[Tesla]

Ricoh's new Aero Bright II lens coating

Ricoh also announced a significant update to their Aero Bright lens-coating technology. Aero Bright is Ricoh's version of nano-particle coating, a technology that's had a big impact on lens flare over the last few years. Manufacturers tend to be pretty tight-lipped about the specifics, but we do know that there are various flavors of nano-structure coatings for camera lenses. They all basically attempt to provide a more gradual transition of refractive index at air/glass boundaries. It's the abrupt change in refractive index that produces reflections; no index change, no reflection, big, abrupt index change, lots of reflection. If you can make the refractive index change gradually instead of abruptly, you can dramatically reduce the reflections.

Ricoh reveals pixel-shifting high-resolution mode: Coming soon to a camera near you? (second half of article)

[Adam]

Cool stuff! It sounds like Ricoh is hard at work on R&D. The pixel shift technology will likely be used to increase sharpness (rather than increasing the image resolution, as per our CP+ interview).

The article also talks about the new clarity enhancement feature in the K-S2.

Ricoh's new "clarity" enhancing function employs a stack of synthetic images, each having progressively lower spatial frequency content. Looking across these images, the algorithm is able to determine areas where local contrast enhancement would improve the image.

This is a really fancy way of saying that the camera has built-in auto shadow/highlight adjustment. I've actually implemented a similar algorithm in a programming class, it was very interesting.

In the K-S2, it works well for good out-of-camera results, but I've found it to clip the shadows a little to enthusiastically.

[Spodeworld]

Sounds so interesting. Wonder what the resulting image will look like compared to those out of the K3 today. Also, as it takes the image moving it over 4 pixels where 1 exists today, I wonder if it will require an exposure four times as long, or if is cumulative and there is no change in exposure time.

[ZeljkoS]

Most interesting part to me was the one about "Taking care of their customers", implying we could see this option inserted into older models, via firmware update.

But wait, there's more! When I asked a Ricoh executive whether this pixel-shifting image enhancement tech might come to any of their existing SLRs via firmware updates, he of course said that he couldn't comment on future developments. He did say, though "But Ricoh believes in taking care of their customers" - and encouraged me to quote him on that. :-)

Not holding my breath, but if they would do this for owners of previous generation cameras ( K-5II/K-5IIs) that would be simply amazing!

[Spodeworld]

Wow, there's a thought that didn't even cross my mind. Too awesome to think about in case it doesn't happen.

Most interesting part to me was the one about "Taking care of their customers", implying we could see this option inserted into older models, via firmware update.



Not holding my breath, but if they would do this for owners of previous generation cameras ( K-5II/K-5IIs) that would be simply amazing!

[goubejp]

This pixel shift technonogy has the advantage of removing moiré (ie false color produced by color interpolation). There can be some interférences with patterns on images, but this result on luminance variations less visible than moiré

[Uluru]

If true, this can deliver ideal Foveon-like colour accuracy. Because 1 pixels shift in each direction grabs R, G and B data for each pixel, without interpolation of any kind which is common to Bayer design. Which is amazing.

I also wonder why this wasn't introduced before, because it is more logical to implement than the AA-simulator. It is far less complicated too.

Also, the graph for the Aerobright II looks uncanny. It is virtually flat, which by definition is called a miracle.

[thibs]

If true, this can deliver ideal Foveon-like colour accuracy. Because 1 pixels shift in each direction grabs R, G and B data for each pixel, without interpolation of any kind which is common to Bayer design. Which is amazing.

I also wonder why this wasn't introduced before, because it is more logical to implement than the AA-simulator. It is far less complicated too.

Also, the graph for the Aerobright II looks uncanny. It is virtually flat, which by definition is called a miracle.

I don't think this is so easy to implement. You have to be very careful to move the SR by EXACTLY one pixel or you're asking for more problems than you're trying to solve. And one pixel left or right will be potentially different for each sensor so this is very camera-specific to implement.

[Zygonyx]

Plus it does also depend on focal values, in the same way as sensorshift.... basically you have to interpolate two movements, the second related/ depending on the first movement which is implemented with a largely variable amplitude.

[RobA_Oz]

Plus it does also depend on focal values, in the same way as sensorshift.... basically you have to interpolate two movements, on a relation depending on the first movement which is implemented with a largely variable amplitude.

I'm not sure what you're getting at, here, when you use the term "focal values". I can understand why Focal Length is an important input for Shake Reduction, but in this case I would have thought that it was irrelevant, unless you mean that shake reduction would have to be overlaid on pixel shift because of the time between successive shots, and the possibility of camera movement.

[goubejp]

You have to shoot 4 images of the same subject, so it'll take more than 0,5 s, so it is applicable for still subjects only. Even for landscapes, as trees and leaves may be moved by the Wind, it can be irrelevent. More for studio shots IMO.

[Uluru]

I don't think this is so easy to implement. You have to be very careful to move the SR by EXACTLY one pixel or you're asking for more problems than you're trying to solve. And one pixel left or right will be potentially different for each sensor so this is very camera-specific to implement.

It is not too difficult once they have AA sorted out. They already can move sensor in tiny sub-pixel movements to eliminate moire, so It is indeed an extension of the same technology.

[derekkite]

Since the K-5 at least the ability to shift the sensor to follow relative star movement has existed.

The difference with this or AA filter or any other implementations is the position has to be maintained and moved across exposures.

I suspect that the changes in the code required are quite extensive, but worth doing for other purposes as well.

[Zygonyx]

I'm not sure what you're getting at, here, when you use the term "focal values". I can understand why Focal Length is an important input for Shake Reduction, but in this case I would have thought that it was irrelevant, unless you mean that shake reduction would have to be overlaid on pixel shift because of the time between successive shots, and the possibility of camera movement.

Yes you get it : anti AA is imho something less accurate because something "looser" than pixelshift, but both are used whilst sensorshift compensation is acting which makes pixelshift really difficult to achieve.

[Uluru]

I'm not sure what you're getting at, here, when you use the term "focal values". I can understand why Focal Length is an important input for Shake Reduction, but in this case I would have thought that it was irrelevant, unless you mean that shake reduction would have to be overlaid on pixel shift because of the time between successive shots, and the possibility of camera movement.

It is important, because one pixel at 15mm FoV, and one pixel at 300mm FoV, correspond to wholly different amount of magnification.