Photography Bay

Digital Camera News, Reviews and Tips

Canon 5D Mark II and Nikon D700 High ISO – Quick Comparison

By

I put together a few quick frames from ISO 1600 to ISO 25600 for the Nikon D700 and Canon 5D Mark II tonight.  Nothing scientific, just a few shots snapped off at my desk.

Settings: tripod, aperture priority @ f/8, evaluative metering, auto white balance, ISO 1600-25600 in full stops – lenses used were 50mm f/1.8 on the Nikon & 50mm f/1.4 on the Canon – RAW capture – exported to JPEG via Lightroom with no edits other than the 100% crops

As you’ll see, the 5D Mark II tended to underexpose a bit, which I left alone for the purposes of this quick test. I’ll go more in-depth in a later post on these two cameras.

Click on any image to enlarge to full size (Note that I mildly resized the 100% crops to fit into the post, so you’ll need to click to get the 100% crop, which are between 650-800px wide for both cameras)  Also, bear in mind that the 5D Mark II full image files go up to 32MB+ in file size, so dial-up readers need not apply.

Canon 5D Mark II ISO 1600

Canon 5D Mark II ISO 1600 @ 100%

Nikon D700 ISO 1600

Nikon D700 ISO 1600 @ 100%

Canon 5D Mark II ISO 3200

Canon 5D Mark II ISO 3200 @ 100%

Nikon D700 ISO 3200

Nikon D700 ISO 3200 @ 100%

Canon 5D Mark II ISO 6400

Canon 5D Mark II ISO 6400 @ 100%

Nikon D700 ISO 6400

Nikon D700 ISO 6400 @ 100%

Canon 5D Mark II ISO 12800

Canon 5D Mark II ISO 12800 @ 100%

Nikon D700 ISO 12800

Nikon D700 ISO 12800 @ 100%

Canon 5D Mark II ISO 25600

Canon 5D Mark II ISO 25600 @ 100%

Nikon D700 ISO 25600

Nikon D700 ISO 25600 @ 100%

I was actually a little surprised at the differences between these two.  Kudos to Nikon for sticking with a 12 megapixel sensor, although I won’t be parting ways with my 5D Mark II anytime soon.  So what do think about the noise comparisons between these two?

Stay tuned to Photography Bay’s Canon 5D Mark II Reviews and Resources and Nikon D700 Reviews and Resources for more coming soon.

 

Comments

  1. I’m a Canon shooter but have messed around with the Nikon D3, I’m impressed with their high ISO, they finally got rid of the noise issues they’ve had since the D70 came out. The Canon noise is much chunkier, but that might be partly due to the off exposure of it. I have zero plans of switching, but I also rarely have a need to shoot above 800iso, so my 5d is fine. Wedding photographers on the other hand, they need to crank the iso when they can’t use a flash during the ceremony, I know of at least one person who sold their entire Canon outfit, including a 1ds mkII and 1d mkIII to switch to a pair of Nikon D3’s just for the noise reason alone, and they almost never use the speed of the D3 (this was prior to the D700 being available too).

  3. wow, i’m surprised for nikon d700 … qood quality. 2300 euros for canon 5d, more then 2500 euros for d700 … i-ve been expecting more quality image noise for 5d II, a big advantage for 5d is the HD filming …

  4. Thanks for this, much more useful than some other ISO crops. I’d say the D700 has an advantage, less noise, and it’s better looking. (I’m amazed that ISO 1600 is just clean as whistle).

    The 5D2 probably would be better if it weren’t underexposing, but clearly resolving more detail.

    Tough decisions deciding on either.

  6. Looked like the D700 was clearly better (to my eyes) up until 12800. At that point, I thought the mkII shots looked decidedly better.

    For what it’s worth, I think both cameras are excellent. I’m a D300 shooter w/ plans for a 700 in the future, but I hold no grudges against the Canons. Great cameras. The competition benefits us all.

    Thanks for the quick test. My only criticism would be that the photos should be grouped w/ their counterparts (ie- mkII shot, D700 shot, mkII crop, D700 crop). As it is, it’s a bit of a pain to scroll back and forth trying to compare the two. I suppose those with dual-monitors or 46″ monitors don’t have that problem :P Other than that, well done.

  7. The D700 at 100% has less noise than the Canon at 100%; however, what is important is when, for example, the Canon’s 21.1mp resolution is downrezzed to Nikon’s 12.1mp; then you can compare them side by side, at the same resolution. This is what it will actually look like when printed on paper of the same size. I would imagine that while Nikon will still have a slight advantage in the noise department, the Canon will be remarkably sharper, especially at high ISOs. Theoretically, a sensor with a higher mp count, using the same technology as a sensor with a lower mp count, should produce better high ISO images when resized to the lower mp camera. The problem is in creating a high mp camera with the same technology; it is intrinsically harder to apply all the technology used in the D3 to something with a higher mp count, as each pixel is smaller.

  8. Nice comparison, but it looked like the focus was changing slightly with each shot? The text on the movie tickets was a great judge of noise when it was in focus. As it changed, it became less helpful. Locking each camera on an identical white balance would have also been helpful.

    Great comparison otherwise! I hate to bite the hand that offered up such a fine set of photos. I’m another D300 shooter who lusts after the D700, but also have nothing against Canon. If anything, I’m often first in line to praise Canon’s excellent lenses over what Nikon currently offers.

    Aside from the noise, did it seem to anyone else that the Canon 25,600 shot had *more details* than any other high-iso shot? It seemed to beat out the Nikon 12800, Nikon 25600, AND the Canon 12800 in terms of detail. This could just be due to finer focus one shot to the next though…

  9. Just can’t wait for Black Silicon CMOS to be worked into DSLR’S. It’s been here a while but not fully operational in the consumer market. The discovery of it in 1999. It should have ISO’s up 102,400+ with very little noise. That will be revolutionary. The draw backs are that it might need more filters on the cmos. Due to its sensitivity. I say 5 years and it will be on the D5x and 1D MK6. This will be a long wait. The only company working with Black Silicon that I know of now is SiOnyx.

  10. @revlimit – Good points on the focus issue. I did re-engage the AF for each shot but I was doing this quick. I’ll take a little more time with the next round of tests. Thanks for the comment and suggestions.

  11. I appreciate this comparo. It contributes to seeing what is possible to current
    DSLR tech.

    A couple of suggestions. Sure, it’s a lot of work if the the cameras metering and/or ISO calibration is off, but I’d like to see the exposures match up.

    Second, if you have the time and the willingness, I’d like to see the AF taken out of the way by manual focus with Live View at 10x. It would make it easier to see what detail is lost to noise and/or noise reduction as opposed to detail that was never there because the lens wasn’t providing it.

    Thanks.

  14. Eric,
    I was wondering, I think you are saying that you did RAW conversion with ACR 5.2 built into Lightroom, right? I have done some tests and I think Adobe released ACR before checking against Canon’s DPP to see if they got it right. Adobe didn’t. There appears there are special characteristics to the 5DMII sensor. DPP does a visibly better job converting, especially with higher ISO noise. Many of the artifacts (regrettably not the banding and I don’t see black dots) disappear if you use DPP. This includes color splotching and grain, and it improves detail vs. ACR. I recommend holding your nose and trying out DPP until ACR is fixed.

    See my web site for an animated GIF that demonstrates the difference.

    http://www.luiswatkins.com/5dmarkii/dppvsacr.htm

  15. no surprise. Nikon has set the bar pretty high and seems canon just wasn’t able to overcome the downsides of higher mp sensor counts. However, canon did a good job compared to the sony A900. It will be interesting to see the successor to the d700 which will probably use the d3x’s sensor. As d3x reviews are starting to come out, we’ll get a better idea about how nikon’s sensor designed (made by sony) compares to the canon and the a900.

  16. I think it is only logical that the photos from the D700 have lesser noise. The D5 II and D700 are both full frame sensors but the D700 has less pixels on his sensor. So the pixels on the D700 sensor are bigger than the pixels on the canon D5 II sensor. Because the pixels on the Nikon sensor are bigger they also are better “light readers” than the smaller pixels on the Canon sensor. Because of that the Nikon sensor produces lesser noise in higher ISO levels than the D5 II. Only I think it would be more fair to compare the Canon D5 II with the Nikon D3x than with the D700.

  17. Thanks for the comparison photos; I wanted to upgrade to Full Frame from my Eos 40D – a cracking camera by all accounts, which produced outstanding photos. There are some funny people out there: the Canon and Nikon losers society, who perpetuate this myth that you are renouncing your religion, family and core values by choosing one brand over another. I was really stuck between the 5d MkII and the D700 and in a typically male OCD way, I read myself to death on all the reviews and finally decided on the D700, as it seemed to suit my needs more (action, low light etc). Looking at these photos only confirms that they are both really capable cameras but if you major in landscapes, go for the Canon, if your stock in trade is more action, then the Nikon is probably the one. I perhaps should not admit this but (as an amateur) the photos look identical to me.

  18. I have had a canon 40D and was waiting for the 5D Mark2. It arrived the other day and after several tries to get a good indor shot at various iso settings I decided that this was not for me, I had just rented a D700 for 10 days while on a cruise and not a bad indoor shot was had. The D700 seems to gather much more light at a given iso setting.

  19. Thanks for the comparison. Have you tried down-sizing the 5D Mark II images and comparing them at the same size as the D700? (Or upsizing the D700 images) The 5D Mark II has more resolution to start with, so it can handle more noise reduction. If you apply NR, then downsize the 5D Mark II images they should look significantly better. I tried it with my 5D2, and the results look good, but I didn’t have a D700 to compare to:
    http://www.andrewyip.com/blog/2008/12/18/canon-5d-mark-ii-high-iso-noise-test

    It looks like it wouldn’t work to downsize this set, since the composition isn’t exactly the same…

  20. Nikon owner D700. both cameras are excellent for low light the D700 shows the least noise but it’s close & the 5D 2 higher iso 12800 & 25600 are better as there is more detail so long as use them for a b&w print.

  21. Well, for all the comments and visual conclusions on which is better, I guess both cameras performed well. But if there’s a real nitpicking on which is really better, I think its not a good idea, since the 5D MKII yields more megapixels. Its needs some magnification before truely seeing all the color noises.
    Compare them to film I guess?

  22. These results are of no surprise to me at all having shot the 5D MK11 along side my own D700. The poorer High ISO performance of the 5D is matched by its lame Auto Focus, terrible UI and slow Flash Synch speed (yes really). Add to that the less than great build quality and Canon owners must be quite disappointed.

  24. I am highly sceptical of this ‘test’ for two reasons.

    White balance is off making it difficult to judge the noise performance alone.

    Nikon employs noise reduction automatically by default. Looking at EXIF, appears Canon does not have High ISO Noise Reduction turned on.

  26. @ mike followed your link and i don’t see how you come to the conclusion re the white balance of the Canon seems ok but the D700 has horrible white balance it’s made a black camera brown.It also seems the 5D mkII use’s a longer shutter speed than the D700.
    In the real world you use high ISO to gain shutter speed .
    I guess that’s why DxOMark Sensor has the D700 at number 1 for High ISO performance at 2303 & the 5D mark II is fourth with 1815 a difference of 418.
    The Nikon D3X ranks third with 1992, a D700X will probably be released later this year
    Canon have done a great job with improving the high ISO noise of the
    5D markII but the D3X with more resolution has done better

  27. DXO is worthless test, they rank 5D below D90 (that’s a joke). D3X is double the 5D Mark II and only slightly slightly better. The real winner is clear 5D Mark II.

  28. mike, your test are clearly unfair. dont post a test if its dishonest and misleading. the people here needs the truth not your make believe test.

  30. If canon 5D II only had 12mp,100% much better noise than d700, hw about low iso 50?
    d3x is 3 times more expensive than 5d II, do not compare them, same as comparing 5d II to nikon d80???, but for 3 times more expensive, d3x is close to 5d II, so the best camera 4 high MP with low noise and price is Canon 5D II,

  32. @Steve I must agree with you that D700X will be built to kick 5Dii’s butt. That is when and if it actually arrives sometime next year. I will dust of all my Nikon lenses for the occasion.

  33. Well to be honest, an adapter is in the mail and on its way ;) Can’t have those old manual Nikkor primes collecting dust on their pouches.

  34. Something is obvious, they are almost equal, but Canon has 21 instead of 12 of Nikon. And as always, Nikon is trying to achieve canon sucess, first, trying to get a full frame, then, having high iso development slr, later, more than 11 af points, now, they took out the d3x, wich 20 mp.

    Nikon will always copy what Canon does, cause Canon EATS Nikon, get over with it Nikon fanboys, Canon is your daddy.

    and dont forget that in long teles, Nikon is out of the fight…

  35. All I like about Nikon is the build.
    By the way, what are the main differences between D3 and D3x? I should say advantages of D3x. The obvious difference is the only High Resolution sensor but Nikon needs to raise the price double to produce the sensor alone?
    When d700x is out, there will be an another camera from Canon (maybe 7D but I don’t know exactly) for Nikon to catch again. Moreover, I don’t think the price of d700x would be affordable for most of the dslr users.
    Maybe D700 is better in NR but 5Dmkii is 21MP. With that pixel count with that much noise, it’s like comparison between an apple and an orange. Dare to compare the sharpness? Look at the letter “T” from both iso1600 pictures. I don’t know how the photographer focused the subject in them but normally Canon make sharper pictures. That’s why I chose Canon for Landscapes. Less expensive, sharper, good for those who work with big prints.
    Anyway, I am not a Canon-faithful. I also use Nikon. Nikon also make very vibrant colorful pictures and some people love this advantage. Since both cameras have pros and cons and choose the one that suit more. Do we still need to argue who is better? No camera is better in everything.

  36. Your focus points were different on the Canon. Seems the blue book in bkgnd was in focus on canon shots, but the ticket in foreground was on the Nikon shots. These two cameras are ~very~ close in performance, but who looks at pixels anyway? My clients dont.

  37. Confusion, confusion, confusion. I read myself to death over these two camera, still couldn’t decide which one to buy.

  38. This comparison would be improved if you changed the layout so that I could see the relevant image pairs side by side.

  40. Well this comparison is not fair because when you take close ups from the distance vs taking pictures from distance with telezoom , you will see more noise if its a closup. so this comparison is nothing more then bull. Distance, zoom, and light should be same in order to compare plus the lense.

  41. D700x probably be about $700 more expensive then MKii. So for me pricing is every thing because for that difference I could buy a better lense. Majority of us do portraits and and land scape so there is no need to spend extra money for little non-noticable difference in picture. The Pixels are every thing because you can crop the picture at will to point where D700 can’t touch it.

  42. Men,you couldn’t post iso comparation between a camera set in correct wb,for the nikon,and a canon setted in incorrect white balance for the reason that the noise in an incorrect wb register the light in a spectar,that could influence the cmos noise due to termal noise

  43. This test is a little in favour of the D700 for a couple of reasons:

    1.) Exposure is better on the D700 and underexposed on the 5DmkII. Noise in digital cameras is exponentially higher in dark areas, so it is absolutely crucial to get the correct exposure!

    2.) The WB is off on the 5DmkII shots, it is to orange, also increasing the noise in the final image.

    3.) Comparing shots at 100% is uninteresting. What should and does matter is how the noise is visible at the full image, print or on screen. I can’t think of any application where per pixel noise is of importance!

    That being said, if you repeat this test again with correct WB and exposure I am almost certain you will end up getting virtually indistinguishable results when viewing the full images (in print or on screen).

    Best regards!

  44. Nikon vs. Canon, Sony vs. Olympus. I think, the lenses and accessories in your camera bag are probably the deciding factor if you have been eyeballing this pair of cameras. Behind the Camera is a photographer and befor the light. For many years, this technik was a wonderfull dream for every photographer. – Now – make pictures, from Cologne, Dirk Baumbach

  46. There is one problem with this test.
    When you under expose a picture it adds much more noise than a properly exposed picture.
    you have to dial up the exposure about 2/3rds
    I have done this test with my 5D Mark II and my mom’s D700 and the difference is not that big but the D700 still wins, it’s just the margin is not that big.
    I also used a Nikon lens adapter, so we even used the same lens.

  47. I agree with PC,
    “When you under expose a picture it adds much more noise than a properly exposed picture”. But there’s more to it….

    Under exposing even by a 1/3 stop can make a Huge difference in the noise seen. Many Pro photographers suggest that when shooting at high ISO, be very careful as to not underexpose your shot. Also, Shoot in Raw and over expose by half a stop and then bring it down in post if needed.

    I have seen images from a 50D @ ISO 3200 where the Noise looks allot better than these specific images of the 5DmkII. (no noise reduction or post processing done).

    In addition, you will find it interesting to note that the lenses used also make a difference in Noise and exposure. Higher quality lenses like Canon L lenses or even the 17-55 2.8 IS show less Chroma noise. And for the latter, you must be wondering “Exposure?”… There are many instances where there are differences in exposures levels between different lenses at the same F, shutter speed and ISO on the same camera, in some cases as much as 1Stop of light.

    So all these testes that specify the “settings” or “setup” used and specify the Same F stop, Shutter speed and ISO are less than 85% accurate.
    More over when you take into account that there is a difference in exposure levels in Cameras of the same manufacturer as well.

Trackbacks