5D Mark III [CR1]

[gmrza]

I read an interesting article on medium format sensors the other day. In it was a discussion on how many mp would be needed before the sensor outperformed the lens. Apparently Canon are said to be working on lenses for 36x24 "full frame" that will work happily up to 60mp, while on the Pentax 645 sensors would need to exceed 110mp before it becomes an issue. Basically, I believe that manufacturers will simply offer better lenses as the MP count creeps ever higher.

That same article was saying that in prints of A3 or bigger, the large MP sensors on good quality medium format systems produced a very obvious quality improvement over the 36 x 24mm "full frame" size sensors. Basically, bigger is still always better.

An interesting question about that will be what lenses like that will cost.  Clearly over time, they will become cheaper, but I expect that we will initially see such lenses being quite expensive (relative to other 35mm system lenses).
On the other hand, that still gives Canon the ability to go hunting in traditional medium format territory, as such lenses might still be cheaper than medium format lenses.  I suspect that Canon (and aso Nikon) will want to hunt as many photographers from the medium format camp as they can. - Ultimately, if your clients do not need resolution of more than 60MP, why buy a 110MP medium format system is a 60MP DSLR will do?  Many photographers who previously shot medium format film have already taken that path and migrated to DSLRs.

I am not saying that there is no need for a 110MP MF camera.  I am just postulating that Canon (and Nikon and Sony) would want to encroach as much as possible into the traditional territory of MF systems, relegating MF to being a more and more specialised niche.  (That is a pity in a way, as that stands to risk MF cameras becoming even more expensive, due to smaller market scale and less competition.)

As for 32MP on a 5DmkIII - anyone who currently makes 20"x30" prints off a 5DmkII would welcome that!

[YoukY63]

I just don't see the point of 32Mp. Why spoiling the burst rate because of this? Such an MP count would just momentarily fill up memory cards and your PC. And as someone mentioned before, shooting with sRAW does not work with some software. With such specs it actually seems dissapointing, because of only increased MP, ISO and AF. One of the previous topics clearly showed that between ISO, FPS, autofocus points and raw video, increased MP was the last thing people wanted. 24MP would still be ok and would give a much better FPS performance.

And if these were the only improvements Canon makes to the camera, which I doubt it will be, then I thing most of us will be very dissapointed.

Then you are not a landscape photographer. If you want burst speed, don't buy a 5D serie Camera, there are not designed for that. What you want is a 1D serie camera: lower MP and higher burst rate, + better AF.
For me, 32MP sounds very logical. I was expecting 30 to 40MP. With an increase of 2 stops of sensitivity, that camera (if informations are true) will make me happy until 5DmIV! 
I shoot mostly landscape and portraits, daytime and also complete dark time. As a landscaper, I never have enough pixels to record all the smaller details from these wonderful Japanese scenery (which explain the success of medium format camera with 80+ MP in landscape community).
The increase of sensitivity will also be very welcome since I mostly use fixed focal lenses for my night pictures, but as they lack stabilisation I sometime fall short with the actual 5DmII sensitivity (even with f/1.4 lenses).

PS: also do not forgot that Sony will be presenting a 24MP APS-C sensor in a few weeks, and is expected to release a 40+ MP FF sensor around fall. Canon made a big part of its success during the past few years on high resolution DSLR, they cannot let Sony's sensors (Sony but also Nikon and Pentax cameras) win this war.

[te4o]

Let us all be just optimistic (it improves the mood) and say, the 5D3 will have at least 32 MP, will be very adaptable and let anyone choose how many of these to use for stills and video, at a perfect ISO & DR, will give you 10.6 FPS at lower MP counts, will serve us all for another 4 years etc etc , and will BE ANNOUNCED in JUNE 201ouch!

[abarraga]

Sort of related to this posting:

I went to Fry's Electronics (to shop around) and there was a fully functioning demo version of the 5DMKII with the 24-105mm f/4L IS.  It was no surprise they did not have the body in stock, but I was offered 5% off the demo kit.

Knowing that the MK III will inevitably come (and my desire to upgrade to a full frame), should I have taken this offer?

[Gcon]

This is a very believable spec sheet.

32mp. Given that 600D, 60D and 7D are 18mp, the same pixel density on full frame is approx 46mp. Thus a move to 32mp is reasonable.

ISO to 25600. I don't expect it to go that high. I have no need for it in any case.

Digic V - to be expected. Hopefully it comes with an overhauled menu system and can do advanced things like electronic leveling.

4.2 shots/sec. A small increase from 3.9 of the 5DII. No threat to the 1DIV.

19 point AF. Taken straight from 7D.

CF card - to keep people with large investments in this card happy. Also it is appropriate for large RAW file transfer speeds.

All the video stuff I don't care for, but would imagine this is a good platform for video, as it's shot horizontal so don't need the vertical grip of the 1D series.

==========
Rant - I was hoping that Canon would bring the 5D line closer to the 1D line with a more rugged body and weather sealing, and using buttons instead of mode dial, and having an integrated viewfinder cover.

It doesn't look they will go that bold, but will keep the status quo and just bump up the features a bit to have a "full frame 7D" but without the same level of weather sealing.

This is a shame really, as Canon users need something akin to the Nikon D700, and it's imminent successor.

[simonfilm]

But 32mp add more problem for video processing in order to avoid line skipping Moire.

I would like a 5D3 with no heat problem, that you can have it recording video hours without heating.

[danski0224]

Sort of related to this posting:

I went to Fry's Electronics (to shop around) and there was a fully functioning demo version of the 5DMKII with the 24-105mm f/4L IS.  It was no surprise they did not have the body in stock, but I was offered 5% off the demo kit.

Knowing that the MK III will inevitably come (and my desire to upgrade to a full frame), should I have taken this offer?

5% off isn't really much of a deal.

I would take it from a new in box item... if I was in the market. I don't think it is enough of a discount on an open item.

[epsiloneri]

32MP? Where am I gonna put all these images into? I just went to Rome for 6 days and I've taken over 64 GB of images! I guess I will need two new 3TB hard disks along with this 5DIII (one for original and one for backup).

32MP for a 5D3 is not that much more than the 21MP of the current 5D2, given that 3 years has passed since the latter's introduction. Hard drives have increased in size much more than those 50%, and storage price has reached 5 cents per GB. You can store all your Roman holiday pictures for $6, including backup. With 32MP it would have been $9. If the pixel count followed Morse's law (like storage does), the 5D3 would have been 80MP. Storing the pictures is becoming easier and cheaper with time despite the MP increases.

[epsiloneri]

When you start having more than one value in the raw file per electron in the sensel, it does not make much sense to increase the amplification further, you can just scale in software. Unity gain (1 raw value = 1 electron) happens quite early, say around ISO 1600 or so. Why some cameras still do "native" amplification past unity gain I don't really know, perhaps there are some read noise advantages or something.

Unity gain depends both on the QE and the pixel pitch. I doubt a new 5D3 sensors would improve on the QE much, but with smaller pixels the unity gain would actually decrease, so you're right in that it wouldn't be much technical point in extending the ISO range even further (there can still be selling points in doing so however). Over-resolving the electron count can give a slight advantage for quantization reasons, since the A/D converters are basically analogue devices that do not register electrons 100% efficiently.

Read more about this at Clarkvision.

[motorhead]

I for one have never seen burst rates as anything important, to me it's another toy to play with when I'm bored. So I much prefer a bigger MP and would happily live with the consequences.

I speak as a died in the wool motorsports photographer as well as a landscapist. I much prefer to shoot in one-shot mode and use my understanding of the particular sport to guide me in my taking a shot. I'm firmly of the opinion that no matter how fast the bust-rate is, its always going to miss that "peak action" moment. I've read  books and articles written by well respected motorsport 'togs and in ever case they have taken a similar view. 

[DavidD]

All these specs sound reasonable.

At 32mp I'd easily upgrade from the 5d2.
At only 28mp - probably not.

I'm hoping Canon will get rid of the anti-aliasing filter
so we get sharper images. (Software can now handle
moire much better than that silly filter.)

And if the rumored "revolutionary" improvements in
noise reduction and back lit sensors arrive in the 5d3,
we'll all be celebrating.

[nex-s]

Then you are not a landscape photographer. If you want burst speed, don't buy a 5D serie Camera, there are not designed for that. What you want is a 1D serie camera: lower MP and higher burst rate, + better AF.
For me, 32MP sounds very logical. I was expecting 30 to 40MP. With an increase of 2 stops of sensitivity, that camera (if informations are true) will make me happy until 5DmIV! 
I shoot mostly landscape and portraits, daytime and also complete dark time. As a landscaper, I never have enough pixels to record all the smaller details from these wonderful Japanese scenery (which explain the success of medium format camera with 80+ MP in landscape community).
The increase of sensitivity will also be very welcome since I mostly use fixed focal lenses for my night pictures, but as they lack stabilisation I sometime fall short with the actual 5DmII sensitivity (even with f/1.4 lenses).

PS: also do not forgot that Sony will be presenting a 24MP APS-C sensor in a few weeks, and is expected to release a 40+ MP FF sensor around fall. Canon made a big part of its success during the past few years on high resolution DSLR, they cannot let Sony's sensors (Sony but also Nikon and Pentax cameras) win this war.

See again, I could say that what you want is a MF camera. Many would love a fast FF instead of an APS-H. And about the cost of storage. Well, HDD prices are falling, but flash memories still cost a lot of money, so... If they allow higher burst rates at a lower resolution, then I`m not that bothered.

What I am hoping is that Canon increases the weather sealing on the 5D mkIII to the 7D's level. If they don't do that, I would be rather dissapointed, cause that is what I'm looking forward to.

[NotABunny]

Some software packages, like DxO, can't process sRAW & mRAW, so some photographers would be forced to use the full sized raw even though they don't want or need the full resolution.

So you want Canon to stop advancing its technology because of some software packages?

[DJL329]

If the pixel count followed Morse's law (like storage does), the 5D3 would have been 80MP.

I believe you meant Moore's Law, which refers to the number of transistors on an integrated circuit (which double every two years), not storage.

[CR Backup Admin]

If the pixel count followed Morse's law (like storage does), the 5D3 would have been 80MP.

I believe you meant Moore's Law, which refers to the number of transistors on an integrated circuit (which double every two years), not storage.

Wikipedia:

Moore's law describes a long-term trend in the history of computing hardware. The number of transistors that can be placed inexpensively on an integrated circuit doubles approximately every two years.[1] This trend has continued for more than half a century and is expected to continue until 2015 or 2020 or later.[2]

The capabilities of many digital electronic devices are strongly linked to Moore's law: processing speed, memory capacity , sensors and even the number and size of pixels in digital cameras.[3] All of these are improving at (roughly) exponential rates as well (see Other formulations and similar laws). This exponential improvement has dramatically enhanced the impact of digital electronics in nearly every segment of the world economy.[4] Moore's law describes a driving force of technological and social change in the late 20th and early 21st centuries.[5][6]