Canon EOS-1D X Mark II Studio Tests

[rishi_sanyal]

This discussion has gotten boring, redundant and reeks of tin-foil hat conspiracy theories.

Interestingly not one of the critics has provided any credible evidence that DPReview's findings are wrong. Instead it is all about little quibbles with the way reviews are written, with the conclusion that the reviews are not glowing enough to satisfy some Canonites.

It's time for everyone to just move on. DPReview is a privately owned site and they are entitled to test cameras in the way they feel is most productive. If you disagree, well...no one makes you go to their site. And, if you think you can design tests and report results in a more objective manner...well knock yourself out. No one is stopping you.

The only question I have is how much longer I will have to wait for the full 1dx review.

I have.

I have linked and posted the image Rishi took in the 5DS/R review that says that the camera couldn't take that image without showing banding and noise in the shadows. That is a lie.

I (and Sporgon) have also posted similar setup images that demonstrate it is a lie.

No, actually, you haven't. None of your images have anything to do with the conditions of my/our images.

And therein lies one of the biggest issues with your type of arguments: you think your own example represents the sum total of all types of conditions a photographer may encounter.

Whereas our example represents one situation a photographer may encounter where the competition would just do better.

 

[rishi_sanyal]

The latest releases from Canon must in some way be threatening to Sony/Nikon, since they find it necessary to send one of their paid evangelists into a Canon forum... defending commercial interest and even accusing people here for being "racists" ???

So when someone accuses our verbiage as possible being unknowingly biased, my coming here (a site I visit all the time personally, having been a Canon shooter for 15 years) to point out the ESL implication is wrong (and offensive) is 'defending commercial interest' and 'paid evangelism'?

Is that what you're saying?

 

[Jack Douglas]

This discussion has gotten boring, redundant and reeks of tin-foil hat conspiracy theories.

Interestingly not one of the critics has provided any credible evidence that DPReview's findings are wrong. Instead it is all about little quibbles with the way reviews are written, with the conclusion that the reviews are not glowing enough to satisfy some Canonites.

It's time for everyone to just move on. DPReview is a privately owned site and they are entitled to test cameras in the way they feel is most productive. If you disagree, well...no one makes you go to their site. And, if you think you can design tests and report results in a more objective manner...well knock yourself out. No one is stopping you.

The only question I have is how much longer I will have to wait for the full 1dx review.

I agree, it's time to get off this hobby horse. Bias or no bias, I believe comments that cause one to question or rethink; comments that cause that uncomfortable feeling, like is this purchase actually wise, ultimately are for our own good.

Jack

Jack

 

[traveller]

Rishi has violated one of the golden rules of any forum:

"Never argue with an idiot, they'll drag you down to their level and beat you with experience."

 

[rishi_sanyal]

We're trying our best to turn around reviews in a timely, accelerated manner this year. There will definitely be a 1D-X II review, and it's largely centered around AF comparisons to D5. Not comparisons against the a7R II, I assure you.

Glad to see the sentiment about 'causing that uncomfortable feeling.' Absolutely agree.

Cheers,
Rishi

 

[Jack Douglas]

The latest releases from Canon must in some way be threatening to Sony/Nikon, since they find it necessary to send one of their paid evangelists into a Canon forum... defending commercial interest and even accusing people here for being "racists" ???

So when someone accuses our verbiage as possible being unknowingly biased, my coming here (a site I visit all the time personally, having been a Canon shooter for 15 years) to point out the ESL implication is wrong (and offensive) is 'defending commercial interest' and 'paid evangelism'?

Is that what you're saying?

Rishi, you have enough intelligence to not be using the race card when it is unjustified. However, I can understand how my statement might touch a nerve and probably I should have worded it a little differently. By responding the way you did and again mentioning ESL here, you display just a little insecurity. That's disappointing.

Here is my quote: "Not everyone has an excellent command of the English language for whatever reason. There are many, including ESL, limited education, disinterest in language as opposed to technology and so forth."

You chose to take the ESL reference personally. Myself, if there was a reference from me relative to you specifically, it would have been that you are strongly focused on technical issues and are less interested in the nuances of the English language. That would just be my guess, of course.

As an engineer I was often chided as being part of a group that did not have strong language skills. The reason I mentioned ESL is because of my personal interactions with my wife that tell me that after 30 years she still doesn't fully understand how a slight change in wording can alter the meaning of a sentence in english.

If your language capabilities are indeed excellent then I have to revert to the camp that believes you are sometimes purposely presenting a bias. What does it matter what I think? Just do your best.

Jack

 

[tpatana]

Half Americans speak worse English than half ESLs I know

 

[rishi_sanyal]

None of any of this has anything to do with the English language, ESL or technical level of interest or what have you. If you can't appreciate that perceptions of bias may be nothing other than misperceived perceptions of bias, then I'm afraid I am also duly disappointed.

Rishi

 

[Jack Douglas]

No more from me on this topic. I should have refrained in the first place. I have encouraged people with poor english to post and couldn't care less if someone is handicapped in that respect.

I simply came with the suggestion that not everyone is capable of perceiving how bias exhibits in english statements - why some recognize bias while others don't. This is a challenging topic because it is so subjective and of course we Canonites tend to be defensive.

Apologies for having contributed to this diversion!

Jack

 

[rishi_sanyal]

No more from me on this topic. I should have refrained in the first place. I have encouraged people with poor english to post and couldn't care less if someone is handicapped in that respect.

I simply came with the suggestion that not everyone is capable of perceiving how bias exhibits in english statements - why some recognize bias while others don't. This is a challenging topic because it is so subjective and of course we Canonites tend to be defensive.

Apologies for having contributed to this diversion!

Jack

That's fine, I understand and appreciate the explanation. I also did not intend for it to be blown out of proportion. Just to raise the awareness of the irony of implicating anything but the possibility of misperceived bias when reading data that doesn't fit your own bias.

But I think each of our intentions are somewhat clear now.

Cheers,
Rishi

 

[Stu_bert]

Hi Rishi

Does Dpreview intend to provide similar studio tests for earlier cameras? It would help people that don't always buy the latest generation (second hand for instance), and also see how much gap there is say between the 1dx I I and a d4s, or a7r and mk ii ?

I'd also like to thank you for discussing the review here - you didn't have to, but I think open debate is good and appreciate the effort you make.

 

[rishi_sanyal]

Hi Rishi

Does Dpreview intend to provide similar studio tests for earlier cameras? It would help people that don't always buy the latest generation (second hand for instance), and also see how much gap there is say between the 1dx I I and a d4s, or a7r and mk ii ?

I'd also like to thank you for discussing the review here - you didn't have to, but I think open debate is good and appreciate the effort you make.

Thanks. All those desired comparisons you'd like to make are already available in our widget. Do you perhaps need a more thorough Instruction sheet for how to effectively use our tools? This is something we're working on by the way.

 

[9VIII]

Because look at Neutral's crops here: the a7R II does have a very slight advantage (less blotchy noise) in the grey patch, and meanwhile the author conflates aliasing artifacts (moire) with 'poor rendering of low contrast areas'. No... That's moire due to the high res sensor and lack of AA filter.

Yes, Neutral's samples are specifically looking at a spot with some very high frequency data, none of the cameras render that well at high ISO.
If you look at other things, like the four main photographs of people's faces, it's obvious that the A7RII is crushing tons of detail. The D5 generally looks best but the original 1DX still pulls ahead in a few instances. Maybe that could be attributed to the random nature of high ISO noise but it's clear the two are performing very similarly, and much better than the A7RII.

Nonsense. Rishi unbiasedly picked a small patch where the SNR of the a7RII is 1/2-stop better. :

Your comment appears to indicate a fundamental misunderstanding of how SNR analyses work. Biggest differences in ISO performance of similar-sized sensors tend to lie in midtones and darker tones, not bright tones where SNR is pretty much only determined by sensor size and efficiency, and where noise isn't as offensive to begin with.

You also can only analyze SNR for detail-less grey patches, but it's not clear you even know that.

So, I analyzed a patch one patch darker than 18% in my analysis, but if you want a proper analysis of a 'midtone' (18%), here it is:

Canon 1D-X II: 4.36
Sony a7R II: 4.97

... or a 14% SNR advantage for the a7R II, which is about 1/3 EV. As you go to brighter tones, the SNR advantage will go to 0, because it's all dependent on sensor size mostly.

Of course, I can actually admit that, because I don't have a bias to start with.

The interesting thing is that the A7RII still has a distinct disadvantage in actually displaying detail in people's faces at high ISO. SNR doesn't explain everything that's going on.

Although the 1D-X II shows significant increase in dynamic range at low ISOs in our dynamic range tests, high ISO Raw performance remains fairly similar to its predecessor, which is actually impressive considering the 1D-X II gains dual-pixel architecture for decisive video AF. Noise performance falls slightly behind the Nikon D5 (and even the Sony a7R II when normalized) at very high ISOs. At these very high ISOs, JPEGs suffer a bit as well, with the a7R II showing the most detail retention in grey tones and in low contrast greenery, despite all cameras starting off with similar detail in Raw. Sony's clever sharpening and context-sensitive noise reduction help it establish its lead, but the Nikon D5 isn't too far behind. Canon's noise reduction is, in comparison, less aggressive overall, but smudges away low contrast detail. While on the surface this may not seem an ideal combination, it's a fair choice in the sense that it avoids obvious noise reduction artifacts.

I agree that the 1DX2 is about the same as the 1DX in collecting detail at high ISO, and that the JPEG's off the 1DX2 for whatever reason look smoothed over, and that the D5 has the best overall noise performance (though it still misses some details that the original 1DX catches), and maybe the A7RII does do best in grey and green, but on the brown tones on people's faces it loses a lot of details.

The A7RII is doing a very nice job given the inherent disadvantage, but at ISO 52100 the girl on the top right practically loses her entire bottom lip, the D5 does an excellent job maintaining that contrast, and the 1DX and 1DX2 are somewhere in the middle.
The praise for Sony in the paragraph I quoted is oddly specific: "grey tones and in low contrast greenery" and then broad conlusions are drawn from those oddly specific examples: "Sony's clever sharpening and context-sensitive noise reduction help it establish its lead, but the Nikon D5 isn't too far behind."
You can find places where the A7RII pulls more detail, but does that "establish its lead" when it obviously fails in so many other instances? This is a race with no clear winner.

 

[traveller]

Hi Rishi

Does Dpreview intend to provide similar studio tests for earlier cameras? It would help people that don't always buy the latest generation (second hand for instance), and also see how much gap there is say between the 1dx I I and a d4s, or a7r and mk ii ?

I'd also like to thank you for discussing the review here - you didn't have to, but I think open debate is good and appreciate the effort you make.

Thanks. All those desired comparisons you'd like to make are already available in our widget. Do you perhaps need a more thorough Instruction sheet for how to effectively use our tools? This is something we're working on by the way.

I don’t understand the fuss that some are making about the choice of cameras for the DPR comparison. If you disagree with the default cameras, it only takes a few seconds to change the choices.

I think the widget works well and is easy to understand. Some people just like to complain.

 

[bwud]

We're trying our best to turn around reviews in a timely, accelerated manner this year. There will definitely be a 1D-X II review, and it's largely centered around AF comparisons to D5.

Rishi -

Thanks for coming to these forums. May I ask what measures you'll take to ensure that the cameras are set appropriately for the scenario you'll test?

Also, I myself am very interested in performance with extended lenses (f/8 max aperture). Any chance you'll compare those capabilities?

 

[East Wind Photography]

No more from me on this topic. I should have refrained in the first place. I have encouraged people with poor english to post and couldn't care less if someone is handicapped in that respect.

I simply came with the suggestion that not everyone is capable of perceiving how bias exhibits in english statements - why some recognize bias while others don't. This is a challenging topic because it is so subjective and of course we Canonites tend to be defensive.

Apologies for having contributed to this diversion!

Jack

That's fine, I understand and appreciate the explanation. I also did not intend for it to be blown out of proportion. Just to raise the awareness of the irony of implicating anything but the possibility of misperceived bias when reading data that doesn't fit your own bias.

But I think each of our intentions are somewhat clear now.

Cheers,
Rishi

Earlier on I menioned that i found the comparison between models pretty useless. What was more important to me were the actual 1dx2 tests. As a professional i am more concerned with the differences between models within the same manufacturer than how a model stacks up to something i would never consider. I dont have enough disposable income to switch everything from one vendor to another. It might make sense to someone purchasing a high end camera for the first time but i think that kind of decision would be more about the system than just sensor performance.

The tests that you posted in the article were helpful in making my decision to buy the new 1dx but it had little to do with Sony and Nikon performance. I just wanted to share that with you and hope in the future you all will focus more on how canon improved things over previous models or the comparisons between new models of the same vendor.

Thanks for taking the time to talk to us even in these angry forums full of negativity. People should try to be more constructive and less antagonistic.

 

[neuroanatomist]

Nonsense. Rishi unbiasedly picked a small patch where the SNR of the a7RII is 1/2-stop better. :

Your comment appears to indicate a fundamental misunderstanding of how SNR analyses work. Biggest differences in ISO performance of similar-sized sensors tend to lie in midtones and darker tones, not bright tones where SNR is pretty much only determined by sensor size and efficiency, and where noise isn't as offensive to begin with.

You also can only analyze SNR for detail-less grey patches, but it's not clear you even know that.

Nor is it clear that I've written Matlab scripts to analyze SNR and other image parameters while developing methods to quantify immunofluorescence, and yet I have.

Of course, I can actually admit that, because I don't have a bias to start with.

Oh, ok. Your writing shows otherwise, but you tell yourself whatever you need to, hope it helps.

Maybe it's an aspirational goal, if so, great! You could write a short piece for your site called, "DPReviewers: Making Small Steps toward Overcoming Their Bias."

 

[scyrene]

Nonsense. Rishi unbiasedly picked a small patch where the SNR of the a7RII is 1/2-stop better. :

Your comment appears to indicate a fundamental misunderstanding of how SNR analyses work. Biggest differences in ISO performance of similar-sized sensors tend to lie in midtones and darker tones, not bright tones where SNR is pretty much only determined by sensor size and efficiency, and where noise isn't as offensive to begin with.

You also can only analyze SNR for detail-less grey patches, but it's not clear you even know that.

Nor is it clear that I've written Matlab scripts to analyze SNR and other image parameters while developing methods to quantify immunofluorescence, and yet I have.

Of course, I can actually admit that, because I don't have a bias to start with.

Oh, ok. Your writing shows otherwise, but you tell yourself whatever you need to, hope it helps.

Maybe it's an aspirational goal, if so, great! You could write a short piece for your site called, "DPReviewers: Making Small Steps toward Overcoming Their Bias."

Everyone is biased. The sooner we accept it, the sooner we can try to address it. Attempting to reduce or neutralise our bias is a worthy goal. Better not to say "I'm not biased" than, "I try not to be biased".

As for organisations, they ought to try and have a range of views, so biases cancel each other out. I think there is a genuine concern (and it isn't just sour grapes) that there's institutionalised bias in some review sites. DPR provides a lot of useful and interesting output, but it isn't perfect, and hopefully they will take constructive criticism on board. It isn't a big deal in the scheme of things, and it's a shame the discussion has become so angry in parts, but then every discussion seems to do that.

 

[unfocused]

It isn't a big deal in the scheme of things, and it's a shame the discussion has become so angry in parts, but then every discussion seems to do that.

+1000

Honestly, my respect for several forum members has really dropped with this discussion.

Let me summarize how it looks to me:

DPR: The new Canon 1DX II appears to have eight angels dancing on its sensor. That is two more than the 1DX I had, but still slightly below the nine angels dancing on the Sony sensors.

Canon Rumors Forum: Nonsense! Anyone can see that Sony sensors only have seven angels dancing on them. You are hopelessly biased to claim nine angels and therefore you are corrupt and on the take. Besides which, you have a funny name!

Me: I don't give a sh*t. Just complete the review and tell me how the camera performs.

 

[neuroanatomist]

Let me summarize how it looks to me:

DPR: The new Canon 1DX II appears to have eight angels dancing on its sensor. That is two more than the 1DX I had, but still slightly below the nine angels dancing on the Sony sensors.

DPR: The new Canon 1DX II appears to have more low ISO DR than the new Nikon D5, but that isn't really all that important in this class of camera. High ISO matters more in this class (although that wasn't true for the 1D X vs. D4). Regardless, the old Sony a7RII outperforms both of its peers in this class. In other news, we've decided – and we've already told you even before the 1D X II started shipping – that the D5's class-leading AF performance is superior to the 1D X II. Now, we're going to go out and test them, because there's always the angels-dancing-on-a-sensor chance that the answer to our question, "Canon Catching Up?," could maybe possibly not be no.

CR Forums: :