Last night the highly anticipated 5D mk III was announced and all official specs confirmed the rumours that have been doing the rounds on the internet. My first impression is that Canon followed the exact same recipe as with the 1Dx (low resolution, improved peripherals) and for that I’m thankful. I knew the launch of these cameras would be interesting, but I don’t think many people saw this reverse of roles in the pixel race. The pixel race is something that degrades image quality by making camera manufacturers money off people’s lack of understanding of true resolution. It’s like an abscessed sore in the camera development world that will hopefully be healed one day.
Nikon has now taken the lead with its 36mp D800, but it wasn’t willing to compete in the pixel race with the model that won it the global sports and journalism market?? They’ve increased the resolution of the D700 by 1.73x, while upping the resolution of the D3 only by 1.22x. There are obviously the speed, focus and noise areas where the D4 needs to perform much better than the D800, but looking at this fact I suspect the D800 is a confident throw of the dice to try and get their foot in a market that is dominated by the 5DmkII. I may have stuck faithfully to Canon over the past 3 years, but the way in which they compromise IQ for specs that sell cameras really pissed me off. The companies obviously just want to make money, but it’s insulting when they sacrifice the performance of their cameras for high pixel-counts that appeal to the Tom, Dick and Harrys that don’t know squat about resolution.
It’s been proven time and time again through many models from every brand that the higher the pixel density, the worse the image quality gets. Nikon faithfully stood by this knowledge and produced cameras with low resolution, great speed and focus that resulted in a great overall camera. Apart from the 1Ds III and 5D II, I can’t say I ever desired one of the non FF Canon cameras. The 1DmkIV was killed on launch by the D3s. The 7D’s pixel density was more than twice that of the 5DmkII and while its IQ did surprise me, it was a very useless 18mp.
Now it seems that Canon has taken the lessons it learned from those over-rezzed models and applied it in the development of what I hope will finally be their comeback from the knock that they’ve taken from Nikon over the past 4 years? They’ve been surviving off the consumer market and the 5D II. The money may be in the consumer market, but it’s the pros that build your reputation and 3 years ago it was like Canon pro bodies were Polish Jews all taken to Auschwitz under the rule of the D3.
Rape Victim of the D3s
Enough speculation, there are solid facts to look at.
22.3mp – I’ve had 21mp for 3 years now and I honestly can’t say I want more unless it’s on a larger format. In what situation will you truly benefit from 7-8mp more? I’m still very doubtful of whether Nikon’s new technology can truly pull out that much detail from a 35mm format lens, BUT, if the D800 can prove me wrong then I’ll get in my car and go buy one.
New AF system – I focus manually most of the time using Live View because you can zoom in 10 times to any part of the frame and make sure focus is dead on. I do however shoot on AF when shooting handheld and I know the focus isn’t reliable, so while it’s not a make or break issue it is a useful improvement for me.
Light metering, video functions and higher fps don’t really tickle my fancy, but it has all been considerably improved.
Improved Weather resistance – After a year of risky seascape shooting and one final wave giving me a sunset shower, my 5D II kicked the bucket. Upon analysis it was constant exposure to sea air and water that corroded the circuit boards around the buttons, so improved weather sealing is very welcome.
New sensor cleaning – They’ve added something at the bottom of the shutter box that absorbs the dust after it has been shaken from the sensor? I’d love to get a better understanding of how this works. All I know is that everyone HATES dust!
Larger, better LCD – Always welcome.
New HDR function – In camera HDR processing…I don’t want to say much about this, but I suspect the results to be very photomatix-like. I’ll wait and see.
Multiple Exposure Function – This I’m very excited for…Something you could do in the film days to create very interesting results. If you don’t know the potential of this then you don’t deserve to get excited about it!
If you go read the page on Canon’s website you’ll see that there are many other cool new functions like comparative playback, in camera processing, variable aspect rations…the list goes on!
100-25600 Native ISO range – Canon says that newly designed photodiodes, a better signal to noise ratio and improved image processing has upped noise performance by two stops. To most people that doesn’t mean much, but for landscapers wishing to go to the night skies it means a lot. I’ve always shot at 3200, which is good enough for web viewing but I wouldn’t print it. Two stops more will mean you can print ISO3200 files and you can shoot at ISO12800. Imagine this…pitch black wilderness sky, a milkyway so bright it casts a shadow under a Namibian tree…ISO12800, 30s, f/2.8. To say that the thought of that gets me very excited is an understatement. If you use a 24mm f/1.4 lens you can get another two stops, and if you’ve got a 1Dx (100-51200 native range) then you’ve got another stop. I’m confident that these new cameras are about to open one massive door to night sky photography and that is the thing I’m most excited about.
Conclusion
I’m going to try and type my conclusion as short as possible without elaborating on points. I would love to hear people’s arguments for or against certain things.
Megapixels – I don’t think that on the medium term the way forward for 35mm cameras is resolution. Even the sharpest lens can only project so much detail in a 36x24mm area and Nikon has gone upstream from their recipe for success of low res, high IQ. I can’t wait to see hi-res samples from the D800 because it will confirm or bust this theory. If you’ve seen the detail that LF film or MF digital produces, you’ll understand.
Live view – Up until the D3s, Nikon’s live view was very 3rd world compared to Canon. The ability to meter and zoom to x10 anywhere in the frame is a brilliant function. If Nikon has improved on this they’ve won serious brownie points from me.
Noise – Canon has a native ISO range going two stops higher, but Nikon has dominated noise performance over the past years. This is going to be interesting.
Highlights – Most readers of my blog know my passionate hate of Nikon’s highlight tones and colors in skies. If this hasn’t improved then there’s no chance of me buying a D800.
Price and availability – The canon is set to be a few $100 more and history has shown that Nikon is better at getting things on the shelf on time…
I honestly don’t know what to expect. These cameras will come packed with 3-4 years of R&D. My main standing point is that I’m doubtful about 36mp in a 35mm sensor. The proof will be in the pudding, so lets hope the pudding hits the shelves ASAP!
Loading...
Thanks Hougaard for combining a lot of info from different websites into one so quick!
To me as an amateur with a low budget the 5DIII is the ideal one-camera-does-it-all. Full frame sensor with 61 focusing points at 6 fps at this price range. An added bonus to me is to get multi-exposures back. Missed it from the film camera days.
On my budget switching to Nikon is not an option. With the 5DIII it is not at all necessary.
Whenever I can afford the 5DIII I will have a great combo in the 5DII and 5DIII.
Comment by Hannes Thirion — March 2, 2012 @ 10:25 am
Good solid objective post.. I like it.
I’m a Nikon shooter and was not so sure about the super high resolution of the D800. I just ordered a D4, with the biggest motivation being that they have kept the pixels relatively low. The increased res from 12 to 16mp give a bit more print res without potential loss of performance caused by pixel compression. I like that..
I’m hoping Nikons D800 will be a show stopper and be as good as they say, but I was also slightly concerned about the camera having to high a res..
In terms of the Canon 5Dmk 3, I’m very happy about for canon users
that they have not made too many dramatic improvements on a camera thats been received so well over the past couple of years.. My big concern is still the auto focus and spread of focus points.. I would have like to see a bigger coverage towards the edges for “focus lock and re- compose” handheld shooting apps.. The increased frame rate will be welcomed by all as will the increased ISO range and tolerance. If all else fails we can at least buy the latest Cell- phone camera from Nokia with 41mp.. I mean.. When needs a Hasselblad when you can shoot 41mp from your cellphone..:-D
Comment by hendre louw — March 2, 2012 @ 10:26 am
My frustration with the big two is that still photographers are having to pay the price for R&D for the improved video features. Whilst this benefits the video guys, there is little to get excited about as a still photographer.
Comment by Jon Reid — March 2, 2012 @ 10:42 am
“The pixel race is something that degrades image quality by making camera manufacturers money off people’s lack of understanding of true resolution. ”
This suggests to me that you don’t understand resolution, although I doubt that is the case! More megapixels=more detail as long as the lens can resolve the pixel density. We arent even close to the lens limit yet. The 5DmkII has the same pixel density as the 8mp 20D. It doesnt take a genius to see that given the performance of the 7D a 50mp camera would produce outstanding results, particularly given the improvements to signal to noise ratio.
“but the way in which..(Canon)..compromise IQ for specs that sell cameras really pissed me off”
“they sacrifice the performance of their cameras for high pixel-counts”
Both of these statements arguably only apply to consumer models. They are definitely not applicable to any of Canons 5D or 1D/1Ds lines. Even in the case of the consumer models I wouldnt say they were worse, just that the improvement was marginal.
“It’s been proven time and time again through many models from every brand that the higher the pixel density, the worse the image quality gets.”
I don’t think this has ever been displayed in DSLRs. The individual pixels may in some cases be ‘worse’ but the image as a whole is generally richer in detail and of higher quality overall.
“The 7D……. was a very useless 18mp”
Disagree. 100% crops bear out the fact that the 7D regularly resolves the full detail of its sensor. The IQ isnt up there with the mkII but if I was shooting telephoto subjects a 7D is clearly a better choice
As for you claim that the 5DmkIII will do well for Canon….At the current price it puts it completely out of the reach of amatuers and consumers who were the main market for the mkII! This camera will only sell well if the price comes down a lot!
“I’m still very doubtful of whether Nikon’s new technology can truly pull out that much detail from a 35mm format lens, BUT, if the D800 can prove me wrong then I’ll get in my car and go buy one.”
There is no doubt that it WILL prove you wrong. As I mentioned previously, that isnt even close to the limit of what lenses can resolve. If this is what matters to you then you really should buy one!
For me, the only point in buying a mkIII if you are a landscape photographer is for astrophotography or a (possible but unlikely) dynamic range improvement. Everything else wont effect the quality of images. Of course the fact is many landscape photographers dont JUST do landscape. There’s also that snob factor if you are a pro of people saying to you “oh, you only have a mkII”
All in all, I’m really happy with this announcement because I can spend my money on lenses instead!
Comment by Alex Nail — March 2, 2012 @ 4:16 pm
With regards to your resolution rant…Would you state absolutely that in no way did Canon sacrifice the overall quality of the image a camera like the 7D produces for the sake of resolution? It’s pretty obvious that a camera like the 7D would have performed much much better in all areas except resolution if it had the same pixel density as the 5D.
“It doesnt take a genius to see that given the performance of the 7D a 50mp camera would produce outstanding results” – Yes, if it does something like 0.2fps and has 3 digic V processors. I’m not saying that cameras shouldn’t go to such resolutions, but while they can already produce such sensors, the processing systems aren’t nearly powerful enough to turn the RAW data into an excellent quality image. They may also already have that technology, but they can’t put into a practical and economical camera they can put on the shelf. We’re on the opposite ends of this argument where I want a cleaner image and you want a larger image and to get the one they have to compromise the other. I would have been very happy if they made the 5D3 around 28mp with the same overall performance of the 5DII. Like I said, if the D800 proves me wrong then I will certainly buy one. I want a 14-24mm with a sore heart. There’s still the suspicion of a 5Dx, and remember that there might still be a large res pro body in the pipeline. Speculating is fun, but lets wait for some results from the D800 then either of us can say ‘I told you so’!
“For me, the only point in buying a mkIII if you are a landscape photographer is for astrophotography or a (possible but unlikely) dynamic range improvement. Everything else wont effect the quality of images. Of course the fact is many landscape photographers dont JUST do landscape. There’s also that snob factor if you are a pro of people saying to you “oh, you only have a mkII”
All in all, I’m really happy with this announcement because I can spend my money on lenses instead!”
I mentioned in my original post that the thing that excites me the most is the potential of astro-photography. My current 5D is old and I fear it may give in any day, so I need a new camera. I’ve also got an opportunity to buy a 2nd hand Ikelite housing for a 5D, and I’m definitely not going under water with my primary body. You know from the old days on DA that I used to be a gear whore, but I’ve shot happily with my 5D for over 3 years now. I want one of these cameras because I need a new and 2nd camera, and I want the one that will suit me the best, but that will come down to seeing some genuine full res results.
I originally paid R32000 for my 5DmkII and it sells for R19600 now. You know the price will drop substantially once those first bodies don’t sell out instantly any more. The prices will be interesting here as Nikon S.A has a much higher profit markup than Canon S.A because they offer free professional services. So chances are they’ll cost the same here, but there’s a reliable 3rd party importer with great prices that a lot of people have been buying from so I’ll probably buy from them.
Comment by Hougaard Malan — March 5, 2012 @ 11:37 am
Hi Hougaard
Photographer Nasim Mansurov has a discussion going on his blog about the D800, and I would also want to elaborate on this.
The question should maybe not be how the D800 will look on a per-pixel basis (worse than the 5Dmk3 I’m sure), but how it will compare after being scaled to the same resolution. Apparently the signal/noise of digital sensors (of a given generation) are similar per physical unit area – thus largely independent of the pixel count.
While it costs extra megabytes, processing time, and discipline, you could see a high MP camera as offering the user the choice of resolution vs noise – low noise being achievable by downsampling. The only thing you cannot regain is dynamic range. Since the D800 has the same pixel density as die D7000 I don’t expect too many dynamic range problems though. (the D7000 is pretty good here).
It will be interesting to see what the tests show (once they are out).
PS: I think the D800E is a useless gimmick for 99% of people. To be blogged about.
Comment by Francois Malan — March 5, 2012 @ 4:35 pm
That’s a very valid point Francois. There are certain competing aspects between the cameras where downsizing can make up for the difference in quality between the two. Not all of them, but quite a lot. I haven’t thought about it.
I can’t wait for in depth reviews of these cameras!
Comment by Hougaard Malan — March 5, 2012 @ 5:29 pm
Yes it goes without saying that you have to compare like with like. You can’t compare a 30″ print on a 40mp camera to a 15″ print on a 10mp camera and, after looking under a microscope conclude that per square inch the 10mp camera produces a better print and is therefore a better camera!
That’s why I dont understand you ‘sabbotage’ statement because there is no evidence showing any of canons digital SLRs getting worse because they have more pixels, largely because the technology has improved.
The 7D would have similar per pixel noise to the mkII if the pixels were the same size, but downsampled they would no doubt be comparable.
Equally introducing the suggestion that high MP are bad because processors/computers etc can’t keep up is shifting the goalposts somewhat. I would of course want the processors to keep up with the sensors, but there is no evidence to suggest that this is an issue. For example doubling the file size and halving the FPS would put similar strain on the processor.
Anyway I’m rambling, its all good
Comment by Alex Nail — March 6, 2012 @ 5:19 pm
While I’m not a full frame user (I use Nikon DX for the sake of weight size and price of the dual system of visible and infrared bodies) I do read reviews now and then. The strong push of the MP count on the FX sensor by Nikon might be valid, because Nikon’s main landscape lens (14-24) resolves much better than Canon’s 16-35II. The 16-35f/4VR is also sharper than the 17-40 f/4 by Canon.
However, I do understand that’s it’s all about pumping more money from the weathly average western European/American amateur user.
Comment by DimensionSeven — May 2, 2012 @ 10:40 pm