing filters to avoid aliasing so as this new approach.
The consequences of aliasing on Bayer sensors are much more noticeable though, because interpolation can additionally cause colour misinterpretation artifacts. Foveon and the new approach will only suffer monochrome moiré if a proper AA filter is not set and the sensor tries to capture high frequency patterns.
Regards