Is the Canon EOS 5D Mark II Really Sharper Than the Nikon D3?

After I published my review of the 5D Mark II, some people were quite surprised that the 100% crops of 5D Mark II photos looked significantly sharper and more detailed than those from Nikon D3 (see sample above, 5D Mark II on the left and D3 on the right). Rather than believing that the difference was due to higher resolution sensor of the 5D Mark II, some people were questioning whether the focus for the D3 photos were wrong, or maybe it was unfair to use a Nikkor 35mm f/2 prime lens on the D3 to compare with the EF 16-35mm f/2.8L USM zoom lens on the Canon. If I gave an unfair advantage to a camera, in fact, it would have been in favor of the Nikon. The 35mm prime should in all fairness be even sharper than the Canon 16-35mm zoom.

However, I am not the only one who can demonstrate this difference. You can go to Imaging Resource’s comparometer, then download full size pics from both the 5D Mark II and D3 to do the comparison ( ). However, it would be best to download the pics to your computer and resize one of the pics to match the 2 cameras files resolution. Either downsize the 5D mark II file to match D3 resolution or upscale D3 file to match 5D Mark II resolution.

I have downloaded ISO 100 photos from both cameras and downscale the 5D Mark II photo to match D3 resolution and made the following 100% crops comparison. Again, 5D Mark II crop is on the left an D3 crop on the right.


I think it is quite clear that the 5D Mark II crop on the left looked sharper and more detailed than the D3 crop on the left. Well, actually, if you look at the D3 photo at 100% on it’s own, it looks pretty sharp. It is only when you compare it side by side with the higher resolution 5D Mark that make the D3 crop looks soft.


27 thoughts on “Is the Canon EOS 5D Mark II Really Sharper Than the Nikon D3?

  1. Roland, thanks for posting again. I too was shocked at first glance. But in reality, what do you expect?

    Take 21 megapixels and down-size it to 12, you’re getting 21 MP worth of detail in a 12 MP image. It’s that simple.

    If you want to compare the 5D mk2 versus the D3 on a per-pixel level, what you’d have to do is either back up or zoom out with the 5D mk2 until you achieve the same image when both images are viewed at 100%. THEN, you’ll be able to see which camera resolves more detail at a per-pixel level. Would you be giving an un-fair advantage to the D3’s larger pixels? Or would the Canon AA filter keep the 5D mk2 in the lead? I dunno, I don’t own either camera…

    (Care to give it a shot, Roland?)

    One thing’s for sure, that D3X sure gives the 5D mk2 a run for it’s money in the high ISO department, shockingly! Here I was, assuming (like everybody else) That Nikon couldn’t do it…


    • Hi Matthew,

      Actually, it’s not what I expected or not. Not the point of this post. It was only in reply to those who doubt the validity of my D3 images and thought I might have misfocused or that the Nikkor 35mm f2 was not a sharp lens to begin with.

      Anyway, that sounds like an interesting test, but would be quite difficult to do and would take a lot of time. It would be difficult know how far to back up exactly with the 5D mark II to achieve the same image size as the D3 at 100%, and cannot be simply judged on the camera itself. It will have to take multiple attempt of shooting, loading on the computer, view it, then try again, shoot again, load agin and view again….. until I can get it right.

      Well, I’ll think about doing it if I have some time after the holiday season. It will continue to be a bit hectic in the coming week. Thanks for the idea anyway!

  2. Hey Roland,

    what I’d do is overlay any 12 megapixel image on top of any 21 megapixel image, and mark how much bigger the 21 megapixel image is. Then capture a D3 image at your initial distance, look at that image, and capture 5D mk2 images backing up in increments until it looks like it is the same difference as the “dummy” 12vs21 already on the computer…

    Haha okay I admit, it’s complicated. And not worth the time for me since I can’t afford either cameras, and will simply be waiting for a 24 MP D700X anyways. 😛

    But you’re welcome to give it a shot, in the name of science!


  3. I am not an expert when it comes to how to best compare these camera. I have seen other reviews show 5D2 having the edge on the D3 to very close to the D3 have slight edge on the 5D2. D3 certainly cost a lot more. Personally, I think the message here is that the 5D2 sensor is performing very well even with it being a 21 pixel sensor. It a $2600 camera and it not a 7K or 10K camera. I am a proud owner now and based on my own general use of the camera so far, it is performing very well and I am getting very sharp pictures.

  4. Hi Roland,

    I value your analysis a lot.

    May I suggest using the same lens on both cameras. This way, you’d be comparing the cameras’ sensors fairly and not worry about differences between lenses. A nikkor-2-eos adapter would allow you to use the Nikkor 35mm on the 5DII.

    Also, how much of an improvement, if any, does the Mark II have in the noise department over the Mark I and others?

    Keep up the good work!


  5. Pingback: More articels about the Canon 5D Mark II on “The World according to Roland” [Lim] « martin zeile photography | photoblog - digital art - virtual bio

  6. Are we all sure we’re looking at “resolution” and not “sharpness”?

    Apply Photoshop Smart Sharpen (100%, 1px Radius, Gaussian Blur) to the D3 shots and they look identical in detail to the 5D2.

    I think a much more important issue is the blown-out highlights on the Nikon. This will ruin images faster than lack of megapixels.

    The Mark seems to smoke the D3 in this regard. Or was the lighting inconsistent?

    Another James

  7. Pingback: M2 Canon 5D Mk II de serie: Pruebas y reviews de la vers. definitiva - Pgina 24 -

  8. What was the incamera sharpening set to on both cameras for these shots? Could you try this? Turn the in-camera sharpening off on both cameras, take your shots and show us. Also, load the shots into photo shop, apply the same sharpening and see how they compare and show us. Normally Nikon’s default incamera sharpening value is lower than Canon’s, so you are seeing post processing differences, not sensor / lense / camera sharpness differences. I also note the writing on the bottle lid is white for the Canon, gold for the Nikon. What colour is it actually? If its gold, then that would also point to more sharpnening being applied with the Canon, as its pushed the gold into becoming white.

    • Hi Paul,

      If you had actually read what I wrote on this post and my article on 5D Mark II review, you would know that I took all the test shots in RAW and in camera sharpening only affects camera generated jpeg files, not RAW.

      I did not add any sharpening in Photoshop either, but if I had added the same amount of sharpening, on both crops, the 5D Mark II crops would still look sharper anyway.

      Again, if you had read what I wrote, you would have known that the bottom crops of the bottle were not taken by me, I copied them from’s comparometer.

      You don’t have to agree with my findings and you can very well have queries about my methodology, but by just looking at the photos without reading through the text, you would not know the reason why I wrote this post and your comments and suggestion would be out of context.

  9. To Roland:
    5D mark II has to be compared in fact with Nikon D3X with its 24 Mp for the sharpness. For D3, the advantage is the image quality with a very high iso level. But with 100 iso and especially in studio shooting, if we add the price comparison, 5D Mark II is the winner with KO.

    I’m wainting for one these period, it will come by the end of April 😀

    Thanks for article 😉

    • Hi Chedy,

      Thanks for the comment. One can compare the 5D Mark II to almost any other camera out there for a whole variety of reasons. IMHO, there is no right and wrong in doing any particular comparisons, so long as one clearly state the purpose of the comparison and what he or she wanted to demonstrate. So you can compare the 5D Mark II to D3X, Sony A900, some p&s digicam or even medium format camera.

      There was a purpose for this article and I have explained it already (in fact more than once), so I am not going to explain my reasons again. Anyone can agree or disagree with me, it really doesn’t matter.

  10. Ooooh – can’t you all see, the Canon just blew the Nikon into da weeeds!! Just look at that detail and… and…

    Seriously though, both cameras are obviously superb and each has particular strengths which for the moment would appear to be class leading. We are all gunning for our chosen ‘brand’ to which most of us were committed some time before these cameras were so much as a twinkle in their creators eye..

    For what it’s worth I just love my mkII and for so many reasons, but if I had a D3 I’m sure it’d be gettin’ that same lovin’ feeling!

    Roland, thanks for taking the time to test this stuff!

  11. I agree with you Mr Lim, it is just a snap comparison, it is about sharpness. Full comparison may also include noise, handling, speed etc.

  12. How the f.. did you get 5D2 to focus? I just sold my 5D2 after 1 months of trial and error.
    The autofocus simply sucks. Even with manual I (and many other, try google this) get soft pictures out of the canon. Buyers beware.

    • HI gribben,

      I have never had any problems getting sharp photos with the 5D Mark II and although I wished the 5D Mark II has a more modern and better autofocus system, it still works fine i most situation and I have no problem getting to focus most of the time.

      Although there are some people that complained about 5D Mark II autofocus, like you. There are also many people who have no problem getting sharp photos and autofocus to work on the 5D Mark II. They just don’t go around shouting on internet like the few who claimed to have problems.

  13. Nikon such as D3 uses fake RAW hence they can make a rebadged Sony sensor to have less noise than Sony A700 using the same sensor by reducing sharpness and details, hence some noise reduction is already applied in Nikon NEF files.

    If you look at the 5DMk2 sRAW1 which is Canon’s own fake RAW you can see it has even less noise at high ISO than D3/D700 but also has less detail than the 5DMk2 full 21MP RAW.

  14. Hi roland,

    I find your blog, just searching for some other issues of the 5dMii i own. It was great reading all the comparisons and the explained new features.
    I would like if you can help me with a question? Why does canon or especifically 5dmii images have such a redish tone, especially on skin tones. Is there any way to correct this without post processing?

    Thanks for your time.

    • Hi Diego,

      I have never experienced reddish skin tone with my 5D Mark II, so I can’t really say for sure what your problem maybe. If you can show a sample somehow, that would help.

  15. that’s seem like very unreal comparison, the nikon images were either out of focus on shooting or blurred on post pro.

    • If you don’t believe my results or those from, then I really having nothing more to say. You can choose to believe what you want to believe that we all purposely either mis-focus the images on the Nikon D3 or blur the D3 images during post-processing. You can believe that Elvis is still alive or aliens killed Marilyn Monroe.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s