Canon EOS 5Ds and 5DsR - 50.3MP with and without AA filters

[shadowblade]

http://photorumors.com/2015/01/30/this-is-the-new-50mp-canon-eos-5ds-eos-5ds-r-full-frame-dslr-camera/#ixzz3QHcLZecl

Looks good so far.

But the key question is whether Canon have made up any ground in terms of DR.

Also, I wish they'd given it a base ISO of 64, or even 50, like the D810 - no-one buys a body like this to shoot low-light action, and a lower base ISO would help with both noise and DR.

[torger]

Also, I wish they'd given it a base ISO of 64, or even 50, like the D810 - no-one buys a body like this to shoot low-light action, and a lower base ISO would help with both noise and DR.

It's not that simple with digital. The original reason why digital cameras had low ISO is that most of the light is not registered at all, so you need to expose it for a longer time -> lower ISO. A low ISO cannot be directly translated to improved quality as you could in the film days (where all light is registered).

Then sensors became better at registering more light, by the use of microlenses and thus base ISO go up. To make ISO down in "a good way" you need higher full-well capacity so each pixel can gather more light before it clips. Some has happened in that area too, but most gains in DR come from lower noise levels in the electronics.

The new Canon looks exciting in any case, can't wait for the official announcement . They could be using a Sony sensor or traded some Sony patents so they may have the DR, if they don't have people will get very disappointed.

[shadowblade]

It's not that simple with digital. The original reason why digital cameras had low ISO is that most of the light is not registered at all, so you need to expose it for a longer time -> lower ISO. A low ISO cannot be directly translated to improved quality as you could in the film days (where all light is registered).

Then sensors became better at registering more light, by the use of microlenses and thus base ISO go up. To make ISO down in "a good way" you need higher full-well capacity so each pixel can gather more light before it clips. Some has happened in that area too, but most gains in DR come from lower noise levels in the electronics.

The new Canon looks exciting in any case, can't wait for the official announcement . They could be using a Sony sensor or traded some Sony patents so they may have the DR, if they don't have people will get very disappointed.

That's why I said the base ISO - the ISO where a full well equals maximum exposure. The lower the base ISO, the deeper the well, assuming the same percentage of photons are registered.

[Bart_van_der_Wolf]

That's why I said the base ISO - the ISO where a full well equals maximum exposure. The lower the base ISO, the deeper the well, assuming the same percentage of photons are registered.

Hi,

Not necessarily the case. ISO rating (REI or SOS) is a calculation which combines the effective sensitivity of silicon based photo-voltaic processes and the optical density of the Bayer CFA filters (and the filter stack and sensor cover-glass). Well depth or storage capacity is more correlated with surface area than with physical depth.

The silicon based sensitivity component is more or less a given. It's just that larger sensel areas allow to store more electrons before saturation occurs. Therefore, smaller sensel areas will increase the base ISO value if everything else is kept the same. The only natural way to reduce that is by using a denser CFA/filter stack.

Other methods will not affect the Signal to Noise ratio, which is the quality improvement one usually hopes to get from using lower ISO settings.

Cheers,
Bart

[torger]

To summarize, lowered ISO through higher full well capacity = good as it means more registered photons; lowered ISO through lower quantum efficiency (QE) = bad as it just means more time is required to gather the same amount of photons.

I guess there's still a possibility to increase full well capacity by having the photo diodes cover more of the sensor surface, but technology seems not to be there yet at this sensor size, as you need to fit the wiring and stuff on the same side.

I don't think we're going to see any revolutionary performance from Canon here. The most likely is unfortunately that pixel design and DR will be the same as other recent Canon sensors, which will be disappointing in this type of camera of course. If the DR is the same as recent Sony sensor I think it will be due to some sort of collaboration patent swap etc with Sony so Canon has been able to implement some Exmor properties in their sensor, and in the least likely case it actually is a Sony sensor. We'll see, I'm very very curious.

[Paul2660]

Actually, if the latest rumors are true, the sensor will be a Sony (not a big surprise) and thus I think you expect pretty good DR.

Sony, Canon and Nikon all should soon have offerings with this new 53MP sensor, Nikon more than likely the last.  It seems Sony will announce the high MP follow on to the A7r with this sensor about the same time the Canon is announced. 

Canon's version of the new sensor has their proprietary AF design built in, and only Canon will have that version of the chip.

Seems that next Friday is the big day, as the non-nondisclosure agreements  are over.   I am sure Lula will be bringing us all the new stuff soon.

Paul

[torger]

While at discussing rumors I've found it interesting that the max ISO is at low 6400, and the rumor has it that the CFA filters have been optimized for color accuracy rather than high ISO performance, something medium format cameras is said to have had all the time. If that's true it would be just awesome.

[davidgp]

Actually, if the latest rumors are true, the sensor will be a Sony (not a big surprise) and thus I think you expect pretty good DR.

For me if that rumor becomes true, even as a landscape Canon shooter it will be a welcome one, will means that Sony is starting to have a monopoly in digital sensor production... I always like the situation where you have two or more competitors, keeping the price check and the innovation going on...

[davidgp]

http://photorumors.com/2015/01/30/this-is-the-new-50mp-canon-eos-5ds-eos-5ds-r-full-frame-dslr-camera/#ixzz3QHcLZecl

If it is what finally it is announce next Friday (I don't doubt it), about f%&/%$& time that Canon puts an intervalometer and a bulb timer in its cameras... (this is something that sadly Canon is not alone... )

[Paul2660]

If it is what finally it is announce next Friday (I don't doubt it), about f%&/%$& time that Canon puts an intervalometer and a bulb timer in its cameras... (this is something that sadly Canon is not alone... )

What has amazed me is that none of the major players have figured this out.  With Magic Lantern installed on 6D, you can run a true time lapse series, longer than 30 seconds.  It's the only built in process I have seen. 

Sony, has nothing, besides their toy app, called time lapse which again limits you to 30 seconds.  Nikon has a "built in" intervalometer, but no bulb timer, so again you are limited to 30 seconds, Fuji et all, Canon et all, 

As it was easy enough to implement with Magic Lantern, I am still amazed that no one else has tried to do this.  It can't be that hard to do. 

So for now, with Canon, Fuji and Nikon, you still need the external cable mounted intervalometer, Wtih Sony, there is not even that.  What a waste.

Paul

[LKaven]

The rumored specs include "regular sensitivity: ISO 100-6400".

Hmm.  If this is an Exmor, I'm surprised to see it top out at ISO 6400. 

[NancyP]

My 50 buck wired intervalometer/ ordinary shutter release does fine.
I will be very interested to see if this is going to be an ideal landscape camera.

[dwswager]

http://photorumors.com/2015/01/30/this-is-the-new-50mp-canon-eos-5ds-eos-5ds-r-full-frame-dslr-camera/#ixzz3QHcLZecl

Looks good so far.

But the key question is whether Canon have made up any ground in terms of DR.

Also, I wish they'd given it a base ISO of 64, or even 50, like the D810 - no-one buys a body like this to shoot low-light action, and a lower base ISO would help with both noise and DR.

Assuming that the rumors of them using a Sony senor are true, this has the makings for a great landscape camera.  My point is that I think by 48-50MP you has overshot the general purpose market and are into a specialty area.  Its going to take a tremendous amount of processing and bandwidth to push 50MPs around inside the camera and out.  I think that is where the D800/D800e surprised people, even Nikon.  I think it was supposed to be a specialty camera, but the market said otherwise and that is why we saw a speedier and more functional update rather quickly to capitalize on that.

As a D810 owner, I will disagree on the whole shoot low light action comment.  I personally think that at the moment, the D810 and the D750 (if your willing to sacrifice 12MP) are the best GENERAL PURPOSE DSLRs on the market.  I know the D810 has great low light capability both from good high ISO performance and the ability to pull detail from the shadows will little noise.  It shoots 5fps standard, 6fps in 1.2X (24MP) and 1.5X (15.6MP) crop mode and will shoot 7fps in DX with the battery grip.  While most of my Sports are outside daytime, I just shot an indoor Futsal soccer tournament with the D810 with no problems at all.  I've only had it since late November, 2014, but the more I use it the more I know this is the camera I had been waiting years to get my hands on.

[BernardLanguillier]

I am a bit surprised by these Exmor rumors in the new Canon. My sources, that seemed very credible, were saying that there was no way Canon was going to use a Sony sensor.

We'll know in a week it seems.

Cheers,
Bernard

[AlterEgo]

I am a bit surprised by these Exmor rumors in the new Canon. My sources, that seemed very credible, were saying that there was no way Canon was going to use a Sony sensor.

even if Canon is not using Sony sensor the mere fact of Canon FF dSLR with purely Canon 50mp sensor will provide more push to make Sony 50mp sensor a reality rather sooner than later... so win-win in any case

[AlterEgo]

The rumored specs include "regular sensitivity: ISO 100-6400".

Hmm.  If this is an Exmor, I'm surprised to see it top out at ISO 6400. 

that's a camera's maker call where to top regular ISO on camera - not sensor maker ...

[AlterEgo]

something medium format cameras is said to have had all the time.

or did they say that just because of CCD sensor tech (readout noise w/o cooling of CCD sensor in a consumer level device) ?

[dwswager]

I am a bit surprised by these Exmor rumors in the new Canon. My sources, that seemed very credible, were saying that there was no way Canon was going to use a Sony sensor.

We'll know in a week it seems.

I have zero sources, but having waited this long and using an out of house sensor at this point would seem to indicate capitulation.  Three years ago is looks like a smart management decision.  Now it shouts total management failure AFTER you squandered a 17% market share lead over your competitor! Which is why I just don't believe they will go out of house.

I'm actually rooting for the in-house guys!  But here is the big question:  How is it received if it in house and does not perform in the same class as the Sony sensored cameras?
 

[fdisilvestro]

But here is the big question:  How is it received if it in house and does not perform in the same class as the Sony sensored cameras?
 

There will be endless threads in on-line forums bashing Canon, mostly by non-owners (and without proper knowledge of digital signal processing). On the other hand there will be great photographers producing outstanding images using those cameras.

[John Koerner]

Is the video 4K?