Canon announce 5Ds and 5Ds-r

[ricktas]

http://gizmodo.com/canon-5ds-and-5ds...ame-1683275915

50MP sensors!!!!!

[MissionMan]

Good stuff. It's always nice to see someone lifting the game a bit. I'm sure the Canon shooters will be happy (along with the computer hardware manufacturers due to the extra processing power required to edit a 50MP RAW file with reasonable performance).

[Arg]

All I can say is, AWESOME!

Company should change its name to Bazooka

[Morgo]

Cool, though not a 5d3 replacement it might fill a need for some.

[dolina]

The 51MP 5DS ($3,699) & 5DS R ($3,899) are high end studio and landscape cameras. Think medium format applications where large or finely detailed prints are required.


These cameras do not replace the 5D Mark III per Canon USA video.





These cameras are positioned against the high megapixel full frame bodies like the 36MP Nikon D810 ($2,997) & 36MP Sony A7R ($2,098).


Canon USA mentioned medium format uses for the 5DS & 5DS R so here are the entry level medium format bodies by year.


2010 $4,500 40MP Pentax 645D
2014 $8,497 51.4MP Pentax 645Z


Now, why the drastic price difference? This all goes down to sensor size as shown below.


The Sony, Nikon and Canon uses a full frame sensor with a dimension of 36x24mm while the Pentax and other higher end medium format cameras tend to use a sensor with a dimension of 44x33mm or larger.

Image deleted. Please do not attach images that are copyright to others.


6X6 medium format film is included for comparison.

[phild]

Noise performance and dynamic range is reported to be on par with the 5DIII.

My rough calculations puts the pixel size the same as the new 7DII sensor so noise and dynamic range should be at least as good as that sensor.

The reviews I've read show that the 7DII is as good as the 5DIII so it will be interesting to see how the new beast performs.

[MissionMan]

You guys can "ooh" and "ahhh" all you want, but at this point, until Canon launches something to rival the Barbie Cam, it's still playing catchup.

[Tannin]

Noise performance and dynamic range is reported to be on par with the 5DIII.

My understanding is that the per pixel noise performance is about the same as the 5D II. (Source: DPR.) Note, that's per pixel noise, so per image noise is of course significantly lower.

I've been saying for years now that Canon has become so obsessed with video that they have forgotten about photographers. These two, together with the 7D II, suggest that they have been listening. Good work Canon!

Will I get one? Possible, but probably I'll look at a 5D IV. Tempting just the same.

[MissionMan]

My understanding is that the per pixel noise performance is about the same as the 5D II. (Source: DPR.) Note, that's per pixel noise, so per image noise is of course significantly lower.

I've been saying for years now that Canon has become so obsessed with video that they have forgotten about photographers. These two, together with the 7D II, suggest that they have been listening. Good work Canon!

Will I get one? Possible, but probably I'll look at a 5D IV. Tempting just the same.

I've noticed Canon and Nikon switch the lead every couple of years. Prior to the D3, Canon spent a while ahead of the game and Nikon was playing catch up.

[bitsnpieces]

My understanding is that the per pixel noise performance is about the same as the 5D II. (Source: DPR.) Note, that's per pixel noise, so per image noise is of course significantly lower.

I've been saying for years now that Canon has become so obsessed with video that they have forgotten about photographers. These two, together with the 7D II, suggest that they have been listening. Good work Canon!

Will I get one? Possible, but probably I'll look at a 5D IV. Tempting just the same.

This is something I wonder about also - the 5Ds and 5Ds R are only full-frame cameras, they're not necessarily medium format or anything.

Will 50mp really be that good in low light?

Yes, there's more megapixels to pack down and reduce noise, so potentially yes, but I still wonder, to what extent?

For example, Sony's a7s is only 12mp (if I remember correctly) and full frame, so the pixel sizes will be much larger than what the 5Ds is using. The a7s has amazing low light performance because it's only 12mp so the receivers (or whatever it's called) are larger, receiving more light, more accuracy, etc.
Packing 50mp worth into a regular full-frame size, these receivers are going to be tiny.

So noise reduction is really going to come down to Canon's engine and the packing down from 50mp - so I'm just wondering, is the low light performance on this beast really going to be that amazing?

As phild was saying - reviews say that the 7Dii is on par with the 5Diii.
The 5Diii is 22mp (I think it was) on a full-frame sensor
The 7Dii is 20mp on an APC sensor.
So in terms of size relativity, it's fairly equal (not really, but for arguments sake, more mp with suitable larger sensor, less mp with a smaller sensor, all fits well).
The 7Dii is running their new dual processor, so again, by receiver size, very similar, it's coming down to the processor.

The 5Ds is running the same sensor as the 7Dii right? Can the processor really be powerful enough to compensate for the extra ~30mp increase / decrease in receiver size?

[end rant]

[TassieSnapper]

Can someone explain the difference between the R and S models?

[ricktas]

Can someone explain the difference between the R and S models?

The 5Ds R cancels out the low pass filter (ie, there is not one). Like the Nikon D800 vs D800e. Result can be slightly sharper images, and on the downside an increase in moire (theoretically). However at the pixel densities we are now seeing, moire is virtually eliminated anyway. http://www.techradar.com/au/news/pho...asing--1160397

[Tannin]

Will 50mp really be that good in low light?

Yes, there's more megapixels to pack down and reduce noise, so potentially yes, but I still wonder, to what extent?

For example, Sony's a7s is only 12mp (if I remember correctly) and full frame, so the pixel sizes will be much larger than what the 5Ds is using. The a7s has amazing low light performance because it's only 12mp so the receivers (or whatever it's called) are larger, receiving more light, more accuracy, etc.
Packing 50mp worth into a regular full-frame size, these receivers are going to be tiny.

So noise reduction is really going to come down to Canon's engine and the packing down from 50mp - so I'm just wondering, is the low light performance on this beast really going to be that amazing?

As phild was saying - reviews say that the 7Dii is on par with the 5Diii.

For low noise, yes, size matters. But it's not the size of the pixels which matter, it is the size of the sensor. For any given sensor size and output resolution, the noise difference between a 12MP camera and a 50MP one is quite small (assuming, of course, that we are comparing implementations of roughly equal quality and technology generation). It doesn't matter whether you up-res the 12MP unit to 50MP or down-res the 50MP one to 12MP, noise performance isn't going to be massively different.

People misunderstand this all the time (I'm not saying you do, just that lots of 'togs who ought to know better do). The brainlessly compare 100% crops and say "look how much better the noise is!" Well duh, of course it is, 'coz they are comparing apples with elephant droppings.

The claim (which I don't make myself, only repeat on trust) is that we are looking at noise performance per pixel equal to the 5D II. Read that carefully: 5D II not III, and per pixel, not per image. If true, this is an astonishing achievement, as (given the huge difference in pixel count) that translates to amazingly good overall noise performance when measured properly (i.e., per image).

With all that said, Canon nevertheless state that the 5D III and 1DX are still their best low-noise performers, which is what you'd expect given that they have much lower resolution. (Noise is mostly about sensor size, but pixel size still helps a bit, so that makes sense. )

Me, I'm still in awe of the noise performance of my ancient 5D II, so any of the latest full frame units would knock my socks off.

Me, I think I'd like a 50MP 5D. I don't reckon my technique is anywhere near good enough to get the best out of one, so I'd have to put a lot of work into climbing that learning curve .... which could only be a good thing!

(Damn it, why haven't I got any grandmothers left? I might have to sell something else instead, and I already sold my left one and perfectly good kidney.)

[Steve Axford]

Sounds like a great camera - if you can use that many pixels, and most of us can't. I can see why Canon have positioned it for a specialist market. It seems to be aimed at regaining the high ground from Nikon, rather than selling a lot of units. I doubt that I'll buy one, though I'd happily accept one that was given to me and I'm sure I'd find a use for it. That new lens (11-24mm) sounds great too, and is even more attractive than the camera, but sooo expensive. Oh well. There's still lots to do with the "cheap" cameras and lenses so I think I'll stay with them for a while.

[MissionMan]

For low noise, yes, size matters. But it's not the size of the pixels which matter, it is the size of the sensor. For any given sensor size and output resolution, the noise difference between a 12MP camera and a 50MP one is quite small (assuming, of course, that we are comparing implementations of roughly equal quality and technology generation). It doesn't matter whether you up-res the 12MP unit to 50MP or down-res the 50MP one to 12MP, noise performance isn't going to be massively different.

People misunderstand this all the time (I'm not saying you do, just that lots of 'togs who ought to know better do). The brainlessly compare 100% crops and say "look how much better the noise is!" Well duh, of course it is, 'coz they are comparing apples with elephant droppings.

The claim (which I don't make myself, only repeat on trust) is that we are looking at noise performance per pixel equal to the 5D II. Read that carefully: 5D II not III, and per pixel, not per image. If true, this is an astonishing achievement, as (given the huge difference in pixel count) that translates to amazingly good overall noise performance when measured properly (i.e., per image).

With all that said, Canon nevertheless state that the 5D III and 1DX are still their best low-noise performers, which is what you'd expect given that they have much lower resolution. (Noise is mostly about sensor size, but pixel size still helps a bit, so that makes sense. )

Me, I'm still in awe of the noise performance of my ancient 5D II, so any of the latest full frame units would knock my socks off.

Me, I think I'd like a 50MP 5D. I don't reckon my technique is anywhere near good enough to get the best out of one, so I'd have to put a lot of work into climbing that learning curve .... which could only be a good thing!

(Damn it, why haven't I got any grandmothers left? I might have to sell something else instead, and I already sold my left one and perfectly good kidney.)

I'm confused. Isn't the exact reason that the size of the sensor impacts the levels of noise because the size of each pixel is larger and receives more light per pixel. I.e. a 24MP sensor on a full frame would have larger pixels than a 24MP sensor on DX and therefore better low light performance?

I also not sure I would agree that the performance would be the same. From what I have heard, the 36MP sensor on the D810 is more prone to camera shake due to the high number of pixels at low shutter speeds where a lower resolution sensor would be less problematic. This may not seem like an issue, but if you're shooting in low light, chances are you may also be shooting at low shutterspeeds which would make the issue more visible in the higher MP cameras like the D810 and the 50MP 5D, than a Sony 12MP camera which also has a full frame sensor.

It's not that I am against increasing sensor sizing, I think it's inevitable and it's great and noise will improve as time goes on so I think this argument is largely mute because our standard of ISO performance increases with each generation. As you mentioned, most manufacturers offer lower MP sensors with better low light noise and I don't think it's possible to have a single camera system that can do everything although this is also largely our perception of what everything is and 10 years ago, even the base level current DSLR's would be better than most of the pro systems so it's the standards of everything that are increasing, rather than what everything is. The only down side really is the potential issue of camera shake along with the increases in processing power and storage to cater for the larger file sizes, but realistically most people should be upgrading their machines on a regular basis so processing and storage issues tend to reduce over time.

[bitsnpieces]

For low noise, yes, size matters. But it's not the size of the pixels which matter, it is the size of the sensor. For any given sensor size and output resolution, the noise difference between a 12MP camera and a 50MP one is quite small (assuming, of course, that we are comparing implementations of roughly equal quality and technology generation). It doesn't matter whether you up-res the 12MP unit to 50MP or down-res the 50MP one to 12MP, noise performance isn't going to be massively different.

Ah, that's what I was kind of trying to get at, sorry, I will never be good at terminology.

Anyways, along with what MissionMan said, technically speaking, if using the same size sensor, so in this case, full-frame, the lower the megapixels, the larger the light receivers, as there's more space for them. This in turn provides better image capture, less noise.

I think the difference in this situation now is the technology and processing behind it. If the lower MP sensor had bad algorithms and didn't compress well, compared to the higher MP sensor, which although hardware-wise doesn't capture as clean an image, but processes so well that it makes it so clean, then the quality can be equal or better.
That's what I'm seeing with Canon's cameras, which is why I was compared the 5Diii and the 7Dii. 5Diii has a larger sensor and has more pixels. 7Dii has smaller sensor less pixels. So hardware-wise, for arguments sake, cancel out to be even. Software-wise, the technology in both is the same, and that's why both have similar ratings for low-light performance (according to one review I briefly read over). They both just equate.

Sony's a7s is a full-frame but only 12mp, so lots of light performance. However, their software-end in processing images is horrible. This is very evident in their RAW files for their cameras. So although it has the hardware advantage, unfortunately, the software doesn't make use of it. So a well programmed camera with full-frame sensor and more pixels could indeed come close in low light performance - I believe only close because image processing can only go so far.

So it's a given, quickly looking at a few reviews, though both 5Diii and 1Dx are their best low light performance cameras, the 1Dx does have a slight edge over the 5Diii and slightly better low light performance, and that's because the sensor size is a little smaller, providing better light capture.

Which is the only reason why I question Canon's 50mp - I believe it'll still be great, and I personally believe it's comparable to the Sony RX10 (I'm not saying the RX10 is better, but I mean the outcome), in the idea of fitting lots of pixels into a sensor, but with decent enough processing, it'll turn out just brilliant.
The RX10 is a 1" sensor, so it's smaller than the usual APC, but packs the mp of an APC sensor camera. So the idea of packing lots of pixels into a sensor, that I guess you could say, doesn't quite fit it, with good processing, can have a great outcome, which the RX10 does (very comparable to APC sensor quality).

I think this is Canon's method to try and bring back momentum into their product line. However, just like Fujifilm (I think it was) that developed their new sensors to have better light capturing for better low light performance (within their sensor sizes), Sony is also developing a new sensor for much better light capturing once again.
I don't know if Canon has developed any new sensors yet or still using the standard type, but just with better processing.
Because if this is the case, only time will time when the quality of the 50mp just can't keep up. Again, I believe the software processing can only go so far.

[Arg]

Is anyone here buying one?

From what I have heard, the 36MP sensor on the D810 is more prone to camera shake due to the high number of pixels at low shutter speeds where a lower resolution sensor would be less problematic. This may not seem like an issue, but if you're shooting in low light, chances are you may also be shooting at low shutterspeeds which would make the issue more visible in the higher MP cameras like the D810 and the 50MP 5D, than a Sony 12MP camera which also has a full frame sensor.

Camera shake is independent of the number of pixels. Photos printed from a 50MP 5D and a 12 MP Sony will have exactly the same amount of camera shake.

Pixel peeping at 100% might look like the high res shot has more blur, but only because the 100% view represents twice as much magnification as it would for the 12 MP. Stop pixel peeping and the shake looks the same on final images.

The reason I, and I suppose others, have posted in the past that you need a tripod with a D800 is because a high resolution image is not high resolution if the pixels are not all different, so you have to hold the camera more stably if you really want it to resolve more than a lower-resolution camera.

[dolina]

For those who reserved what will you be using the 5DS or 5DS R for? Wildlife? Birds? Landscape? Studio?

[ricktas]

For those who reserved what will you be using the 5DS or 5DS R for? Wildlife? Birds? Landscape? Studio?

Funnily they can take all of these.

[dolina]

Funnily they can take all of these.

I'm sure it can take a great passport photo as well but it's serious amount of money and I'd like to know what people who reserved this camera will be using it for.