Michael, have I missed your 5D3 review?

[shadowblade]

I think that a 5D mkIII review and contrasting it with the D800 is well worth the effort. The 5D mkIII has lots of improvements over 5D mkII and in many ways a more general purpose camera than the D800.

I wouldn't say that. They both seem pretty general-purpose to me.

The D800 has the advantage if you need to print large - you can print 27% larger on each dimension while maintaining the same resolution (upsampling is useful, but doesn't give you more actual detail). At 24" print width, you're getting 307ppi with the D800, vs 240ppi with the 5D3 - not all that noticeable. At 36" width, you're getting 205ppi vs 160ppi - starting to be noticeable. At 60" width, you're getting 123ppi vs 96ppi - this is very noticeable.

Both are good in different forms of challenging lighting. In extreme low-light situations, the 5D3 seems to have a slight edge at ISO 25600 and above; they are about equal at ISO 3200-12600 (I couldn't tell the difference on 16x24" prints) and the 5D3 is decidedly inferior below that (similar noise on the pixel level, but the D800 has 70% more pixels). So, if you're regularly shooting at greater than ISO 12800, the 5D3 might be better. But, really, if I were shooting at such high ISOs routinely, I'd be looking at the D4 or 1Dx, rather than either the 5D3 or D800. In situations of high dynamic range, there's no contest - the D800 is simply better.

That the D800 came around at 36MP should not really have been a big surprise since we already saw what such a sensor could bring from the D7000 sensor and other cameras using the same sensor. The next step up for Nikon (and Sony)  for full frame would logically be a FF sensor based on the 24MP APS-C sensor giving a 54MP sensor.

To me, this is a huge milestone - 150ppi at 40x60" print size, or 300ppi at 20x30" print size. The issue, of course, is lenses can keep up in the corners (time for a 14-24 Mk II from Nikon, maybe, and definitely time for a 16-35 Mk III (or a 14-24 Mk I) from Canon). Canon's TS-E 17mm and 24mm are definitely up to the task; I wouldn't mind seeing a TS-E 14, though, for tnose situations where you want to capture all the width in a single frame, but need the tilt function for focal plane control (would be OK even if the shift were limited, as long as there is tilt). Nikon's PC-E lenses probably need a revamp - and, given that Nikon's latest camera is pushing the boundaries of medium format, and many medium-format photographers need their movements, there's every reason for Nikon to be heading in this direction (the recently-patented PC-E 17 may be indicative of this).

What is more surprising is that Canon chose to stay at 22MP and not base a new FF sensor on the 18MP APS-C sensor which would give a 46MP sensor.

Because it's sufficient for event and wedding photographers. But, with no high-resolution replacement for the 5D2 and 1Ds3, it's left studio and landscape photographers deeply disappointed. After all, many people bought into the 5D2, coming from MF film, when it came out for its high resolution - 21MP was huge in 2008, when the standard was 12MP - not for its other features. These same photographers are still shooting the same things - requiring high resolution and dynamic range, not so much high ISO or frame rate - but, for their purposes, the 2012 model is no better than the 2008 model.

Thoughts about these choices and why they were made and where we are heading for the future (and as mentioned what is relevant for real photography of various types) might be more interesting than how much you can lift shadows from the D800

For sure, but definitely not to downplay the DR advantage. The superior shadow detail is a huge deal. If you've ever taken a landscape with a recent MFDB (single frame, not multiple exposures), exposing for the highlights then recovering detail in the shadows, you know how detailed and noise-free the shadows are. You can do the same with the D800 and D7000. With the 5D2, all you get is mush and pattern noise. There are many situations in photography where you can't take multiple exposures (moving objects), can't use GNDs (uneven horizon) and can't use fill flash (distances too great) but need detail from shadows as well as highlights (usually involving backlit objects during the 'golden hours', or silhouetted against the sky).

Come to think of it, the D3x was announced on December 1, 2008, when the 5D2 had been available for barely a month, and still had the 'black dots' issue. If, instead of releasing a $8000 camera, Nikon had put the same sensor in the D700 and called it the D700x, the 5D2 would have been stillborn - many of those who moved from MF film to the 5D2 would have, instead, moved to the D700x. Similar price band, higher resolution, better DR, better AF, better weather-sealing. Slower frame rate, but no-one ever bought the 5D2 for action shooting anyway.

[Ray]

P. S. Given that extreme closeness, I am not sure where Ray gets his claim of better high ISO noise performance for the 5D3, let alone his apparent sentiment that this high ISO noise advantage plus 6fps vs 4fps was enough reason for Canon to choose a sensor design with substantially inferior resolution and dynamic range. Perhaps some comparison involving a choice of lenses that eliminates or reverses the pixel count advantage of the D800, making the DxO "screen" comparison more relevant.

Why so negative, BJL? Try to think positively. 

This is what I wrote regarding the 5D3's high ISO performance:

"Although the D800 has an impressively high pixel count which exceeds the pixel count of the 5D3 by a worthwhile margin of 60% or so, its high-ISO performance is no better than the 5D3, and if anything, slightly worse."

The expression, "if anything, slightly worse", means if one were to find a difference through extreme pixel-peeping, such difference would favour the 5D3, according to my interpretation of the DXO results.

Another trend in the pattern of Canon's release of new models, which I think is relevant here, is its history of always releasing its highest-resolving sensors first in its expensive, fully professional models, the 1Ds series.

This policy is presumably a business decision rather than a technological limitation. The original 12.7mp 5D followed the 16mp 1Ds2. The 21mp 5D2 followed the 21mp 1Ds3. If this trend were to continue, the 1Ds4 would be the next camera with an increased pixel-count, possibly followed by a 5D4 with the same or similar pixel-count.

Nikon have been very keen to break out of this pattern, to their advantage. It was a big surprise to everyone when Nikon introduced a more affordable D700 so soon after the release of its first full-frame DSLR, the D3. The D700 seems to have equal specs in most areas, yet was a lighter camera and much more affordable. The sort of camera that appeals to me.

With the release of the D800, Nikon have completely broken out of the mold and produced a 'prosumer' camera at a relatively affordable price, which exceeds in performance in certain significant areas, such as resolution, their fully professional models, such as the D3X.

It is in this sense they have stolen a march on Canon and changed the paradigm in the process. I can imagine that Canon has a 1Ds4 in the development pipeline, with maybe a 46mp sensor, perhaps scheduled for announcement at Photokina in September.

However, the Canon Board of Directors and the Marketing Department are probably now wondering how their 30mp, or 40mp or 46mp top-of-the-line model which they intended to price at $8,000, can compete with the Nikon 36mp prosumer model with its two option of D800 and D800E.

It would be interesting to eavesdrop on the board room discussions. 

[BJL]

Ray, you still leave me wondering why you have repeatedly raise this slight or non-existent advantage 5D3 as if it somehow balances the vastly greater IQ advantages of the D800.

As to Canon always releasing its highest resolution sensor in its top model, you seem to have missed the memo: times have changed, Canon has clearly described the 1DX as _the_ new high end model, as surely as the D4 is _the_ new high end Nikon model, both in the sense of professional build quality, price and so on. There are no new Canon 1Ds or Nikon D?X models coming: those big, square, heavy, bullet-proof, maximum resolution, lowish frame rate, medium format challenging $8000 models are gone, and the two new 35mm format models from each of Canon and Nikon are it for this year. And I am sure that a great many landscape, architectural and other high resolution seeking photographers are very happy for the saving in weight and money that they are getting from new breed, both the D800 and 5D3.

You can imagine all kinds of wonderful new sensors and models coming, but I prefer to deal with the available facts.

P. S. when it comes to thinking negatively, your persistent attention on small to non-existent disadvantages of the D800 is a fine example!

[Ray]

Ray, you still leave me wondering why you have repeatedly raise this slight or non-existent advantage 5D3 as if it somehow balances the vastly greater IQ advantages of the D800.

Crikey! I'll explain it further. The thrust of my point was that at high ISO the 5D3 is the equal of the D800. The reference to the 5D3 possibly having a slight edge, if one were to engage in pixel-peeping, was merely to reinforce the point that at high ISO the D800 really does have no advantage, in case anyone is doubtful. I can't be more explicit than that, surely.

The reference to the 7D was to make the point that the cropped-format of that camera, used in conjunction with the 5D3, would tend to cancel some of the benefits of the D800, specifically that the D800 in DX mode is close to being a D7000, useful to extend the reach of one's longest lens.

As to Canon always releasing its highest resolution sensor in its top model, you seem to have missed the memo: times have changed, Canon has clearly described the 1DX as _the_ new high end model, as surely as the D4 is _the_ new high end Nikon model, both in the sense of professional build quality, price and so on. There are no new Canon 1Ds or Nikon D?X models coming: those big, square, heavy, bullet-proof, maximum resolution, lowish frame rate, medium format challenging $8000 models are gone, and the two new 35mm format models from each of Canon and Nikon are it for this year. And I am sure that a great many landscape, architectural and other high resolution seeking photographers are very happy for the saving in weight and money that they are getting from new breed, both the D800 and 5D3.

You seem very confident about this, BJL. It seems clear to me that the 1Dx is a discontinuation of the 1D series, not a discontinuation the 1Ds (or full-frame) series. Whether it's called 1Ds or 1Dx is irrelevant. A name is a name.

The fact remains, the 1Dx is a full-frame camera with the sort of performance one expects from a 1D series camera, such as a very high frame rate. The price in Australia is around the $8,000 mark. The weight is a secret, so I think it would be reasonable to guess it's probably as heavy as the traditional 1Ds. Just why the weight of the 1Dx is not included on all references to its specifications, that I can find on the internet, is a mystery. People are apparently placing pre-orders for a camera without knowing its weight.

Until DXO tests the 1Dx, its sensor performance is speculation. There's no reason to suppose than Canon is not working on a lighter and more affordable 46mp 5D4, with a much slower frame rate than the 1Dx. Such a camera may have begun its development as a 1Ds4, but as I mentioned, if you read my post more carefully, the release of the D800, and no doubt the forewarning of the D800 that Canon would have been aware of, before its announcement to the public, would have caused a rethink about an expensive and high resolution 1Ds4.

Reading your arguments here, I'm reminded of a popular thread a few years ago on this forum in which I presented the argument that Nikon would eventually have to offer a full-frame format in order to compete with Canon. You insisted there was no evidence at all that Nikon was working on a full-frame model, and you appeared to firmly believe they would stick with the DX format.

P. S. when it comes to thinking negatively, your persistent attention on small to non-existent disadvantages of the D800 is a fine example!

That's a very strange comment considering I've already mentioned that I've placed an order for a D800E. I've also put down a deposit. I expect to receive the camera before the end of the month.

[Hans Kruse]

I wouldn't say that. They both seem pretty general-purpose to me.

The D800 has the advantage if you need to print large - you can print 27% larger on each dimension while maintaining the same resolution (upsampling is useful, but doesn't give you more actual detail). At 24" print width, you're getting 307ppi with the D800, vs 240ppi with the 5D3 - not all that noticeable. At 36" width, you're getting 205ppi vs 160ppi - starting to be noticeable. At 60" width, you're getting 123ppi vs 96ppi - this is very noticeable.

The 6fps on the 5D mkIII is a real difference from the 4fps on the D800. For action photography I would have no doubt which one to choose. So the camera that covers different disciplines I find the 5D mkIII a more general purpose camera than the D800. As far as I can tell the only two advantages the D800 hold is the larger DR at low ISO and the resolution advantage which more or less disappears at higher ISO.

Regarding resolution difference I would like to see a double blind test of prints made from both cameras with equal (good) technique and see when and if a difference can be validated in a scientific test. I even doubt there would be a very significant difference observed at least until you go to A0 print size and bigger. I'm sure a reviewer would see a difference before that, but he has the problem of knowing and therefore suffers from confirmation bias.

Both are good in different forms of challenging lighting. In extreme low-light situations, the 5D3 seems to have a slight edge at ISO 25600 and above; they are about equal at ISO 3200-12600 (I couldn't tell the difference on 16x24" prints) and the 5D3 is decidedly inferior below that (similar noise on the pixel level, but the D800 has 70% more pixels). So, if you're regularly shooting at greater than ISO 12800, the 5D3 might be better. But, really, if I were shooting at such high ISOs routinely, I'd be looking at the D4 or 1Dx, rather than either the 5D3 or D800. In situations of high dynamic range, there's no contest - the D800 is simply better.

To me, this is a huge milestone - 150ppi at 40x60" print size, or 300ppi at 20x30" print size. The issue, of course, is lenses can keep up in the corners (time for a 14-24 Mk II from Nikon, maybe, and definitely time for a 16-35 Mk III (or a 14-24 Mk I) from Canon). Canon's TS-E 17mm and 24mm are definitely up to the task; I wouldn't mind seeing a TS-E 14, though, for tnose situations where you want to capture all the width in a single frame, but need the tilt function for focal plane control (would be OK even if the shift were limited, as long as there is tilt). Nikon's PC-E lenses probably need a revamp - and, given that Nikon's latest camera is pushing the boundaries of medium format, and many medium-format photographers need their movements, there's every reason for Nikon to be heading in this direction (the recently-patented PC-E 17 may be indicative of this).

Nikon has the 14-24 lens, but on Canon there is the TS-E lenses 17mm and 24mm which have no Nikon equivalent. At 14mm there is the Samyang 14mm f/2.8 UMC which is as good as the Nikon 14-24 at 14mm. There is Zeiss lenses as well with a Canon mount, so I would say Canon is not left behind here and actually has an edge over Nikon. On the longer lenses Canon has the 70-200 f/2.8L IS II which is the best lens in this category ever made. That lens is very useful for landscape photography.

Because it's sufficient for event and wedding photographers. But, with no high-resolution replacement for the 5D2 and 1Ds3, it's left studio and landscape photographers deeply disappointed. After all, many people bought into the 5D2, coming from MF film, when it came out for its high resolution - 21MP was huge in 2008, when the standard was 12MP - not for its other features. These same photographers are still shooting the same things - requiring high resolution and dynamic range, not so much high ISO or frame rate - but, for their purposes, the 2012 model is no better than the 2008 model.

For sure, but definitely not to downplay the DR advantage. The superior shadow detail is a huge deal. If you've ever taken a landscape with a recent MFDB (single frame, not multiple exposures), exposing for the highlights then recovering detail in the shadows, you know how detailed and noise-free the shadows are. You can do the same with the D800 and D7000. With the 5D2, all you get is mush and pattern noise. There are many situations in photography where you can't take multiple exposures (moving objects), can't use GNDs (uneven horizon) and can't use fill flash (distances too great) but need detail from shadows as well as highlights (usually involving backlit objects during the 'golden hours', or silhouetted against the sky).

I shoot a lot of landscapes using a Canon 1Ds mkIII and yes, I would not downplay the resolution and DR advantage of the D800, but in my opinion there is quite some misinformation around as if shadows can't be lifted at all on a Canon sensor. That is clearly not true. I do that all the time. Extreme lifting of shadows like more than 3 stops does not work well and then exposure blending comes in which is pretty easy to do using Lightroom and Photoshop Pro and bracketing is needed anyway these days as you can't see the true clipping of the sensor and what Lightroom 4 can recover without loss by the new algorithms (I mean on the camera LCD you can't see this). You can see examples of this here http://forums.dpreview.com/forums/read.asp?forum=1032&message=41483832 and you can see the originals here http://forums.dpreview.com/forums/read.asp?forum=1032&message=41484204 and this is really an extreme example where exposure blending could result in a slightly better shadows rendering. But don't forget that we are looking at this in 100% view and in a normal print the noise would not even be seen.

Come to think of it, the D3x was announced on December 1, 2008, when the 5D2 had been available for barely a month, and still had the 'black dots' issue. If, instead of releasing a $8000 camera, Nikon had put the same sensor in the D700 and called it the D700x, the 5D2 would have been stillborn - many of those who moved from MF film to the 5D2 would have, instead, moved to the D700x. Similar price band, higher resolution, better DR, better AF, better weather-sealing. Slower frame rate, but no-one ever bought the 5D2 for action shooting anyway.

In my opinion it was Nikon who had a camera that was stillborn, namely the D3X which was the last of it's kind and the 5D mkII was the camera that ended the high priced, high resolution models. This we have seen now with the 1DX ending the two camera 1D and 1Ds series. The D800 is not getting a higher prices Nikon sibling in the $8000 range and on the Canon side the 1Ds mkIII was the last of it's kind. I'm sure Canon will release a high resolution body later on. Surely many Nikon shooters using the D700 did not switch to the 5D mkII for landscape since they were confident that Nikon would launch a higher rez model at some point. This was true although it took longer than expected and probably because of the catastrophis in Japan and later in Thailand.

Please don't get me wrong, I think the D800 and D800E are great new cameras, but it does not mean that the competition is not there at all.

And yes, I have seen files from MF cameras and they can provide higher resolution, but the DR is a bit over rated from what I have seen, but I don't own one so bare with me

[shadowblade]

The 6fps on the 5D mkIII is a real difference from the 4fps on the D800. For action photography I would have no doubt which one to choose. So the camera that covers different disciplines I find the 5D mkIII a more general purpose camera than the D800. As far as I can tell the only two advantages the D800 hold is the larger DR at low ISO and the resolution advantage which more or less disappears at higher ISO.

I'd consider fast-FPS action shooting to be a pretty narrow field - you're either into it or you're not. If action photography is your thing, you'd probably choose the 1Dx or D4, or even an older 1D4 or D3s, over either of these cameras. If action photography isn't your thing, then 4fps is more than fast enough anyway. After all, it's still faster than the 5D2...

Regarding resolution difference I would like to see a double blind test of prints made from both cameras with equal (good) technique and see when and if a difference can be validated in a scientific test. I even doubt there would be a very significant difference observed at least until you go to A0 print size and bigger. I'm sure a reviewer would see a difference before that, but he has the problem of knowing and therefore suffers from confirmation bias.

The difference in linear resolution is similar to the difference between the D700 and 5D2. One doesn't have to print too large to see the difference between these two - you can easily tell the difference at a 16x24" print size. Given that 22MP equates to around 30% greater linear resolution as compared to 12MP, you would expect to see a definite difference at 24x36" print size.

Nikon has the 14-24 lens, but on Canon there is the TS-E lenses 17mm and 24mm which have no Nikon equivalent. At 14mm there is the Samyang 14mm f/2.8 UMC which is as good as the Nikon 14-24 at 14mm. There is Zeiss lenses as well with a Canon mount, so I would say Canon is not left behind here and actually has an edge over Nikon. On the longer lenses Canon has the 70-200 f/2.8L IS II which is the best lens in this category ever made. That lens is very useful for landscape photography.

I wouldn't disagree on that. But Canon doesn't have a high-resolution full-frame camera to take advantage of all that...

I shoot a lot of landscapes using a Canon 1Ds mkIII and yes, I would not downplay the resolution and DR advantage of the D800, but in my opinion there is quite some misinformation around as if shadows can't be lifted at all on a Canon sensor. That is clearly not true. I do that all the time. Extreme lifting of shadows like more than 3 stops does not work well and then exposure blending comes in which is pretty easy to do using Lightroom and Photoshop Pro and bracketing is needed anyway these days as you can't see the true clipping of the sensor and what Lightroom 4 can recover without loss by the new algorithms (I mean on the camera LCD you can't see this). You can see examples of this here http://forums.dpreview.com/forums/read.asp?forum=1032&message=41483832 and you can see the originals here http://forums.dpreview.com/forums/read.asp?forum=1032&message=41484204 and this is really an extreme example where exposure blending could result in a slightly better shadows rendering. But don't forget that we are looking at this in 100% view and in a normal print the noise would not even be seen.

I'd be very hesitant to lift even two stops on a 5D2. But, then again, I make very large prints, so require fairly tight tolerances in image quality.

In my opinion it was Nikon who had a camera that was stillborn, namely the D3X which was the last of it's kind and the 5D mkII was the camera that ended the high priced, high resolution models. This we have seen now with the 1DX ending the two camera 1D and 1Ds series. The D800 is not getting a higher prices Nikon sibling in the $8000 range and on the Canon side the 1Ds mkIII was the last of it's kind. I'm sure Canon will release a high resolution body later on. Surely many Nikon shooters using the D700 did not switch to the 5D mkII for landscape since they were confident that Nikon would launch a higher rez model at some point. This was true although it took longer than expected and probably because of the catastrophis in Japan and later in Thailand.

Exactly - the D3x failed because of the 5D2's existence. The 5D2 made a high-resolution (for its time) sensor available for around $3-4k, although without the AF and other features of the 1Ds series. It was certainly no EOS-3 in digital form. The D800 completes that transition of high-resolution bodies to the $3-4k price point, by including a good autofocus system. But the delay in bringing out a high-resolution, low-cost body has cost Nikon dearly - when the 20MP threshold was crossed, a lot of photographers were comfortable moving from MF colour film to full-frame digital, since similar image quality was now possible. Since Canon had the product, and Nikon did not, many of those moved to Canon.

Now, let's imagine that, less than a month after the 5D2 hit the shelves, Nikon had announced not the $8000 D3x, but the $3500 D700x, with the same sensor as the D3x, but in the body of the D700. Then, on the one hand, you'd have the 21MP 5D2, with poor weather sealing and poor autofocus, shooting at 3.9fps. On the other hand, for a similar price, you'd have the 24MP D700x, with weather sealing and AF equal to the D3, great dynamic range, shooting at around 1.5fps. Guess which one would have sold like hotcakes, while the other sat on the shelves? The fps difference is moot - after all, no-one bought either the 5D2 or the D3x to shoot action. The 5D2's AF couldn't keep up, while the D3x's frame rate couldn't keep up.

By releasing the 24MP Exmor-type sensor in a D3-style body, rather than in a D700-style body for half the price, Nikon made a huge strategic error, essentially handing Canon the high-resolution studio and landscape crowd (the D3x isn't good for much else, so it can hardly be said that they weren't chasing that crowd) and ensuring the 5D2 would dominate for a number of years, despite its numerous shortcomings.

[BJL]

You seem very confident about this, BJL. It seems clear to me that the 1Dx is a discontinuation of the 1D series, not a discontinuation the 1Ds (or full-frame) series. Whether it's called 1Ds or 1Dx is irrelevant. A name is a name.

There's no reason to suppose than Canon is not working on a lighter and more affordable 46mp 5D4, with a much slower frame rate than the 1Dx.

You insisted there was no evidence at all that Nikon was working on a full-frame model, and you appeared to firmly believe they would stick with the DX format.

Ray,
As to whether the 1DX is the only top model in the Canon line, I prefer to go by what Canon has said rather guessing based on the name! I suggest you reread what Canon has said. But it is impossible to absolutely prove a negative prediction like this, in the face of generic, irrefutable but utterly unpersuasive double-negative  arguments of the form "there is no reason why Canon is not working on ..." I have stated evidemce from Canon's own statements and the history of its product release cycle for my prediction; beyond that, the only way to prove a negative prediction is to wait.

By the way, as to Nikon and "FF", if you check, my skeptical comments were based on refuting recurring specific rumors that such a camera was about to come, and I was right every time: those rumors were (rather clearly) false and wishful thinking. When the real rumors of the D1X came, backed by sources like Thom Hogan, I dropped my skepticism.

By the way, if you want to argue ad hominiem about a single wrong prediction in the past, that is easy to do, is it not? Us long-time LL discussants have all made mistakes in numerous discussions over the years. For example you rather persistently predicted that the size advantage of smaller formats, in particular 4/3" vs larger formats, through the telephoto/macro reach advantage, would go away, because pixel sizes would equalize between the various DSLR formats, so that using the same focal length in a larger format and cropping would be a match. Back in the early days, 4/3” was at 5MP while APS-C was at 6MP (or 8MP in one Canon model) and 35mm format was at 14MP from Kodak, 11MP from Canon. Today, 4/3" offers 16MP, APS-C systems offer from 18 to 24 MP depending on brand, and in 35mm, Nikon offers up to 36MP, Canon offers up to 22MP. The ratios of pixel counts have roughly held steady (as I instead predicted), except that Canon's 35mm format offerings have had less resolution growth than the rest, and overall the pixel count increase has been slower in 35mm format than in the smaller formats. This is due I think to the law of diminishing returns when pixelx counts get so high: more than about 20MP is a rather specialized need, even forprofessional photography. So I am not holding my breath for a 35mm format DSLR with the 64MP sensor needed to match 4/3” for resolution with a crop using the same telephoto or macro focal length. Even the APC-C sensors of 30MP from Sony or 25MP from Canon as needed for equal pixel size are not in sight.

[BJL]

@shadowblade and @hans,
It seems that me might all three agree that the 5D2 severely undermined the market for the D3X, and I would go further: the 5D2 was disruptive, in moving the maximum resolution and image quality of a 35mm format system to a far lower price level, severely undermining the market for the entire category of cameras like the D3X and 1Ds series, which for a substantially higher price, add only "around the edges", not to the core of image quality. And it seems that both Nikon and Canon have responded to this by upgrading those far more afforable product lines with just the most important things like top of the line AF and VF, to produce the 5D3 and D800 as their new generation of highest resolution models, and ending the $7000-8000 1Ds and D?X product lines.

I see this as a positive trend, in greatly lowered price of access to the best image quality that 35mm format can deliver and eliminating any "hobbling" of those models with inferior AF or such.  However, I am sure some people will interpret it as a negative prediction, and as always, the only way to resolve negative predictions is to wait. Photokina 2012 should be long enough to wait in order to have an almost certain picture of the complete new generation of 35mm format cameras from Canon and Nikon.

[budjames]

I don't think that I need a Canon 5D MarkIII review from Michael. I have mine, and it's awesome compared to the 5D Mark II.

The focusing system and FPS improvements are radically better. The image quality is better and I really love experimenting with high ISOs that used to be worthless on my 5D Mark II and previous Canon bodies.

It so good that I'm selling my Canon 1Ds Mark III with less than 10,000 actuations on it. Anyone interested in making me an offer for it?

Cheers.
Bud James

[ndevlin]

If you want to see a good comparison between the 5D3 and D800, just watch The Camerastore's excellent video's here:

http://www.youtube.com/watch?v=omTo7UxbJX8&list=UUqpOf_Nl5F4tjwlxOVS6h8A&index=3&feature=plcp

http://www.youtube.com/watch?v=4W9EeDCaVFM&feature=relmfu

These show how the cameras differ for real photographers.

- N.

btw: the suggestion that the D3x was stillborn struck me as curious. Nikon sold boatloads of them at $8K a pop! That price includes about $4K of cream for corporate, and they milked that bitch for like four years. "You want Nikon FF, you pay." Pure. Genius.  The mediocrity of the 5DII as a camera, and Canon's normalt-to-wide lenses, along with the size of their pro install-base, is what let Nikon get away with this.  The D800 is a recognition that that gig is up.
 

[Ray]

Ray,
As to whether the 1DX is the only top model in the Canon line, I prefer to go by what Canon has said rather guessing based on the name!

BJL,
I was addressing your comment that "those big, square, heavy, bullet-proof, maximum resolution, lowish frame rate, medium format challenging $8000 models are gone".

It seems clear that only the 'maximum resolution' aspect is gone from those heavy and expensive cameras, for the time being at least.

The current top-of-the-line Nikon D3X and D4, Canon 1Ds3 and 1DX are all heavy and expensive cameras, irrespective of their pixel-count. The street price in Australia ranges from around $7,500-$8,000, and they all weigh around 1.3Kg or greater, including the new 1DX.

There would appear to be a paradigm shift led by Nikon, in the sense that the highest resolving cameras from both companies are now in the more affordable prosumer category. Whilst Canon's highest resolving camera is only 1mp higher than the 1Ds3 and of no significance resolution-wise, it is perhaps symbolic and a portent of things to come.

I suggest you reread what Canon has said. But it is impossible to absolutely prove a negative prediction like this, in the face of generic, irrefutable but utterly unpersuasive double-negative arguments of the form "there is no reason why Canon is not working on ..." I have stated evidemce from Canon's own statements and the history of its product release cycle for my prediction; beyond that, the only way to prove a negative prediction is to wait.

Of course it's impossible to absolutely prove a negative prediction. I thought the discussion is about what is likely, what is reasonable, and what is sensible. I doubt that Canon will allow Nikon to completely dominate the full-frame prosumer market for DSLRs.

For example you rather persistently predicted that the size advantage of smaller formats, in particular 4/3" vs larger formats, through the telephoto/macro reach advantage, would go away, because pixel sizes would equalize between the various DSLR formats, so that using the same focal length in a larger format and cropping would be a match.

Nope! My predictions tend to be positive. Any persistent claims I've made regarding the future of the 4/3 formats have been along the lines that the smaller format can never compete with the larger formats in terms of ultimate image quality, all else being approximately equal or equivalent.

I've always recognised the weight and bulk advantage of the smaller formats and in particular the high quality of many of the Zuiko lenses. Each of us has to make compromises regarding price, weight, performance flexibility, and ultimate image quality, and such compromises will be different according to different purposes.

For the past few years years I've frequently walked around carrying two cameras when I'm out shooting, such as the Canon 5D with Sigma 15-30 plus the Canon 20D with 24-105/F4, or the D700 with 14-24/2.8 plus the Canon 50D with 17-55/2.8 or more recently the D7000 with 24-120/F4 in place of the 50D.

However, I appreciate that the combined weight of two DSLRs and two lenses around one's neck and shoulders would be too much for some folks. That 14-24/2.8 is a heavy lens. One advantage of the D800E for me is a weight reduction of around 800gms. I'll carry just the one camera with two lenses, the disadvantage being I could miss a good shot whilst changing lenses.

[BJL]

BJL,
I was addressing your comment that "those big, square, heavy, bullet-proof, maximum resolution, lowish frame rate, medium format challenging $8000 models are gone".

It seems clear that only the 'maximum resolution' aspect is gone from those heavy and expensive cameras, for the time being at least.

Exactly! That is all I was saying: these big rugged "square" bodies are now back to what they were in the film era: all-round professional tools for sports, action, photojournalist and such, but not for the very high resolution "medium format replacement" sector where the extra burden of cost and weight does not deliver sufficient benefit.

Let me make another wild prediction: the next top of the line "square" Canon model will be in spirit a "1D X Mk II" with the 5760 x 3840 resolution of the 5D3, for its perfect match to decimation/downsampling to 1920x1080 HD video output for journalists and such. Or maybe just a little more like 6144x4096, if the wider aspect ratio 2K video format, 2048 pixels wide, become commercially important. (That 2K fits better with the currently dominant digital cinema projectors.)

Canon might well have made a sound business decision that this roughly 22MP, "video friendly" pixel count is a sweet spot of resolution for its high end gear, and one that could still make sense even if it can substantially reduce its noise floor and so increase the dynamic range in its next generation of sensors.

[BernardLanguillier]

Exactly - the D3x failed because of the 5D2's existence.

By releasing the 24MP Exmor-type sensor in a D3-style body, rather than in a D700-style body for half the price, Nikon made a huge strategic error, essentially handing Canon the high-resolution studio and landscape crowd (the D3x isn't good for much else, so it can hardly be said that they weren't chasing that crowd) and ensuring the 5D2 would dominate for a number of years, despite its numerous shortcomings.

It seems clear that Nikon should have released a D700x, but I disagree about the D3x being a failure.

It did sell well for many years, was a much superior camera compared to its main competitor, the 1Ds3, and simply redefined our expectations in terms of DR. While expensive it was competing performancewise with MF backs costing twice as much.

Besides, it served a purpose in clarifying to the market that Nikon is aiming, successfully, at delivering the highest level of performance in DSLRs. Want a good deal? Buy Canon or Sony. Want the best? Get a Nikon. Things get of course interesting when the D800 is both the highest performing camera and the better deal. :-)

In terms of system, the 17mm T/S, a very niche product, is probably the only lens in Canon's line up that delivers something the Nikkor really cannot do. For the rest we are talking about tiny differences with one brand topping the other in one direction or opposite. Nikon is rumored to release a 17mm T/S soon following the publication of a patent. Considering the overall domination they have in wide angle technology, there is no reason to think it will not be best in class.

Cheers ,
Bernard

[Josh-H]

was a much superior camera compared to its main competitor, the 1Ds3

UTTER BS.

[BernardLanguillier]

UTTER BS.

Why do you say so?

The D800 and D3x have roughly the same performance sensorwise at lower ISO and the D800 is said to be significantly superior to the successor of the 1D3s. The D3x was simply at least one generation ahead of its time.

On other metrics, we can mention AF (the Rob Galbraith saga), auto ISO, weather proofness (the Patagonia trip),...

Does it mean that the 1Ds3 is not an excellent camera? Not at all obviously.

Cheers,
Bernard

[tom b]

I went to the PPA show last year in Sydney. Lots of A2 size prints from both Canon and Nikon photographers. The quality was excellent and I don't think that anyone would have walked away from the show thinking they were looking at anything but professional quality images from both camps.

What I think of is a worry for Nikon and Canon is that there is no buzz for either the D4 or 1D X. Maybe the market has matured enough that it is the same as for computers, what we have is good enough.

Cheers,

[kers]

I am sure the Canon is a good camera; It is an improved version of the 5MK II in line of its name.
But the vision of Canon remains unclear. To me it seems their focus is at 4K video more than on static Images.

With the D800 Nikon shows more progression and vision. There is so much improvement even over their former flagship the D3x.
Their problem is that only a few Nikon Lenses deliver 36MP

So now we have Nikon with the best DSLR body and lenses for 24MP and on the other side Leica With 18 MP bodies and 50+ ( who knows?) lenses.
best 50mm Nikon lens  €350  vs best Leica 50mm lens 7000€

In that respect Canon is more balanced.

[dreed]

FWIW, it is widely believed that the 5D2 also killed a large portion of the market that was interested in the 1Ds3.

The 5D2 was a very disruptive product. You've got it (and the 7D) being used in TV shows and movies. From "House" to "The Avengers". As a camera, pro's were using it for considered work as well as amateurs that saved their pennies. IMHO, the 5D3 is attempting to bank on the success of the 5D2 and not much more as if you're not already sold on Canon, why would you pickup the 5D3 now and not the D800?

[shadowblade]

I went to the PPA show last year in Sydney. Lots of A2 size prints from both Canon and Nikon photographers. The quality was excellent and I don't think that anyone would have walked away from the show thinking they were looking at anything but professional quality images from both camps.

What I think of is a worry for Nikon and Canon is that there is no buzz for either the D4 or 1D X. Maybe the market has matured enough that it is the same as for computers, what we have is good enough.

Cheers,

A2 isn't very large - you'll mostly see the differences at 20x30" and larger, and more so on glossy/metallic surfaces.

Regarding the D4 and the 1Dx, I think this is mainly because their only advantage over the 5D3 and D800 is relevant only for one kind of photographer - those shooting high-speed action who absolutely require the 11/12/14fps frame rates. In other words, mainly sports photographers - the pixel density just isn't there to make the bodies particularly attractive for wildlife photography. We haven't seen what their ISO performance is like yet - if they have a significant edge over the D800 (after scaling to equal resolution) they may also have an audience in those shooting in extreme low light (starlit/moonlit landscapes without trails or with moving elements, e.g. waves, or concert/stage photography). For all other photographers, the smaller bodies have the advantage of less weight, greater resolution, less cost and, in the case of the D800, more dynamic range. If you don't need the high frame rate, why would you spend twice as much for a body with lower image quality, then shoot it at 1 frame every 2 seconds because that's all you need anyway?

After all, the D800 is essentially a 1Ds3 updated for 2012.

[dreed]

Regarding the D4 and the 1Dx, I think this is mainly because their only advantage over the 5D3 and D800 is relevant only for one kind of photographer - those shooting high-speed action who absolutely require the 11/12/14fps frame rates. In other words, mainly sports photographers - the pixel density just isn't there to make the bodies particularly attractive for wildlife photography.

I'm curious as to whether they'll be a hit with runway model photographers, unless that also qualifies as "sports photography."