Did Canon criple the 5D Mark III sensor on purpose?

[photogaz]

Judging by the 1D X samples, the high ISO performance is outstanding and a true 1 stop improvement over the 5D Mark III. Something that strikes me though is 4 Megapixels can't make that much difference.

Canon opted to put a nice new focus system in the 5D Mark III, but it seems to avoid people not buying the 1D X and pro's opting for a Mark III instead, they limited the ISO performance boost. With a boost there wouldn't be much between them but now the 1D X is more appealing with its ISO quality jump!!

 

[wockawocka]

It's not about the lack of pixels but the space it affords them.

The pixels are what? 10% bigger than the 5D3's?

ISO stops = 100,200,400,800,1600,3200,6400,12800,25600,51200 - 10 stops.

10% bigger pixels = extra stop of performance.

Hardly scientific and probably incorrect but we should really think about this before asking.
The flagship body will have the best of everything in, best circuitry, best capacitors, best chips, everything better.

There's a reason why the 5D3 is half the price.

 

[briansquibb]

I would hazard a guess that Canon is pitching against the D4.

I dont believe the 1D4 was impacted by the 1DS3/5DII - I suspect the same will be case with the 1DX and the 5DIII - totally different use and therefore userbase

 

[Razor2012]

Let me see, the $3500 camera doesn't quite match the performance of the $7000 camera. Do I really need to ask why?

 

[heptagon]

Let me see, the $3500 camera doesn't quite match the performance of the $7000 camera. Do I really need to ask why?

Yes, you need to ask why. What is it that makes one camera perform vastly better than another? Why are Sony sensors better while the cameras using them are cheaper? Why does the 7D use the same sensor as the 550D? Because the price is not in the sensor and the 7D doesn't make better pictures once the light gets through to the sensor than the 550D. So we have to ask the question what makes the sensor better, even if the answer is that it is cooking the raw data. Let's find out.

 

[SandyP]

I got a better question: why do people ask such stupid questions that are useless, and will get absolutely no where? Answer: N/A

 

[Tammy]

No. The 5D series is typically one stop under the performance of the 1D series. This is acceptable, especially given the price. The 5D2 had the same MP as the 1Ds III but wasn't crippled to perform slightly worse etc.

 

[LetTheRightLensIn]

Judging by the 1D X samples, the high ISO performance is outstanding and a true 1 stop improvement over the 5D Mark III. Something that strikes me though is 4 Megapixels can't make that much difference.

Canon opted to put a nice new focus system in the 5D Mark III, but it seems to avoid people not buying the 1D X and pro's opting for a Mark III instead, they limited the ISO performance boost. With a boost there wouldn't be much between them but now the 1D X is more appealing with its ISO quality jump!!

The 1DX sensor is an absolute revelation at high ISO. Looking at the flower comparison shots, to beat the 5D3 by a solid 2/3 stops is just insane. This is the all-time low light king ever and it's got to be absolutely at the complete theoretical limit for what is even possible from this sort of design. I'm stunned.

OTOH you did bring up a good point. Sadly, it seems what one connected-one said is true. He said months ago something about how we'd see that Canon marketing thought that they could leave the sensor sort of the same since they were giving us so much in terms of the body specs and that they saved the new process for the 1DX sensor only and wanted a key 5D3 design goal to make it the highest profit per copy yet for a 5 series body. I guess the D800 sensor taught them.

Not that the 5D3 sensor is bad at high ISO though, it's one of the best. But the 1DX is insane. Wow, I mean it's hard to absolutely be sure from that sort of test, it's never more than a rough guess, but I think it's pretty safe to say that at worst the 1DX is as good as any other DSLR at high ISO and it's likely the best in history and perhaps so near the limit that nothing of any sort of even vaguely standard design will beat it.

I wonder how the 1DX will fair at low ISO though. How will the max DR be?
That is where the 5D3 really looks old school.

 

[LetTheRightLensIn]

It's not about the lack of pixels but the space it affords them.

The pixels are what? 10% bigger than the 5D3's?

10% bigger pixels = extra stop of performance.

Hardly scientific and probably incorrect but we should really think about this before asking.

pretty sure that is definitely incorrect ;D

 

[LetTheRightLensIn]

No. The 5D series is typically one stop under the performance of the 1D series. This is acceptable, especially given the price. The 5D2 had the same MP as the 1Ds III but wasn't crippled to perform slightly worse etc.

Not really.

5D2 had better SNR than the 1Ds3 (although slightly worse DR at the lowest couple ISOs and a lot more banding in shadows down there)

 

[neuroanatomist]

10% bigger pixels = extra stop of performance.

'Stops' are a log scale. One stop is a doubling or halving, i.e. a 100% or a 50% change, depending on direction. 10% << 50% or 100%.

But I like the 'new math' approach...

 

[briansquibb]

No. The 5D series is typically one stop under the performance of the 1D series. This is acceptable, especially given the price. The 5D2 had the same MP as the 1Ds III but wasn't crippled to perform slightly worse etc.

Not really.

5D2 had better SNR than the 1Ds3 (although slightly worse DR at the lowest couple ISOs and a lot more banding in shadows down there)

Yep the 1DS3 is superb in the iso 50-400 range, pretty much noise free and devoid of banding. The 1DS3 also has decent AF, making it a very decent alternative to the 5D3 if you dont want to shoot at iso 1600 or higher

 

[jaduffy007]

Judging by the 1D X samples, the high ISO performance is outstanding and a true 1 stop improvement over the 5D Mark III. Something that strikes me though is 4 Megapixels can't make that much difference.

Canon opted to put a nice new focus system in the 5D Mark III, but it seems to avoid people not buying the 1D X and pro's opting for a Mark III instead, they limited the ISO performance boost. With a boost there wouldn't be much between them but now the 1D X is more appealing with its ISO quality jump!!

Sorry but in regards to the 5D3 performance / price, ratio, it's called milking your customers. That said, there are other major differences in the cams. AF tracking, fps obviously is huge, etc. I have a feeling the 1D X is gonna kick some serious butt for action, low light shooters.

 

[preppyak]

Yes, you need to ask why. What is it that makes one camera perform vastly better than another?

Well, the price for one thing. If the 1DX didn't do anything better than the 5dIII, nobody would buy it for twice the price. So its understandable that Canon would use certain items that are better in the 1DX to differentiate it.

Why are Sony sensors better while the cameras using them are cheaper?

Because a sensor is one item in a camera, and the best sensor doesnt change a blurry or missed shot. Also, if we're talking video, I'm not sure I'd agree the Sony sensor is better (reviews pretty universally favor the Canon footage). But for photos, I'd agree that the Sony sensor has some potential advantages

Also, its coming from different companies...so maybe Nikon is taking a loss on the D800 to get people to convert to their system (Sony sold the PS3 at a lost for a few years, for example). Maybe the D800 has really a really cheap build. Maybe Canon is fleecing their customers. I'm not sure you can answer that question in any reasonable way since it involves multiple companies, multiple strategies, and thousands of different parts.

Why does the 7D use the same sensor as the 550D? Because the price is not in the sensor and the 7D doesn't make better pictures once the light gets through to the sensor than the 550D.

And the answer there is easy, as just reading the spec sheet would show you all the differences in the camera. Again, the best sensor in the world doesn't matter if the object is out of focus, or if the camera can't process the shots fast enough to even take the shot.

 

[Kernuak]

Probably one reason Nikon has been able to keep the price down on the D800 (not the only one by any means) is that they haven't really changed the AF compared to the D700, in fact other than the sensor, there are probably minimal changes in the scheme of things. To a lesser degree, it may also be true of the D4. Of course, when you look at the AF, build quality and one or two other things that were "fixed" in the MkIII, the 5D MkII had a lot of catching up to do in those areas compared with the D700. That must have taken significant R&D costs to do and was an area that Nikon was almost certainly able to save money on, as they would have felt it was sufficient for the user base. Even the 1D X has improved AF according to both Canon and the few reviews over the 1D MkIV. It's certainly a completely new system, regardless of how much better it may or may not be.

 

[RLPhoto]

Yes, So they can make more money. 8)

 

[wickidwombat]

the 1D series are the flagships of the canon line they are the pinaccle of canon cameras at the time
while they miss out on many features lower models have they are the ultimate as far as image quality goes
expecting a lower model camera that costs half as much to be the same as this is somewhat naive

 

[Policar]

It's not about the lack of pixels but the space it affords them.

The pixels are what? 10% bigger than the 5D3's?

ISO stops = 100,200,400,800,1600,3200,6400,12800,25600,51200 - 10 stops.

10% bigger pixels = extra stop of performance.

This is irresponsibly terrible logic. You found the number 10 twice (incorrectly, and in arbitrary places, I might add) and concluded that there's a relationship. I'm not sure if you're parodying terrible logic or are serious; my apologies if your post was a joke (subtle humor is lost on me over the internet!).

Assuming gapless microlenses, the 1DX has pixels that are 23% larger by surface area (5760/5184)^2, not 10%. And it doesn't matter because the surface area that collects light is roughly the same on both cameras. If the 5D has smaller pixels, it has more of them, and this cancels out almost exactly except maybe in terms of collecting light at very oblique angles. The D800 is maybe 3-4% less light efficient than the D4 despite having more than twice as many pixels--that's the tiny extent to which megapixel count effects light sensitivity (assuming you have a very clean ADC, as those cameras do).

The issue is likely with read noise. The 5D III (if I remember correctly) is >50% efficient, an improvement over the mark II that is achieved with gapless microlenses, but it has very high read noise. There 1DX likely has very similar quantum efficiency but a better ADC. It obviously has a better sensor but not because there are fewer pixels.