Fuji LX: No Microlenses No IR Filter ..

[bwcolor]

Here is a best guess summary of this new camera in a clear/concise format (all assuming that the blurred photos and other data associated with the leak are genuine):

Fuji LX Mirrorless

It seems to me that in order to meet promised full frame performance from an APS-C, or APS-H sensor that this camera will use the new organic CMOS sensor technology as mentioned in other threads and first seen in a 2008 Fuji Japanese patent.

This sensor should provide much better peripheral image quality while using very compact lenses and may be the new technology that gives Fuji a real competitive advantage.

 

[N.delaRua]

Here is the R&D pdf describing what an organic CMOS sensor is. Its a fun read:http://www.fujifilm.com/about/research/report/055/pdf/index/ff_rd055_004_en.pdf

Enjoy.

 

[Arjay]

The R&D report is very interesting as it is pointing out that such a design will

1. have a high apparent pixel aperture of 85% because the pixel electrode is occuping almost the entire area of the geometric pixel area,
2. its organic photoelectric conversion layer has such a high absorption coefficient that there's virtually no spectral crosstalk between neighboring pixels which helps to avoid having to use microlenses, and
3. the spectral response of the organic photoelectric conversion layer can be fine-tuned by a diligent choice of layer materials, so that the sensor can be made insensitive to IR radiation by design, making IR cut filters superfluous. This helps to avoid problems that arise with varying light incident angles which could cause a shift in IR cut wavelength and would thus generate inhomogenous color hue changes across the image.

To be quite clear however - this image sensor still has to use a Bayer filter pattern . So it's still very different from the Foveon sensor principle.

• Bullet point 1 could translate into good s/n figures even for high pixel count sensors. Does anybody know the apparent pixel apertures for conventional imaging sensors?
• Bullet point 2 would mean that microlenses are no more required. This would yield a lot more flexibility for lens design.
• Bullet point 3 could mean more color reproduction uniformity, and maybe a slightly improved apparent resolution.

Since the sensor still has to use a Bayer patterned color filter, some anti-aliasing filtering will still be required. 🙁

 

[willie_901]

A minor point: The Foveon system produces luminance ( not color) Moiré. Any time an analog process is digitized the Nuquist Nyquist–Shannon sampling theorem applies.

However the research PDF indicates the organic Bayer array has real promise.

 

[Jamie123]

Nerds



😉

 

[Matus]

Since the sensor still has to use a Bayer patterned color filter, some anti-aliasing filtering will still be required. 🙁

Well, not necessarily - there are already several cameras without anti-aliasing filters (Leica M9 and Ricoh GXR M-mount come to my mind) and most users are very happy about that.

I would love to see a Foveon sensor in such kind of camera too, but that technology seems to still have a long way to go - even the $$$ Sigma SD1 seems to be far from matching current CMOS performance in most aspects (minus detail on per pixel basis)

 

[iandiam]

Nerds



😉

Really interesting stuff, really. I agree, though. Heavy nerdery...🙂

Film, please!

 

[Sparrow]

.... never mind all that stuff, get your priorities correct, where's the Fiji Guys's un-boxing video?

 

[N.delaRua]

I just thought it was fun to read the R&D paper. Sounds like it has promise, but the organic layer could be a decent marketing tool as well as a actual functional improvement.

It just sounds cool... My CMOS is Organic... It has a soul, its made from carbon based materials, its also friendly to the environment and to my digestive tract like free range chickens. Those non-living Cannon and Nikon CMOS senors are soulless. I can already feel the warmth and life of an image captured on and organic sensor....

 

[bensyverson]

WOW.

This is basically digital film. A conventional CMOS sensor has a little light sensor (photodiode) at every pixel. Amazingly, this doesn't. Instead, it's a CMOS array of simple charge readouts, with an incredibly thin layer of organic* photoelectric film laid on top.

That's huge, for so many reasons. There's only a tiny gap (0.24 microns) between the electrodes to prevent them from touching, so the pixels are nearly as large as they could theoretically be. That translates to phenomenal low light performance.

Because the charge circuit is so simple, the overall noise is naturally low. And because there's no microlens array, there's no requirement for telecentric lens designs. There's no IR filter, so colors straight out of the sensor are more natural and require less processing. With less stuff in front of your pixels, resolution is at or near the theoretical peak.

Normally I would be skeptical of this making it to market, but the CMOS part of the design is much simpler than a CMOS image sensor, so it should be cheaper to fabricate. The film part of the design would be challenging for any company not named Kodak or Fuji. It would still cost millions to set up a production line for it, but Fuji could actually do it.

* One thing to note is that "organic" in this context means some kind of polymer, not "organic matter."

 

[bwcolor]

So, as pointed out previously, it is unlikely that Fuji would risk the introduction of a new sensor technology in a large sensor camera. If true, how are they going to give us full frame quality using an APS-C / APS-H sensor? Will they use a 5N sensor with micro-lenses and without AA filter? How do they get there, or are they just blowing smoke?

One last thought, this probably will be a relatively low volume camera, when compared to more mass marketed cameras. So, this might actually be a reasonable test platform. We would accept a high price for such a camera and it would certainly sell in numbers similar to the X100. Take the NEX-7 and add an extra $800-$1000 for the new tech sensor and I'm still onboard. I bet that most of you would also be there.

 

[bensyverson]

The question is, would Fuji risk killing this technology in the cradle by pricing the first body above $2000? I could see $2000 for the kit though...

 

[Trius]

Does this mean that when you retire the camera you can compost the sensor?

 

[stompyq]

edible sensor!!! wait..... no that can't be right......

 

[noimmunity]

"- Hybrid viewfinder... 18mm/2.0 pancake, 35mm/1.4,. Probably a zoom and a 60mm prime lens coming too."

This sounds excellent!!!

The organic CMOS sensor sounds, to my layman's ears, like it will provide a platform for M lenses with adapters that may be better-suited to corner performance etc, but with the array of lenses Fuji is planning, adapted lenses may be superfluous.

 

[Bobfrance]

Well Fuji's new orgasmic sensor already seems to be delivering. 😀

 

[Pherdinand]

the only thing is, you can only buy it in organic shops and therefore will cost a lot more.

 

[gavinlg]

One of the big things I see from this organic sensor is that the lenses will be able to be tiny. If you don't need microlenses, you don't need strict telecentric lens designs - e.g. you can make rangefinder sized lenses that are good optical performers.

If this organic sensor business is true, this will be one seriously good little system...

 

[Harry Caul]

I wonder if Leica will pony up and pay Fuji to make a FF version to replace their Kodak sensor.... low-volume, but high margin!

 

[noimmunity]

I wonder if Leica will pony up and pay Fuji to make a FF version to replace their Kodak sensor.... low-volume, but high margin!

Who cares about Leica?

What I wonder is rather how long before Fuji, Zeiss and Cosina team up to bring out a digital ZI/GS-M? :angel: