Imaging Resource takes a look inside Nikon's sensor design studios, solves Sony misconceptions

[ND64]

11 minutes ago, MdB said:

This comes at the expensive of high ISO noise. For some people one is an advantage and for others the other. That doesn't make a single clear cut 'better' sensor. 

 

Read noise at ISO 6400: A7R3> 1.05e D850> 1.2e

You can't see this difference in images. 

[MdB]

2 minutes ago, Eric Calabros said:

Read noise at ISO 6400: A7R3> 1.05e D850> 1.2e

You can't see this difference in images. 

And nor can you in the extended DR. Plus the A7R III has greater DR at the same ISO. Who really cares? 

[Andrew Reid]

@MdB

Go look at the D850 4K in full frame, A7R III and let me know what you see there as well!

[MdB]

3 minutes ago, Andrew Reid said:

@MdB

Go look at the D850 4K in full frame, A7R III and let me know what you see there as well!

Ok? I guess I must see Nikon watermarks that prove that it's a Nikon sensor? 

I'm not sure what you think your point is here? Both are horribly compromised designs when it comes to 4K, with lower resolution sensors being far better suited. 

How about you compare the flagship D5 in 4K against the budget A6300/6500 and A7 III. What an absolute joke! But I guess Nikon always have the better sensors because they are always better at 'designing' them? 

[no_connection]

If they designed the sensor and processor together with readout to make it "the best", why is the rolling shutter so bad? Did they wake up launch day and go "oops, we forgot about that".

It would make sense for them to be all about image quality and dynamic range but to drop the ball so bad when it comes to other parts makes you wonder if they just don't care about the moving part of moving pictures.

[MdB]

8 minutes ago, no_connection said:

If they designed the sensor and processor together with readout to make it "the best", why is the rolling shutter so bad? Did they wake up launch day and go "oops, we forgot about that".

It would make sense for them to be all about image quality and dynamic range but to drop the ball so bad when it comes to other parts makes you wonder if they just don't care about the moving part of moving pictures.

Almost certainly this part will be Sony's fault. Because, reasons. 

[Robert Collins]

22 hours ago, Andrew Reid said:

Next innovation to be widely used is Quad Bayer, I think. It allows 2 exposure sensitives in one pixel, so HDR on chip, essentially. That's design, not just manufacturing. Patents are key and Nikon's team have quite a few.

Canon with Dual Pixel AF have more.

Design is important guys.

Noone denies that Nikon designs their own chips as does Fuji for that matter but Sony is the one producing them.

If you take the stacked cmos sensor in the Sony A9, it is clearly far more advanced than any other camera manufacturer - camera sensor.

I am sure Nikon can design a FF sensor with a stacked sensor but who else can manufacture it? Samsung maybe but they dont seem interested in this space.

If you want an idea of how the technology is likely to progress, look into Pixim (seawolf security sensors) which Sony acquired in 2015. In that technology every individual pixel has its own exposure (as opposed to single or dual exposure over the whole sensor) allowing ultra wide dynamic range. Currently it is restricted to 1mp but in a few years....

BTW I dont see Sony's stranglehold on the camera sensor business as a positive thing, it is the sort of monopoly that will ultimately be to the detriment of consumers in terms of both price and innovation.....

[Nikkor]

Do you guys hold stock options at Sony? 

[MdB]

16 minutes ago, Nikkor said:

Do you guys hold stock options at Sony? 

No. I just use my ability to interpret BS. 

Funny considering your own title on this forum...

[Nikkor]

3 minutes ago, MdB said:

No. I just use my ability to interpret BS. 

Funny considering your own title on this forum...

I like my nikon glass, but I mostly use a phase one. Funny, isn't it?

Fanboyism is nothing new,neither is forum narcissism.

[Andrew Reid]

Camera fanboys are as bad as Trump fanatics.

[Mokara]

The sensors made by Sony for Nikon are not the same as the ones Sony uses, although they may seem so superficially. The way it works is that the client would design architecture that is compatible with the manufacturers processes and essentially send them the spec. The manufacturer would then make it based on their own technology and what is permissible to use. Significant changes are routine at this point. So, the sensor may include features normally seen in Sony sensors because, well, that is what Sony makes and how they make them. They will NOT include features that are outside of Sony's manufacturing processes. In a contract situation like this the client will set general specifications but exactly how those specifications are achieved is left to the manufacturer to do (and usually those specific manufacturing processes are proprietary to the manufacturer). Nikon would not have designed BSI for example, that part would have been done by Sony. Likewise the sensors would include focusing points that are allowed by licenses Nikon holds. For example, Sony can make DPAF sensors (they have for Samsung, who hold a license from Canon), but they can't do that for Nikon since they DON'T have a license from Canon (and never will).

All that Nikon design does in this is come up with a workable spec that is compatible with Sony's processes and allowable under licences/patents held by Nikon. They talk to engineers at Sony's manufacturing facilities and figure out what can be done in general. Then they will come up with small scale prototypes made in the R&D labs and send those over to Sony along with the design they used. The engineers at Sony then modify that to be compatible with their processes (likely significantly changing aspects of the design in the process) and practical to make. If Nikon is OK with the final prototypes that come out of that process, it goes into production, with all of the sensors resulting from that going to Nikon (Sony will be contractually barred from selling them to anyone else). This is how everything else is made in the tech industry, so no doubt the same holds true for Nikon/Sony sensors.

[Andrew Reid]

What tends to be the fashion in 2018, is that people ignore the facts, and keep repeating their divine personal opinions like a religion, repeating their personal mantra again and again until they are blue in the face and the discussion about the facts is long dead. I think our culture has gone seriously wrong.

1 hour ago, Mokara said:

All that Nikon design does in this is come up with a workable spec that is compatible with Sony's processes and allowable under licences/patents held by Nikon. They talk to engineers at Sony's manufacturing facilities and figure out what can be done in general. Then they will come up with small scale prototypes made in the R&D labs and send those over to Sony along with the design they used. The engineers at Sony then modify that to be compatible with their processes (likely significantly changing aspects of the design in the process) and practical to make.

And what qualifies you to have insider knowledge about the process to that degree of detail? You're coming up with this stuff, from literally a speculative nowhere.

[Andrew Reid]

https://www.nikonprecision.com/products-and-technology/

5nm.

[MdB]

3 hours ago, Andrew Reid said:

What tends to be the fashion in 2018, is that people ignore the facts, and keep repeating their divine personal opinions like a religion,

Says the guy just repeating the Nikon PR hype train. YOU have no experience on this, you wrote an article repeating another article. If anyone says otherwise apparently they are fanboys, trump or divine personal opinions. It's called a discussion Andrew, you might want to look it up. 

And PS people have been around this industry a while, perhaps some of us are better at sorting the wheat from the chaff when it comes to marketing hype. 

Let me put it this way: The only way ANYONE is getting a completely custom sensor design is to make it themselves (at huge cost). Like Canon and Sony do. Otherwise they are getting a stock design with some tweaks to suit their implementation. 

The sad thing is, nobody who is suggesting that these are Sony sensors (with OEM tweaks) are saying that that is a GOOD thing. Nobody is saying 'NO they are Sony sensors because Sony are the BEST', which is what you are implying with all the nonsense 'fanboy' talk. In fact I would argue it is the Nikon fans saying 'NO they Nikon sensors because Nikon are the BEST' (exhibit: your comparisons between sensors, which does zero to prove who designed them, only who use them). Nobody thinks it's a good thing, or the best, or whatever fanboy response you think they are having. Its just not PR peppered BS.

Why does every Nikon fanboy think their company is so much more special than Phase One or Hasselblad? Because the marketing dept wants them to feel that way. Those companies proudly boast they are using sensors from the best manufacturer in the world (which is largely true). But Japanese pride and the fact that Nikon and Sony are competitors stop Nikon from doing that. It is really that simple. ESPECIALLY when they are on the eve of launching a competing mirrorless system to Sony who are their lunch. 

 

I hate to use the term 'fake news', but fake bloody news Andrew! 

[ND64]

17 hours ago, MdB said:

And nor can you in the extended DR. Plus the A7R III has greater DR at the same ISO. Who really cares?

With ISO 64 you can expose for an extra 2/3 of a stop without clipping highlights. Thats "real" benefit. 

 

16 hours ago, Nikkor said:

Do you guys hold stock options at Sony? 

If they had stock, they knew Sony Semiconductor is completely independent from Sony Imaging, and has strategic relationship with Nikon. 

 

6 hours ago, MdB said:

Why does every Nikon fanboy think their company is so much more special than Phase One or Hasselblad? Because the marketing dept wants them to feel that way. 

Funny thing is Nikon's marketing is the worst in the entire camera/optics industry. They let Dave to visit their not-visited-before R&D rooms, but what Dave wrote is just his reaction to what he experienced. Simple as that. 

[Robert Collins]

30 minutes ago, Eric Calabros said:

If they had stock, they knew Sony Semiconductor is completely independent from Sony Imaging, and has strategic relationship with Nikon. 

 

Or possibly not....

https://petapixel.com/2017/03/22/sony-keeps-best-sensors-cameras/

[Andrew Reid]

10 hours ago, MdB said:

Says the guy just repeating the Nikon PR hype train.

You can go on contradicting the man who has visited the Nikon lab until you've worn yourself out and you need a lie down, but the fact is, you have never been to the lab, you have never been to a Sony factory and until you do, you will keep contradicting the facts in Dave's article until you're blue in the face. Your knowledge is mishmash of third-hand speculative nonsense sourced by reading rumors sites.

[MdB]

2 hours ago, Andrew Reid said:

You can go on contradicting the man who has visited the Nikon lab until you've worn yourself out and you need a lie down, but the fact is, you have never been to the lab, you have never been to a Sony factory and until you do, you will keep contradicting the facts in Dave's article until you're blue in the face. Your knowledge is mishmash of third-hand speculative nonsense sourced by reading rumors sites.

Mate, READ the article. I don't need to do anything of the things you sugggest because it is literally there in black and white. The guy writing the article even says he doesn't know enough technically to understand some actually pretty simple things. The whole situation was CREATED by the Nikon PR team. He states specific technologies as being used in Nikon sensors, but doesn't state that that technology COMES from Nikon or even has anything to do with Nikon (other than its appearance in the sensors they happen to use). When he does explicitly state Nikon have some actual input, guess what it is about?! Packaging and sensor toppings. 

Just read it again. I mean how long does he go on about LVDS as if it is 'special' to Nikon, just because they happen to use it? 

Im frankly stunned at just how literally people seem to be reading it. Dominant readings all day everyday from the fanboys who already believed it, but just needed that 'credible' article to prove it was all true. 

You stand there and defend it, but you haven't once been able to answer any of my questions.

7 hours ago, Eric Calabros said:

With ISO 64 you can expose for an extra 2/3 of a stop without clipping highlights. Thats "real" benefit. 

No, that's an assumed benefit. It's actually considerably less than that. With a greater loss in high ISO. 

7 hours ago, Eric Calabros said:

Funny thing is Nikon's marketing is the worst in the entire camera/optics industry. 

According to whom? 

7 hours ago, Eric Calabros said:

They let Dave to visit their not-visited-before R&D rooms, but what Dave wrote is just his reaction to what he experienced. Simple as that.

Yes, that is what happened. He wrote about a bunch of things Nikon have in their sensors, many of which have nothing to do with Nikon, but the way it is written assumes that it does. He writes about LVDS as if it were a Nikon thing, solely for Nikon sensors. Which is either poor writing, or the guy actually didn't know any better. So which is it? Doesn't really matter. The point is there is nothing in the article that points specifically to Nikon actually having much say about anything in the 'design' process. Just a lot of jargon about the amazing technology Nikon uses (thanks mainly to the underlying Sony base designs). 

[webrunner5]

"However, in 1992 Apple Computer needed a method to transfer multiple streams of digital video without overloading the existing NuBus on the backplane. Apple and National Semiconductor (NSC) created QuickRing, which was the first integrated circuit using LVDS. QuickRing was a high speed auxiliary bus for video data to bypass the NuBus in Macintosh computers. The multimedia and supercomputer applications continued to expand because both needed to move large amounts of data over links several meters long (from a disk drive to a workstation for instance)." Wiki

So it Ain't a new thing as they say.