Canon EOS-1D X Mark II Image Quality Comparison

[Canon Rumors]

<p>Imaging Resource has posted their image quality comparison between the Canon EOS-1D X Mark II and cameras such as the original Canon EOS-1D X and Nikon’s latest flagship, the D5.</p>
<p>Canon EOS-1D X Mark II vs Canon EOS-1D X at ISO 1600</p>
<blockquote><p>The 1DX Mark II continues to resolve fine detail a little better than its predecessor here at ISO 1600. Luma noise appears to be a bit higher in the shadows from the Mark II, but chroma noise is better controlled which may be why the 1DX II doesn’t do as well with our tricky red-leaf swatch (the fabric has probably also faded slightly since the 1DX was shot). Still, overall image quality is similar. <a href="http://www.imaging-resource.com/PRODS/canon-1dx-ii/canon-1dx-iiA.HTM#image_quality">See the full comparison</a></p></blockquote>
<p>Most sites and reviewers are finding that the new EOS-1D X Mark II does have better image quality than its predecessor, but that the differences are small.</p>
<span id="pty_trigger"></span>

 

[biglipps66]

Yes

 

[Deleted member 7828]

Oddly enough, there have also been reports that the Mark II produces a softer overall image than the original...the DP Review shots (and a couple other independent tests) appear to support that...others (like this one) seem to say otherwise (contradict). Consensus...?

 

[Sporgon]

Oddly enough, there have also been reports that the Mark II produces a softer overall image than the original...the DP Review shots (and a couple other independent tests) appear to support that...others (like this one) seem to say otherwise (contradict). Consensus...?

Could be slightly larger output size.

 

[Loswr]

Oddly enough, there have also been reports that the Mark II produces a softer overall image than the original...the DP Review shots (and a couple other independent tests) appear to support that...others (like this one) seem to say otherwise (contradict). Consensus...?

DPR is biased, as far as I can tell, IR is not. I'd take anything DPR states with a grain of salt. Rather more than a grain, actually.

For example, regarding the 1D X II, DPR concludes:

[quote author=Rishi 'I'm not biased' Sanyal]
While the EOS-1D X II shows big improvements in base ISO dynamic range relative to previous Canons, high ISO performance remains stagnant, falling behind the Nikon D5, and showing no benefit over the higher resolution Sony a7R II at common viewing size.
[/quote]

DxO's measurements show that the 1D X II is has higher SNR at high ISOs than the the D5, and the same SNR as the a7RII. You can also see their bias in the phrasing: "...showing no benefit over the higher resolution Sony a7R II at common viewing size," or put differently, when you factor in the benefit of downsampling the a7RII by >50%, it performs as well as the 1D X II.

 

[ritholtz]

Oddly enough, there have also been reports that the Mark II produces a softer overall image than the original...the DP Review shots (and a couple other independent tests) appear to support that...others (like this one) seem to say otherwise (contradict). Consensus...?

DPR is biased, as far as I can tell, IR is not. I'd take anything DPR states with a grain of salt. Rather more than a grain, actually.

For example, regarding the 1D X II, DPR concludes:

[quote author=Rishi 'I'm not biased' Sanyal]
While the EOS-1D X II shows big improvements in base ISO dynamic range relative to previous Canons, high ISO performance remains stagnant, falling behind the Nikon D5, and showing no benefit over the higher resolution Sony a7R II at common viewing size.

DxO's measurements show that the 1D X II is has higher SNR at high ISOs than the the D5, and the same SNR as the a7RII. You can also see their bias in the phrasing: "...showing no benefit over the higher resolution Sony a7R II at common viewing size," or put differently, when you factor in the benefit of downsampling the a7RII by >50%, it performs as well as the 1D X II.
[/quote]
Yah. They went great lengths to explain issues with ETTR and importance of iso invariance sensor for 7d2. Then for D5, they said its DR is fine unless you are a hardcore landscape shooter.

 

[Ozarker]

I'd still be happy with a 1Dx.

 

[Deleted member 7828]

Back in the "good old days" DPReview was an excellent resource and very fair in their reviews, I thought.

I think nowadays they focus a bit too much on dynamic range and shadow recovery abilities of sensors (and weight their reviews accordingly). Sure, it's nice to have more DR/editing latitude, but it's not essential to all. Plus, it seems to come at the cost of high ISO IQ. I think that's a point I don't see mentioned very much in the discussion.
_____________

Having said all that, I wasn't referring to what DPR SAID in their review of the camera - instead, it comes from the studio image samples they offered - and I do acknowledge that the difference is extremely subtle and requires a sharp eye to spot - but I insist it is there - their 1DX II studio test image appears just a touch softer than the test image for the 1DX (and 5D3, for that matter). Most noticeable at base ISO (and with the deck of cards). The same lens (85 1.8 ) appears to have been used...so unless there is a testing inconsistency I'm unaware of...

http://www.dpreview.com/news/8090146652/canon-eos-1d-x-mark-ii-studio-tests/2

Colors are also different - seems to be more neutral, less "punchy". That's also apparent in the IR comparison.

 

[scottkinfw]

Back in the "good old days" DPReview was an excellent resource and very fair in their reviews, I thought.

I think nowadays they focus a bit too much on dynamic range and shadow recovery abilities of sensors (and weight their reviews accordingly). Sure, it's nice to have more DR/editing latitude, but it's not essential to all. Plus, it seems to come at the cost of high ISO IQ. I think that's a point I don't see mentioned very much in the discussion.
_____________

Having said all that, I wasn't referring to what DPR SAID in their review of the camera - instead, it comes from the studio image samples they offered - and I do acknowledge that the difference is extremely subtle and requires a sharp eye to spot - but I insist it is there - their 1DX II studio test image appears just a touch softer than the test image for the 1DX (and 5D3, for that matter). Most noticeable at base ISO (and with the deck of cards). The same lens (85 1.8 ) appears to have been used...so unless there is a testing inconsistency I'm unaware of...

http://www.dpreview.com/news/8090146652/canon-eos-1d-x-mark-ii-studio-tests/2

Colors are also different - seems to be more neutral, less "punchy". That's also apparent in the IR comparison.

To my old eyes, the 1DX DPReview images look better or about the same in some cases vs. the 1DXII. So my question is what do we make of this? Shouldn't sensors in new cameras be way greater? Aside faster fps, why upgrade to the v. II? Slight reported DR? Will sensor be improved with a firmware update? Any ideas/suggestions?

Thanks.

sek

 

[retroreflection]

Shouldn't sensors in new cameras be way greater?
[/quote]
No.
Expecting to see an obvious difference between an 18 and 20MP sensor built to an essentially common design is asking too much. Focus accuracy and observer bias can easily swamp the tiny differences that might exist. Then, if you want to talk about color accuracy, that test shot methodology is not designed to assess color over several years. Mutable color of the objects, and probable variation in the lighting would have to be controlled.

Fundamentally, we should start to accept the fact that we are in a time of diminishing pace of sensor advances. I will not discount the opportunity for Canon's sensors to improve relative to noise floor performance. I will say that the vast majority of photographers live in such light polluted areas that astrophotography will remain a niche pursuit. The rest of us have a selection of sensors from many manufacturers that can deliver stunning images, within a few constraints. We should accept those constraints and shoot to them. If you look at each picture and attach blame for lost potential because those constraints exist, you will never be happy.

Why should there be new models?
1. Other aspects of the camera system have not reached the same level of technological saturation. Autofocus systems, control ergonomics, video capabilities, data handling... all are worthy of continued development. (Should you pay for the new ones on each new model? Business and consumerism decisions like that are personal choices.)
2. Consumers continue to expect a development cycle that gives Santa something to do each and every year. It wasn't always that way in photography. It will not always be that way.

 

[3kramd5]

Fundamentally, we should start to accept the fact that we are in a time of diminishing pace of sensor advances.

Perhaps for single-layered bayer CMOS sensor, but certainly there are large advances to be made with different approaches (stacked sensors, sCMOS, rolling buffers aka modulo, etc).

 

[scottkinfw]

Shouldn't sensors in new cameras be way greater?

No.
Expecting to see an obvious difference between an 18 and 20MP sensor built to an essentially common design is asking too much. Focus accuracy and observer bias can easily swamp the tiny differences that might exist. Then, if you want to talk about color accuracy, that test shot methodology is not designed to assess color over several years. Mutable color of the objects, and probable variation in the lighting would have to be controlled.

Fundamentally, we should start to accept the fact that we are in a time of diminishing pace of sensor advances. I will not discount the opportunity for Canon's sensors to improve relative to noise floor performance. I will say that the vast majority of photographers live in such light polluted areas that astrophotography will remain a niche pursuit. The rest of us have a selection of sensors from many manufacturers that can deliver stunning images, within a few constraints. We should accept those constraints and shoot to them. If you look at each picture and attach blame for lost potential because those constraints exist, you will never be happy.

Why should there be new models?
1. Other aspects of the camera system have not reached the same level of technological saturation. Autofocus systems, control ergonomics, video capabilities, data handling... all are worthy of continued development. (Should you pay for the new ones on each new model? Business and consumerism decisions like that are personal choices.)
2. Consumers continue to expect a development cycle that gives Santa something to do each and every year. It wasn't always that way in photography. It will not always be that way.
[/quote]

Thank you Retroreflecction for your insights. I purchased the 1DXII coming from a 5DIII, and obviously the camera is a huge upgrade for me on many levels. When I saw the reports and pic comparison, I got "buyer's remorse" despite the fact that the camera is great. Human nature I guess.

sek

 

[arthurbikemad]

Sek, why the remorse, do you feel you have lost IQ? Most images I took with the 1DX Mk1 stood up with my 5D3 easy, I want to get a 1DX Mk2 but am hanging on for the spec of the 5D Mk4, trying to avoid any remorse IQ wise, problem is I want linked meter modes and the other benefits of the pro body, but as a pixel peeper I am afraid I will be dissapointed with the Mk2.

 

[arcer]

I'm no expert but.......

Is the softness due to the splitting of the photodiodes due to the introduction of DPAF into the 1DX2? As commented by DPR.

If that is the case, would we be seeing similar softening in the 5D4 if it will have a DPAF sensor also?

 

[privatebydesign]

Having said all that, I wasn't referring to what DPR SAID in their review of the camera - instead, it comes from the studio image samples they offered - and I do acknowledge that the difference is extremely subtle and requires a sharp eye to spot - but I insist it is there - their 1DX II studio test image appears just a touch softer than the test image for the 1DX (and 5D3, for that matter). Most noticeable at base ISO (and with the deck of cards). The same lens (85 1.8 ) appears to have been used...so unless there is a testing inconsistency I'm unaware of...

http://www.dpreview.com/news/8090146652/canon-eos-1d-x-mark-ii-studio-tests/2

Colors are also different - seems to be more neutral, less "punchy". That's also apparent in the IR comparison.

Your link is to JPEG OOC, all you are comparing is different rendering intents.

If you go to the RAW comparison http://www.dpreview.com/news/8090146652/canon-eos-1d-x-mark-ii-studio-tests/2 then you have a slightly more meaningful tool.

But what you are seeing is not normalised or colour calibrated. The more pixels the bigger the magnification, so you are effectively looking at a bigger print/crop from the 1DX MkII, so if they look the same the bigger image is 'better'. As for the colour, you can't get around the fact that it is entirely made up by the software, use another program or make a profile and the colours can be whatever you want, even OOC Jpegs if you take ten minutes to sort out what you personally want your jpegs to look like.

But, looking at unedited RAW files is not really very helpful either, who doesn't edit the RAW file? You are seeing the effect of AA filters etc not the practical output quality of the file.

Truthfully the comparison tool is very limited in its use. Far and away the best thing to do is download the RAW files from the links and put them into your own editing software and have a play with them. Given that they contain a standard colour checker you can even make corrected colour profiles for them.

 

[Jack Douglas]

privatebydesign, good points. It all becomes somewhat subjective even when one tries to be objective. It seems to me that there isn't too much point in "which is best" when products are that close to one another. Like my 300 2.8 II I finally said to myself I may not have the sharpest copy but a lot of folk are loving what I'm doing with it and so do I, and that solved my issues. I just use it and love it.

I'm improving because I'm trying to focus on various things that are not gear related and I now appreciate what features are important to me, like the lit focus points, so I'll be happy with a 1DX II. The low FPS of a 5DSR is what I don't like about my 6D so that's out, given I need faster and I doubt that the 5DIV will be to my liking after having owned the 1D4 but if it's out by September, I'll maybe consider it.

One thing I've learned, second guessing after a purchase just leads to frustration and wasted time/energy and is best avoided. That's why I've held off and kept investigating before my proposed September purchase date.

Jack

 

[Deleted member 7828]

PBD - that's exactly what I did...viewed the RAW comparison, downloaded the files, played with them and came to a conclusion. I've also downloaded RAW files from other sites. I stand by my observation.

On average it took about 0.5 extra sharpening in DPP for the 1DXII to match the others in sharpness. Even then, I still found certain fine details remained "smoothed out". Again, to most people this is splitting hairs, or "pixel-peeping" or whatever you want to call it - and this doesn't mean the camera is bad (I'm still in the market for one should the push arise)...but it is something I've noted and will be keeping a close eye on with the upcoming 5D4. My guess is a stronger AA filter but I don't know.

This is mainly at lower ISO settings - once you get to ISO 3200 and higher, dialing down the noise reduction a couple notches below the default on the 1DXII image more or less appears to equalize things.