Canon 7Dii vs Nikon D750 Dynamic Range Test

[privatebydesign]

Specifically on the testing I'd say you have a couple of issues.

The first, from your strips the Canon's are displaying more DR than the Nikon, why do I say that? Because the black is darker, DR is the difference in luminosity between the brightest white and the darkest black, in those simple terms the Canon strips perform better. Now I am not suggesting the Canon has 'more DR', what I am saying is there is a flaw in your methodology. Both black points should display the same value, 0.

Second, for the 7D MkII to actually test better than the 5D MkIII also seems to point to an error in methodology. We all know the downsampling meme, without knowing sensor area sampled and reproduction ratios we don't know if we are comparing like for like or turnips to tulips.

Third, the wedge numbers seem to indicate you are not getting even illumination across your step wedge. Surely all the numbers should have the same values? You are getting falloff across the wedge. The illumination has to be completely even.

Lastly, whilst these tests do have some merit, I find a far more beneficial test is actual post processed results, that is, don't use zero post processing because it isn't 'fair' and isn't how we actually get the best out of our gear. Besides, Nikon have long been suspected of cooking their RAW files. A truer test of actual end results is to process both files to the best abilities you can with the software you would use, and dn't be surprised if other people get different results with different software.

 

[YuengLinger]

Why not convert with Canon DPP? It shipped with your 7DII.

 

[AlanF]

I was happy to get my Canon 7Dii yesterday and being in the middle of reviewing the Nikon D750, I decided to put them both through a Dynamic Range test I have been working on. I know it is an Apples to Oranges comparison, but I was curious none-the-less.

Im not a huge fan of, nor believed a certain testing website for various reasons, and have always read their scores with skepticism because their methodology is mostly proprietary and scoring system whacky. In fact, I have long wanted to come up with a DR test simply to confirm or deny their DR tests to some degree.

I am long time (12 years Canon owner and shooter vs 3.5 years of Nikon), so if anything I would naturally be more biased towards Canon, but in this test I think we are looking at a DR for the 7Dii very similar to that of the Canon 5Diii.

In the past, some of you have been very insightful with suggestions and tips to improve my tests, and therefore I invite your scrutiny and suggestions. Many others of you have ridiculed me, which also actually improved the quality of my tests, so I guess I invite that as well, as long as there is some creativity involved in the ridiculing.

What do you guys think I can do to improve this beyond the obvious (cropping straight).

What would be a good way to figure out total DR and noise cut offs?

http://www.michaelthemaven.com/?postID=3431&canon-7dii-vs-nikon-d750-dynamic-range-test-by-michael-the-maven

In any event enjoy!

Does calling yourself "The Maven" mean that you consider yourself the expert?

 

[MichaelTheMaven]

The 5Diii test is there as well, thats full frame to full frame.

Same strip, same exact settings on both cameras same everything. I have a hard time differentiating between stops 32/33 and higher on the Canons due to their noise. The D750 is much cleaner and easier to differentiate, more visible, distinct shades in deeper tones. Its clear to me. I try to frame up the strip edge to edge in both cameras. when I shoot it so yes the focal lengths are different.

The strip is calibrated plus the strobe & tripod do not move, not sure how that would explain such different results between the two.

There was some exposure and shadow recovery on both, no noise reduction. Ill have to play with it more once the 7Dii files are supported by Camera RAW.

Canon DPP will only convert to tiff and JPEG files, unless I am mistaken.

Thank you for the feedback.

M

 

[MichaelTheMaven]

A Maven is an both an expert in a particular field, AND someone who is willing to share his knowledge about that field, I first learned about it from the book "Tipping Point" by Malcolm Gladwell, and it has stuck for me.

I consider myself a student of all things who is willing to learn as much as possible and then share. Whether someone considers me an expert depends on their personal perspective. I teach beginning photography, so from the students perspective in terms of learning photography, I better be an expert...right?

M

 

[East Wind Photography]

The 5Diii test is there as well, thats full frame to full frame.

Same strip, same exact settings on both cameras same everything. I have a hard time differentiating between stops 32/33 and higher on the Canons due to their noise. The D750 is much cleaner and easier to differentiate, more visible, distinct shades in deeper tones. Its clear to me. I try to frame up the strip edge to edge in both cameras. when I shoot it so yes the focal lengths are different.

The strip is calibrated plus the strobe & tripod do not move, not sure how that would explain such different results between the two.

There was some exposure and shadow recovery on both, no noise reduction. Ill have to play with it more once the 7Dii files are supported by Camera RAW.

Canon DPP will only convert to tiff and JPEG files, unless I am mistaken.

Thank you for the feedback.

M

As I mentioned before you are not testing the complete DR of the sensors. You also need to show the white highlight area. Showing the shadow part of the DR is not the complete range. I'd like to see how the other end looks at the same exposure and where the highlight detail is lost.

The only thing you care comparing here is shadow detail at ISO 100, not dynamic range.

 

[jrista]

Thanks for the test, Michael. I'm glad you ran the test.


Given that you lifted the shadows, I think that can explain why they look brighter on the D750...there IS a lot more dynamic range to be had there. In my experience, lifting Canon shadows looks exactly like your Canon results, where as lifting A7r shadows look exactly like your D750 results.


I do think that showing the entire wedge, from the highlights to the shadows, both unprocessed as well as processed, and including the settings you used to process them, would help provide a more balanced review. It should head off the most predictable of human responses (which is quite prevalent on these forums) as well. ;P


Anyway, looking forward to seeing the full step wedge in both OOC and processed versions.

 

[Mt Spokane Photography]

Now the most important question for me. How do I stop the website from telling me this thread exists?

You can go to your profile and choose the ignore boards option. Now, if there were only a board for DR and for trolling posts (2 boards). Then we could ignore them and those interested could use them.

 

[ajfotofilmagem]

This is not a test of DR. The way it was done is a test "lift the shadows." :

A true test of DR will compare similar cameras, photographed with equal illumination at ISO 100, 200, 400, 800, 1600, 3200, 6400. 8) I know many photographers who never use ISO 100, but interestingly, many of the DR testing has only ISO 100, deliberately under-exposed, and corrected in PP.

Shooting with the correct exposure is so difficult today?
Is it really necessary to do tests that simulate user error? :-X

 

[MichaelTheMaven]

Thanks for the test, Michael. I'm glad you ran the test.


Given that you lifted the shadows, I think that can explain why they look brighter on the D750...there IS a lot more dynamic range to be had there. In my experience, lifting Canon shadows looks exactly like your Canon results, where as lifting A7r shadows look exactly like your D750 results.


I do think that showing the entire wedge, from the highlights to the shadows, both unprocessed as well as processed, and including the settings you used to process them, would help provide a more balanced review. It should head off the most predictable of human responses (which is quite prevalent on these forums) as well. ;P


Anyway, looking forward to seeing the full step wedge in both OOC and processed versions.

I see what you are saying here. Its black and white to me, but without showing what I am actually doing and how I am trying to get these 41 calibrated steps all in one shot, I can see how it would be confusing. Im out shooting now, Ill post something tonight to show how I am clipping off that first step in RAW, and then squeezing everything I can out of the shadows.

I know some people believe this is not useful or not a DR test, but I think the results are pretty obvious. If you have a better solution or idea, I would still love to hear it.

M

 

[takesome1]

It would be nice to see an overall test of the DR of the 7D II vs whatever. This test isn't that, it is just a shadow test in a narrow set of parameters.

I bought the 7D II as a wildlife camera for focal length limited situations and for packing.
Testing the 7D II at 100 ISO is useless for the the type of photography that this body is designed toward.

So I will watch for the next test to come along to see if it will show how the camera performs in real world situations.

Or maybe I will just check it out myself once mine arrives.

 

[jrista]

Thanks for the test, Michael. I'm glad you ran the test.


Given that you lifted the shadows, I think that can explain why they look brighter on the D750...there IS a lot more dynamic range to be had there. In my experience, lifting Canon shadows looks exactly like your Canon results, where as lifting A7r shadows look exactly like your D750 results.


I do think that showing the entire wedge, from the highlights to the shadows, both unprocessed as well as processed, and including the settings you used to process them, would help provide a more balanced review. It should head off the most predictable of human responses (which is quite prevalent on these forums) as well. ;P


Anyway, looking forward to seeing the full step wedge in both OOC and processed versions.

I see what you are saying here. Its black and white to me, but without showing what I am actually doing and how I am trying to get these 41 calibrated steps all in one shot, I can see how it would be confusing. Im out shooting now, Ill post something tonight to show how I am clipping off that first step in RAW, and then squeezing everything I can out of the shadows.

I know some people believe this is not useful or not a DR test, but I think the results are pretty obvious. If you have a better solution or idea, I would still love to hear it.

M

I'm someone who believes that the D750 has significantly better DR than any existing Canon sensor. I'm right there with you. I've been waiting for Canon to fix their sensor read noise problems for years, and I'm pretty fed up with their lack of progress. They may have reduced banding, that's good...but your test indicates that overall...nothing has really changed. Color noise is just as big a problem on the 7D II as it has always been (compared to the 5D III, you don't even need the D750 image to see that.) Personally, for all the time Canon invested in the 7D II, I find that extremely disappointing. It's disheartening, but, that's just my opinion. (I can only hope that sometime next year, Canon demonstrates some radical new sensor technology that improves IQ across the board, reduces read noise to unprecedented levels, etc. (I say unprecedented, because with Canon's turnaround rate for releasing new models, any new sensor technology they release will need to stand the test of at least three years...and all their competitor sensors that come out within that time.)


I just know that the resistance to and downright denial of the notion that improved DR matters, or that anyone is even remotely capable of performing a valid IQ test, or that Canon cameras might have a problem with anything, is extremely high here. I think your test is valid. It should may DTaylor happy, however I'm pretty sure if/when he comes along, he'll have some way of 'debunking' what your test clearly shows.


Your test may need fine tuning to get the best apples-to-apples comparison, and without a doubt more information will help keep the skeptics satisfied. I would show the full length of the test strip, show exactly what settings you are using in RawThearapy for your processing, and even share the RAW files themselves. That's the only thing you can do if you really want anyone to take you seriously. (And even if you provide everything, there are still going to be people who deny that any of it matters...)


I personally take your results seriously, but I've had my own first hand experience with cameras that use Sony Exmor sensors, so I'm not surprised by the results.

 

[Loswr]

I'm someone who believes that the D750 has significantly better DR than any existing Canon sensor.

Me, too.

I just know that the resistance to and downright denial of the notion that improved DR matters, or that anyone is even remotely capable of performing a valid IQ test, or that Canon cameras might have a problem with anything, is extremely high here.

No.

There is general acknowledgment that more DR is desirable. Certainly some people/organizations are capable of performing valid IQ tests, but there are also many poorly done and biased tests out there, and cases where valid measurements are interpreted/summarized in a biased manner. As for Canon cameras having problems, that has certainly occurred and been widely acknowledged.

What there is high resistance to is the idea that 2-3 stops less DR and more noise when shadows are pushed >2-3 stops mean that Canon sensors deliver (in your words), "...poor, sub-par, unacceptable IQ." There is downright denial of the idea that DR is the only performance metric of a camera system that matters to most people.

I personally take your results seriously, but I've had my own first hand experience with cameras that use Sony Exmor sensors, so I'm not surprised by the results.

I doubt anyone would be surprised that a FF Exmor sensor has more DR than an APS-C Canon sensor. However, I would not be at all surprised to see certain individuals claiming that means the 7DII delivers poor IQ and shows a lack of innovation by Canon, and other such foolishness.

 

[privatebydesign]

I just know that the resistance to and downright denial of the notion that improved DR matters, or that anyone is even remotely capable of performing a valid IQ test, or that Canon cameras might have a problem with anything, is extremely high here. I think your test is valid. It should may DTaylor happy, however I'm pretty sure if/when he comes along, he'll have some way of 'debunking' what your test clearly shows.

What an inflamatory thing to say, kinda like shouting fire in a cinema, and I believe in the USA that is illegal. It is also a gross misrepresentation of the opinions that people here actually have, but you know that and did it anyway.........

So anybody that questions methodology, as I legitimately did and included specific areas of concern, is automatically labeled in denial of the holy Exmor scripture?

Way to go to incite the response you criticise people for and lay the foundation for the kinds of discussion you always claim you are interested in.

 

[ajfotofilmagem]

We all know that (until proven otherwise) Exmor sensor that has more DR than Canon sensors IN LOW ISO. So, DR tests would be most useful to compare Exmor versus Canon in ISO1600 and above. :

Someone wants to apply? 8)

 

[StudentOfLight]

We all know that (until proven otherwise) Exmor sensor that has more DR than Canon sensors IN LOW ISO. So, DR tests would be most useful to compare Exmor versus Canon in ISO1600 and above. :

Someone wants to apply? 8)

Scientific results should be repeatable. The fact that we have multiple methods to come to similar results would lend credibility to given claim. The process of verifying a claim is not a waste of time. If all people would just accept anything at face value then liars, cheats and deceptive quacks would run amok.

I've looked though my LR catalogue and here are my percentages of shots taken at different ISO settings:
50&100 - 29.3%
200 - 16.1%
400 - 14.6%
800 - 8.1%
1600 - 11.8%
3200 - 5%
6400 - 4.8%
12800 - 1.5%
25600 - 0.2%
51200 - 0.0%
102400 - 0.5%

Until I actually did this exercise I didn't realize how much I shot at low ISO. If I had just stuck with what my feelings were I would believe that I shot a lot more at 800-3200, but in reality that simply is not the case. Numbers don't lie (until they get to the statistician)

 

[Don Haines]

A true test of DR will compare similar cameras

There are lots of valid reasons for testing dis-similar cameras... One might be asking how the 7D2 compares to the 5D3.... one might want to know what the differences are between old tech and new tech.... between FF and Crop, between 4/3 and crop, between Canon, Nikon, and Sony... Lots of questions to be asked here and even more answers sought. We THINK we know the answers, but until testing is done, we are just guessing.

 

[Marsu42]

... dr ...

...

...

Yay, ladies and gentlemen, the game is ON again after one of the contestants withdraw from the first round from pure exhaustion. How many pages will the second round last? Make your bets and watch the show, now live on CR! :->

 

[ajfotofilmagem]

A true test of DR will compare similar cameras

There are lots of valid reasons for testing dis-similar cameras... One might be asking how the 7D2 compares to the 5D3.... one might want to know what the differences are between old tech and new tech.... between FF and Crop, between 4/3 and crop, between Canon, Nikon, and Sony... Lots of questions to be asked here and even more answers sought. We THINK we know the answers, but until testing is done, we are just guessing.

Don, you forgot the most important part of my message: Tests comparing a parameter (for example DR) must minimize all other possible variables.

I understand that some want to know "how much" of a difference between equipment of different categories and purposes. But the bottom line is always getting stuck in tests at ISO100 purposely under-exposed to lift the shadows in PP. :-[ This seems to me an excuse to prove to yourself something like "my dick is bigger than yours"... :-X

 

[jrista]

... dr ...

...

...

Yay, ladies and gentlemen, the game is ON again after one of the contestants withdraw from the first round from pure exhaustion. How many pages will the second round last? Make your bets and watch the show, now live on CR! :->

I have no intention of playing any games. I said what I said, in the way that I said it, because I believe that is how a core group of members in this community...operate. They break down and deny every test ever done, and always find some way of labeling them flawed, or incomplete, or unfair, or something like that. I call that denial.


People are free to take offense to that, or ignore it and just keep on making photos, or maybe actually break out of their comfort zone and give another brand a try if they are really curious about the truth (damn what anyone else says, including me, prove it to yourself if you are wondering what's what...hopefully with the realization that it isn't just low ISO DR that is better on competitors anymore...the A7s brings better DR throughout the entire ISO range). Otherwise, I'm not here to participate in a war.