Canon EOS 5D MkIII

[klindup]

I have a very simple view when considering a new camera such as the 5D MKlll.  When I bought my 40D it met my needs and my needs have not changed ergo the 40D still meets my needs.  I reckon that the thing that limits the quality of my images is still my ability as a photographer.  If I want a really large print I revert to film and a 6x6 camera.  Not sure how a planar lens compares with a Canon lens for sharpness but I guess it would give it a good run for its money.

Ken

[Bob Atkins]

That's an excellent philosophy if you have the self control! Those new cameras and lenses are so tempting and there's always a little voice at the back of your head that's telling you that if only you had a new camera your pictures could be so much better. Of course the important word there is "could" and not "would". As you so rightly say, unless you're such a good photographer that you are pushing the limits of your existing equipment right to the edge, an upgrade probably isn't going to do a lot for your photography.

I think it all boils down to what you shoot and how you shoot it, not what you shoot it with. Of course if everyone realized that the economy would tank and Canon and Nikon would go out of business, so don't tell anyone else the secret.

[Mike]

Not afraid that Lytro style cameras in a few years would push everything we own out of the business :-P LOL

P/S I decided that I might wait a few month... and see where the prices settle... if it still be at $3500, I will just do a gap plug and buy 7d on the cheap...

[bmpress]

I was almost ready to rob a bank (LOL) and purchase the Mark III, and then read the sensor report by DXO Labs comparing it to the new Nikon D800. Even though the high ISO performance of the MarkIII is far superior to the other, the performance of the Canon falls short in their tests for normal ISO shots.

So Bob, what does this mean? Is the Canon not up to snuff? Will this get folks switching to Nikon? Or is DXO wrong?

Thanks,
Barry

[Johnny]

DXO is not wrong. IQ is as expected. It is marginally improved. From what I have seen you can push the MIII files more without seeing banding than you could do with the MII. So it looks as they have at least some improvements compared to the MII.

IMO increasing DR at low iso`s is a matter of priority from Canon. I think Canon see the current DR at low iso`s as "sufficient". I also believe they can improve it if they wish to do so. I think it is a calculated choice.

Switching is expensive. Canon knows that too.

If I really needed two more stops of DR for landscape work I would buy myself a Nikon. Or maybe a K5. They are getting cheap now. And a couple of fast primes from Sigma.

[Bob Atkins]

The D800 looks like a very good camera indeed. If I didn't have any investment in DSLRs or lenses and I wanted to spend around $3000 on a full frame camera body at this point I'd buy into the Nikon system. It seems like their sensor readout electronics are cleaner than Canon's.

The D800 is also cheaper than the 5D3 and it has a built in flash (which I think can act as a wireless controller). Overall I'd say it's a better camera at a lower price as far as image quality goes at low ISO.

The 5D3 probably has a high ISO advantage, not least because it's capable of ISO settings two stops faster then the D800

It's yet to be fully determined how well the two AF systems operate and I don't know what video comparisons have been done.

I suspect that both the 5D3 and the D800 will deliver equally good results for 95% of the normal needs of most users. If a 5D2 was good enough, the 5D3 will be better. But if we are talking about "spec sheet" king of the hill in the $3000 DSLR region, it looks like the D800 gets the prize.

If you have a bag full of lenses and a 5D MkII, not upgrading seems like a reasonable action at the moment, unless you simply can't live with the AF system of the 5D2.

[bmpress]

It would be interesting to know which company or companies manufacture the sensor chip and the readout/amplification chip for the Canon and Nikon cameras, and where exactly is the problem with the Canon devices compared to the Nikon. If Canon's weak link is due to deficient  internal chip manufacturing capability, they could be thinking of outsourcing. I suppose that I view this latest Mark III deficiency as a major debacle for Canon since we have all become highly sophisticated users who want nothing less than the best that technology has to offer, and I feel pretty certain that executive heads may be living in fear at this moment.

[KeithB]

I have heard that Sony makes the Nikon chips and that Canon makes their own.

I think Canon is focusing (hah!) more on video.

[bmpress]

And we can be sure that Canon has been cutting one of the Nikon/Sony chips apart to understand how it has higher DR with smaller pixels. And if it is not the chip, than it must be the OpAmp module. That may be outsourced too. I believe that Burr Brown made them for several camera manufacturers a few years ago.

[Bob Atkins]

Canon make their own DSLR sensors in house. With CMOS, much of the readout circuitry is built onto the CMOS chip (that's one advantage of CMOS). CCDs are much more likely to use external amplifiers and readout electronics. I'd assume that all the circuitry that Canon use is developed and manufactured in-house.

Up to now Nikon have been using Sony sensors a lot. I don't know if they have officially acknowledged Sony as the manufacturer of the D800 sensor, but I assume is is outsourced as I don't think Nikon have in-house semiconductor/sensor fabrication facilities capable of making full frame CMOS sensors.

At this point it's not 100% clear why the D800 images seem to be cleaner in the deep shadows then the 5D3 images. There is some evidence that they are, at least at lower ISO settings. Whether this is from intrinsic pixel/sensor construction, readout electronics, amplifier noise or is due to some additional noise suppression electronics prior to digitization it's hard to say since neither Canon nor Nikon ever give design details of their sensor circuitry.

I'm following the discussions and at some point when the fog starts to clear I'll probably do a writeup on the D800 vs 5d3 situation. Right now I haven't had my hands on either of them for extensive testing but I should get a 5D3 to look at at some point. Whether I could get hold of a d800 at the same time I don't know, but I will give it a try.

If Canon do learn anything from Nikon, I don't think we'll see the results very soon in a FF DSLR. They've only just released the 1D X and 5D MkIII so there's not really any room in the lineup for something else without treading on the toes of their latest cameras. Not only that but it probably takes a minimum of 12-18 months to get a camera into production. I suppose it's possible that we might see something interesting as the next generation of the EOS 7D though. The 7D is now 2.5 years old. I have absolutely no complaints about mine and no desire to "upgrade" it, but Canon certainly must have the 7D Mk II in development by now. 3 years is about the lifespan of a DSLR model these days. Maybe something with 25MP, low noise, high ISO and high DR, priced in the $1600-$1800 region would shake things up.

[Johnny]

...we have all become highly sophisticated users who want nothing less than the best that technology has to offer...

Agree. Have you seen this http://www.fredmiranda.com/5DIII-D800/index_controlled-tests.html ? It is interesting to see the difference in the shadows. It is huge.

I am still convinced this is a matter of priority and money and I am not so sure we will be seeing something completely in a long while.

Another thing to consider is the new and advanced raw processing capabilities. Adobe has taken another step forward with the new 2012 processing engine. I have been trying it out and I am impressed. Every camera will of course benefit from this but cameras like this new Nikon will no doubt benefit even more from it.

[bmpress]

Yes...very big issue for Canon. People are now talking seriously about switching to Nikon and some, like myself, are not going to purchase a Mark 3 until Canon gets it act together.

I would not be surprised if Canon holds production of this camera until they update the sensor issue, and then refit those cameras in circulation.

As far as switching is concerned my feeling is that for non-pro's like me, it is totally out of the question. In fact Canon's deficiency is probably wonderful news because Nikon is forcing the acceleration of product development and release timing.

But it surely is a most interesting time to be into photography!

[KeithB]

Nonsense.
Canon will do nothing about the sensor "problem". Unlike Nikon, they are marketing the camera to two markets: still photography and Video. Canon will not release an "improved" model, nor retrofit any old cameras.

I don't think Canon cares about pixel-peepers or DXOMark.

[Bob Atkins]

I agree. Canon made the camera that Canon wanted to make. They made the improvements they wanted to make, mainly in the AF system. They would not recall the camera because there is nothing wrong it. It's a perfectly fine camera. a better camera then the 5D MkII is, which itself is a perfectly fine camera.

The very last thing Canon will do is say "Opps, we made a mistake, the Nikon D800 is so much better that we're going to have to recall all the 5D3s and retrofit them". Never, ever going to happen. Not in a million years.

However, Nikon seem to have set a benchmark, which is good. You can bet that the next Canon EOS to be released, which I'm guessing may be a 7D update, will be designed keeping the D800 in mind.

However I don't think Canon will ignore pixel peepers, reviews and DxOMark. I think they wiill take note of what's being said when setting the direction of their R&D efforts.

[KeithB]

It seems to me that the latest Canon offerings are really aimed at control/software improvements rather than sensor hardware.

For examples look at the almost confusing AF systems and the "silent control" :
http://www.learn.usa.canon.com/resources/articles/2012/5d3_silent_control_video.shtml?categoryId=12

(Which is really aimed at videographers.)

I wonder if Canon has some sensor improvements coming down the pike, and they wanted to test some non-sensor struff to put into a new flagship camera.

However, what we don't know is how much the C and 4K video cameras are diluting still camera and sensor development. After all, almost any of the sensors developed in the last 5 years are way overkill for the video market.