Facts and speculations: Update on the organic sensor

Facts and speculations:Update on the organic sensor

In 2014 i wrote this about the organic sensors and what role it could have in the future digital imaging market. Much of it still holds true and there has been very little news on the subject until last month. Now Panasonic has gone public with some more information about the technology itself and way to apply it to reduce motions blur and flicker when using what we today would call an electronic shutter.

First, lets ask ourselves why they go public at all with this! They could have just kept silent about it until they had an actual product to sell. There are a number of possible reasons for going public, everything from that they don't have the resources to finalise a product themselves to that they want to start a buzz or wanting to protect their patents. We can't really know what combination of reasons that made them disclose the information, but we can draw conclusions about what they are not planning. They do not intend to take the world by surprise, suddenly shipping cameras with organic sensors. That makes perfect sense too, the digital sensors of today are getting rather close to what is theoretically possible in terms of low noise and high DR. At some point in the not so far off future, the best way to compete in that market will no longer be by performance, but rather by getting the production cost down. Whenever an organic sensor would reach the market, that future would probably already be upon us and getting the prices down is all about volume and production efficiency and almost nothing about virgin tech.

With that said, lets go back to the announcements. What first strikes me is that they have upped the promised dynamic range from 88dB to 123dB by applying Fujis old Super CCD  SR II structure of mixing large and small photodiodes. From the experience of the Finepix S5 Pro we know this works remarkably well. That old sensor from 2006 with 6Mp of high sensitive and 6Mp of low sensitive was very popular by wedding photographers for its high dynamic range and still beats modern cameras in that department. It didn't deliver the same detail as a 12Mp sensor of the time, but still a lot more than a 6Mp one. 123dB of DR would amount to almost 41 stops, compare that with the best 24x36mm sensors of today that has just past 15 stops.

Further more, have a look at this passage: "In OPF CMOS image sensor, charge-storage function and photoelectric-conversion function can be set independently." If you look here and here. The orange rectangles called charge storage node and floating diffusion is the charge-storage function in the above quote. A floating diffusion is usually an isolated part of a chip that can store charge without being affected of what goes on in the rest of the chip. The photoelectric conversion is done by the film, the organic part of the sensor. Combine this information with: "Wide incident angle (60 degrees), high sensitivity, high saturation and highly-functional circuits due to a unique feature of OPF, in which an OPF for photoelectric-conversion and a readout circuits are independent." and focus of the last 5 words. This sensor can actually be built into 3 different parts that can work independent of each other. For example, one could have a large OPF with charge storage behind all the film, but the chip that actually does the readout could be a lot smaller and travel behind the film to cover the whole area. This would make it even more suitable for very large formats than i previously believed. Instead of the cost increasing linear to the sensor area compared to exponential for conventional sensors the cost of a twice the size organic sensor would be less than twice the price if one accept a longer readout time. For a large format photographer a few seconds of readout time is nothing, as long as the capture time is the same all over the image.

In the 2014 text, i also mentioned the state of the camera market as such. Not much have changed, but there are a few things worth mentioning.

1. There are now 645 full frame sensors available on the market. They are very expensive, but not that much more expensive than crop 645 sensors were a few years ago. Production has thus become more efficient and it will become even more efficient before we see an organic medium format sensor.
2. Sony acquired Toshiba's sensor department. Toshiba was important it was one of the few remaining competitors to Sony in the small frame sensor market. Nikon for example, can no longer ask if Toshiba can make a cheaper sensor than what Sony offers when they need a new one.
3. Canon is lagging even more behind in the DR department. They still have almost 50% of the market, but sooner och later this will catch up with them and they know it. Not upgrading their fabs to modern standard for 10 years was a very smart and brave decision and has so far not cost them that much while saving a lot of money, but the time brave turn into foolish is getting closer every day. 
4. Fuji film are slowly increasing their semi-conductor production. They just announced the establishing of a new plant. It's not a big one, the investment is about $8.4 million, but it tells us that they are expanding nevertheless. 
5. When i wrote the old text Fuji X system and the Samsung mirrorless systems were up and coming. Now Fuji X is established and Samsung looks to be abandoning their mirrorless effort.

Speculations

The facts that Panasonic is this public about the organic sensor, that the competition in the small frame sensor market is decreasing and that Canon has still not caught up with Sony in DR is opening up a slim chance that Panasonic is aiming for something bigger than i expected in 2014. I say slim, because the problem of getting the production cost down so early in the development is almost impossible. But somehow it does fit the facts and if Panasonic could get Canon onboard they would at least have the volume needed to actually aim for the small frame mass market.

Still, if that would become true, the impact on the small frame market would still be nothing compared to what it would do to the large and medium format markets.

Enough speculation, in a few more years, we may actually know something.