Tuesday, February 02, 2021

Sony Global Shutter Sensor Designers Interview

Sony publishes a large interview with its global shutter image sensor designers titled "Pregius S—to be a bridgehead for pioneering new markets such as robot arms through the pursuit of high speed, multi-pixel as well as size reduction." Few quotes:

"...there is a question why CIS-based GS did not exist before, and the answer is partly because a new structure was needed to add the GS function to the previous CIS. To realize this, a diversity of technology was required for microfabrication, such as memory storage potential structure design, design to embed it in the photodiode, technology to transfer stored information without a loss, pixel designing technology, micro-processing technology, etc., to a greater extent than that required for conventional CIS.

Our competitors produced CISs with superior energy efficiency or speed (higher frame rates), but we heard that many users complained of the noises compromising the picture quality.

Against this backdrop, we decided to develop a CIS for GS, Pregius, leveraging our expertise in pixel technology which we cultivated through SSS Group’s pioneering RS mode CIS and CCD image sensors.

...finally, we delivered increased multi-pixels and size reduction while at the same time achieved to eliminate noises, which competitors could not solve.

Pregius by SSS Group had impressive responses for its low-noise picture quality that was far superior to its competitions, but the market for industrial machines was moving fast, and there was an increasing demand for smaller (miniaturized) and faster GSs with more pixels.

The back-illuminated structure enabled enhanced light saturation and increased pixels, but the most challenging task for development of Pregius S was to shield the pixels so that the light did not seep into unintended parts."


  1. Credit to Sony for implementing GS in a stacked BSI structure. It certainly improved pixel size. But GS (or snapshot) shutter was around from the very beginning by extending the micro-CCD in each pixel by one stage and used CDS for low noise.

  2. Ah, but this is so Sony. They wrote an entire article with no performance data in it.

    1. Albert Theuwissen - Harvest ImagingFebruary 3, 2021 at 11:50 AM

      That's the real marketing strength of these companies : getting the attention but don't tell anything.

    2. Hi Albert, what type of information do you miss most?

    3. I do miss numbers and figures. Especially about the PLS not that much information is out in the field.
      If a company claims that they are the best in a particular field, e.g. noise, then why do they still hide the hard numbers ?

  3. Its not the pixel performance. The reason for the success of the Pregius is the family approach. Lets be honest - industrial is not a astro application. The sensors are all good enough. Industrial design in hurdle is the main factor. The fact that you can design in one sensor and use many across the same interface is what makes the pregius attractive - not its pixel performance. I am surprised to see again and again (also here in the comments...) this focus on pixel performance. The integration of a sensor (or also camera from a industrial camera manufacturer) is a huge effort. The pixel/sensor properties get designed into the recipes and machine generations and all the small mechanisms that define the behaviour of machines e.g robots across the generations. You dont take a camera of vendor a and replace it by vendor b and get the same result even if both run on 'standards' like genicam, (not even if both use the same sensor). This is also one factor that justifies the high prices in industrial market - the lockin effect. If you designed in a family they can charge you what they want because you have no easy alternative.
    The main benefit of sony is that they have different sensors with the same optical behaviour and the same firmware behaviour. Some even can use the same PCB (but this is what i consider secondary to firmware, even though this is what the vendors say makes a 'family'. for example: if 2 sensors use the same PCB but one has a bug where every second line is shifted 1 pixel and the second/later sensor fixes this shift - it is NOT a "brother" sensor even if it uses the same PCB. It is a "cousin" - same pcb but different firmware. We had this case (not with sony...), this was the argument why we did not use the quite attractive second sensor.
    The problem of for example the new ams sensors competing Pregius (as discussed in a post few days back) is not the pixel performance. it is the missing family. You never just need 1 type of sensor. If you only have one, it has to be superior in one parameter (the Gmax0505 was such an example, no other sensor filled cmount space with so many small GS pixels than this one. This justifies design in).

    1. Albert Theuwissen - Harvest ImagingFebruary 3, 2021 at 11:48 AM

      Great comment.
      My first claim : "It is very difficult to kick your sensor competitor out of an existing camera, but the same is true for your competitor if he/she wants to kick you out of an existing camera !"
      My second claim : "It is difficult to get your sensor in a new camera, but once there, it is sometimes even more difficult to step out !"

  4. Lot of comments from engineers. The good product is not the one with the best performance, it is the one that sells. If you have designed a technically superior product but failed the market, I am sorry for you.

    1. Like an old (engineering) professor of mine used to say:

      "All (young) engineers should be forced to work in sales for at least 1 year...only then are they allowed to return to FFTs & integrals.


    2. This is true - but also true in opposite direction. Sales guys should work for example in a complex software project under production pressure for a year before they are allowed to wear their suits ;-)
      Sales guys that never did this tend to underestimate technological complexity (in multiple dimensions, e.g product and tool generations across product lifecycles) and the impact of changes/requests. Developers on the other hand also tend to underestimate a lot of factors, also they tend to overestimate the importance of details (pixel performance for example ;-)
      One of the secrets of successful teams or companies is the mix of people/skills.


All comments are moderated to avoid spam and personal attacks.