CCD / CMOS benefits and drawbacks.

[mjh]

Advertisement (gone after registration)

So do they count the 'blind' pixel as one of the 29Mpixels? So it is really a 14.5Mpixel sensor?

No, the MP figure specifies the number of photosites; pixels just serving for storing electric charges don’t count.

[mjh]

If you want FF CMOS on a rangefinder with short flange-focal distance wide angles it will need some radical design including a backlit design and microlenses.

Whether this becomes an issue also depends on the pixel pitch; that’s why backlit designs are employed for small sensors in compact digicams where the pixel pitch is now less than 1.5 microns.

 

Whether or not a backlit design will be necessary (I suppose the microlens shifting in the X100 is already quite extreme and it didn’t require BSI technology), those sensors will be developed if they are not already under development. Sooner or later there will be FF EVIL cameras and since it isn’t likely that these will use near telecentric lenses, there’s an issue that needs to be solved. And thus it will be solved.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[colorflow]

Sony did a custom microlens design for a single, low-volume model, the X100 – there is no other camera using that sensor. If Leica were to order a custom design, the number of units ordered wouldn’t be that much different. Obviously a FF sensor would cost much more, but it costs much more for anyone so that’s beside the point.

 

That is my point it would cost much more for anyone. So who would do this for Leica? It is a much more difficult & costly problem. Further speculation is pointless. Time will tell.

[mjh]

So who would do this for Leica?

Kodak does when CCDs are required; the question is who will do the same with CMOS technology. Previously we didn’t know whether Sony’s semiconductor division was above doing custom designs for a (relatively) low-volume production run. Now (because of the Fuji X100) we know. Surely there are other sources for a CMOS sensor but Sony shouldn’t be discounted.

 

And obviously any custom FF sensor design for an M was and will be expensive; in that respect nothing will change, whether it is CCD or CMOS.

Edited June 22, 2011 by mjh

[jaapv]

And, as I said, there is a Leica-Sony connection through Fujitsu who are the mastermind behind Sony's sensors and the provider of Leica's Maestro chip.

[colorflow]

And, as I said, there is a Leica-Sony connection through Fujitsu who are the mastermind behind Sony's sensors and the provider of Leica's Maestro chip.

 

Interesting, why do you think Fujitsu is the mastermind? Sony has been doing cutting edge CCD and CMOS for a long time. Possibly before anyone else.

[jaapv]

Advertisement (gone after registration)

It is easily googled - there is an advisory agreement between Fujitsu and Sony regarding sensor technology. Fujitsu advising Sony that is.

[colorflow]

It is easily googled - there is an advisory agreement between Fujitsu and Sony regarding sensor technology. Fujitsu advising Sony that is.

 

I only know Sony is outsourcing CMOS production to Fujitsu. Would you kindly provide a link?

[jaapv]

Regoogled: I think you are right I and I got it wrong. Sony has Fujitsu as producer. It does not make any difference to my remark about the connection between the two in the field of camera sensors however.

Edited June 22, 2011 by jaapv

[colorflow]

Kodak does when CCDs are required; the question is who will do the same with CMOS technology. Previously we didn’t know whether Sony’s semiconductor division was above doing custom designs for a (relatively) low-volume production run. Now (because of the Fuji X100) we know. Surely there are other sources for a CMOS sensor but Sony shouldn’t be discounted.

 

And obviously any custom FF sensor design for an M was and will be expensive; in that respect nothing will change, whether it is CCD or CMOS.

 

Fuji's projected volume for a $1200 camera is about the same as a $10K+ camera? Highly implausible.

 

Sony is not a niche player. Their fabs and R&D are not geared for that. In any case they have more work than they can handle right now. In the longer term, they may do it if its just a spin-off from the direction they are going anyway. In the short run Leica needs a niche player like Kodak (in semiconductor fabrication) to do their FF M CMOS. Who knows...

[mjh]

Fuji's projected volume for a $1200 camera is about the same as a $10K+ camera? Highly implausible.

The X100 is a niche camera. It’s not even a system camera such as the M10 would be. How many people have the least interest in an expensive camera with a fixed lens and a fixed focal length when you could get a NEX for less? There’s a lot of interest in the X100 among Leica users because Fuji is targeting photographers that would naturally also be interested in Leica products. But that users of one niche camera show great interest in another niche camera doesn’t imply that the camera will sell in great numbers (which Fuji isn’t expecting anyway). And the nature of the sensor is such that its microlens pattern is optimized for the lens of the X100 so there’s no hope Sony could sell it to anyone else, just as Kodak couldn’t hope to sell the sensor of the M9 to anyone else. So now we know that Sony is not above accepting a low-volume custom order.

[IkarusJohn]

That specific sensor is FF 35mm. It is 29 meg output, not 14 meg. 29 meg is 29 meg. I don't know of any consumer camera that is FF, so I'm not sure what this sensor would be for other than something like the M9. It seems no more limited than any other CCD, just higher resolution ... in fact less limited if the Specs are correct from Kodak:

 

Kodak states the KAI-29050 is one of the first sensors devices to use the Kodak TRUESENSE color filter pattern claimed to provide 2X to 4X increase in light sensitivity by adding panchromatic pixels to the standard RGB elements that form the sensor array.

 

To me that sounds like the higher ISO some want with the IQ of a CCD like the M9 provides ... without using a CMOS with an aggressive AA filter to mitigate noise.

 

It's all a moot point ... Leica will use a CMOS sensor in the next M digital, so the technology must exist, and Leica must see that technology as not compromising the IQ their buyers expect.

 

We'll see.

 

-Marc

 

Hi Marc,

 

I'm not sure I follow your logic. It sounds like this new Kodak CCD sensor (or a further development of it) could be just what Leica needs for its M10 (which I guess just moved some years out with the release of the M9-P).

 

So why do you then make the comment that Leica will use a CMOS sensor in the next M digital?

 

Did I misunderstand you post, or am I missing something else?

 

Cheers

John

[mjh]

It sounds like this new Kodak CCD sensor (or a further development of it) could be just what Leica needs for its M10

The KAI-29050 is just another sensor in a line of products for industrial applications (“demanding applied imaging applications such as industrial inspection, aerial photography, and security”, to quote Kodak) where it doesn’t matter when you need active cooling to keep the sensor from overheating. This family of sensors has been on the market for years (there was the KAI-11000 with 11 and the KAI-16000 with 16 MP) and none of these has ever made it to a camera in the usual sense of the word. Over the years, whenever Kodak announced the availability of one of those KAI sensors people got excited, assuming there would soon be cameras using it. It never happened.

[IkarusJohn]

The KAI-29050 is just another sensor in a line of products for industrial applications (“demanding applied imaging applications such as industrial inspection, aerial photography, and security”, to quote Kodak) where it doesn’t matter when you need active cooling to keep the sensor from overheating. This family of sensors has been on the market for years (there was the KAI-11000 with 11 and the KAI-16000 with 16 MP) and none of these has ever made it to a camera in the usual sense of the word. Over the years, whenever Kodak announced the availability of one of those KAI sensors people got excited, assuming there would soon be cameras using it. It never happened.

 

Ah, I see. Thanks Michael.

[jaapv]

So why do you then make the comment that Leica will use a CMOS sensor in the next M digital?

 

Because there is a very strong undercurrent amongst Leica-watchers that this will be the case

[Wozza604]

Sony on the other hand is an old hand at producing CCD sensors, and could do it for Leica if need be. In their professional cinema applications and joint ventures with Panavision, Sony produces professional CCD chips. Their home movie cameras have traditionally also been CCD powered, though I'm sure they're mostly CMOS by now.

 

The Sony sensors for cinema applications aren't "Full Frame" as we know it though. A traditional film gate in a movie camera is smaller than in a still camera. The main reason they use CCD is because it doesn't have the rolling shutter issue that CMOS cameras (like the Red One) have. The CCD chips used in cinema applications have completely different design requirement, to the point where they're not really comparable.

[IkarusJohn]

Because there is a very strong undercurrent amongst Leica-watchers that this will be the case

 

I get that, but I was wondering what it is based on.

 

Cheers

John

[noirist]

I don't think the theoretical advantages of CCD vs. CMOS matter to anyone except a camera sensor designer. As photographers, we care about the actual measured performance of a camera we can buy. How much dynamic range? resolution? signal-to-noise under varying levels of illumination? power consumption? data rates?

 

And by those measures, the best full frame CCD sensor in a commercially available camera (in the M9) is soundly beaten in every respect by the commercially available full frame CMOS sensors. It's even beaten in many respects by some smaller APS-C size CMOS sensors in budget DSLRs, like the Nikon D5100. Take a look at the camera sensor rankings on dxomark.com for details. So whatever hypothetical advantages that CCD might have versus CMOS are completely obviated by the very real advantages that actual CMOS sensors have over actual CCD sensors.

 

With the M9-P about to be announced and the talk here that this may be the last opportunity to get a CCD sensor M, can anyone explain the advantages of CCD over CMOS?

I've heard that the advantages with a CMOS sensor are better low-light performance and better battery performance but what are the plus points for CCD?

 

Thanks,

 

Mike.

[Wozza604]

I don't think the theoretical advantages of CCD vs. CMOS matter to anyone except a camera sensor designer. As photographers, we care about the actual measured performance of a camera we can buy. How much dynamic range? resolution? signal-to-noise under varying levels of illumination? power consumption? data rates?

 

And by those measures, the best full frame CCD sensor in a commercially available camera (in the M9) is soundly beaten in every respect by the commercially available full frame CMOS sensors. It's even beaten in many respects by some smaller APS-C size CMOS sensors in budget DSLRs, like the Nikon D5100. Take a look at the camera sensor rankings on dxomark.com for details. So whatever hypothetical advantages that CCD might have versus CMOS are completely obviated by the very real advantages that actual CMOS sensors have over actual CCD sensors.

 

I'm not sure I agree with you there. Just because a camera gets higher DXO scores does not mean that it takes better photos. My NEX-5 has a higher DXO Mark than my M8, but my M8 produces far more pleasing photographs with the same lenses. In fact, I prefer the images from my DP2 as well.

 

If, as photographers, all we cared about were actual measured performance of a camera, none of us would own a single digital Leica camera. 35mm film also falls short of the best digital cameras in a lot of way, especially sharpness, but to my eye often produces a more pleasing image. The technical spec race with cameras is more about convincing people to upgrade so they can sell more products and has very little to do with producing better images. But to some, I suppose photography is more about pixel peeping than about composition. Sad really.

[Advertisement]

Hallo,

Look here: Leica M

[xrogers]

As photographers, we care about the actual measured performance of a camera we can buy.

 

...Take a look at the camera sensor rankings on dxomark.com for details....

 

As a photographer, I care about the actual performance of a camera I can buy, and don't much care about measurements.

 

Dxo's one-number-tells-it-all ranking assumes a weighting of attributes that may not match with how you shoot. It doesn't take other relevant factors into account (the absence of an AA filter, or amazingly noise-free shadows at low iso, or non-sensor attributes like size and weight, or the ability to mount a Summicron).

 

Dxo provides a sometimes useful (but often abused) data point. That is no substitute for shooting a few photos with cameras that interest you, and then picking the one that will make it easiest to get the photos you want.

 

Until later,

 

Clyde