Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

The megapixel sweetspot. Canon 5D mkIII or Nikon D800?

  1. Mar 3, 2012 #1
    Recently Canon and Nikon announced successors for their lower price (but still 3k-ish) full frame cameras. The successor of the Nikon D700, the D800, boosts the original 12 megapixels to a whopping 36 megapixels, however Canon made even more impression by boosting the original 21 megapixels of the 5D mkII to an impressive 22 megapixel of the 5D mkIII

    So people go: what the bleeb is going on here?

    First question is, wasn't Canon unable to follow the megapixel race? Nope, they patented a 50MP sensor some years ago already, but it hasn't shown up in any camera yet. Also if you consider that they can put 10-14 megapixel in some 4 x 6 mm sensor of pocket cameras then at the same pixel count per area, that would be 360 - 500 megapixel on a full frame.

    Apparantly there is more to pixel counting that meets the eye and there must be an optimum (sweetspot), but is Nikon right or Canon?

  2. jcsd
  3. Mar 3, 2012 #2


    User Avatar

    Staff: Mentor

    I wonder about the same, but not because of these cameras.

    I guess some time ago Canon made a decision Mark III will have such a sensor, while Nikon made a different decision - I don't think there is much more behind. It will make an interesting comparison when both cameras will show their potential in the future.

    However, in my experience taking pictures that really make use of EOS 7D sensor (pictures where every pixel counts) is already pretty demanding, even in controlled environment (studio, tripod). I wonder if a full size sensor with larger pixels is not giving better results on average - but I have no way of testing it by myself, I don't know anyone with 5D here.
  4. Mar 3, 2012 #3
    I have a 5D mark 2 and I have never ever once wished for more megapixels. In most cases, the quality of the optics limits the total resolution more than the pixel density.

    The weak point of the 5D2 is the AF. That seems to have improved a lot in the 5D3.

    I am extremely happy with the full frame sensor. Often for tele pictures like birding I use it as a 200% viewfinder for an APS-C-sized crop. For wide to medium focal lengths I rarely crop. The 24-105 is a great match for the camera.
  5. Mar 3, 2012 #4
    Yes that's important, but mainly because of the difraction that occurs at higher aperture values (physical law), even with ideal glass this difraction limit is inevitable and with current pixel densities this starts around F8-11.

    On the other side, at less than F2.8 lens errors usually start limiting sharpness also exceeding the sensor resolution, giving a usual aperture sweet spot between F2.8 and F8 where sensor and lens are best match. So, extra pixels can only be useful around that range.
  6. Mar 3, 2012 #5
    Have they released any info about the exact physical size of the sensors? I would think such a high pixel count would also introduce a large amount of noise.
  7. Mar 3, 2012 #6

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    Hee hee- Nikon and Canon finally have cameras that can compete with my 2-year old Sony.
  8. Mar 3, 2012 #7
    Yes standard SLR. 24 x 36mm
  9. Mar 3, 2012 #8
    Let's have a look at that with the very interactive standard DPR studio comparising tool.

    I used the Sony NEX-7 test with the APS-C sensor with the highest pixel count (25m) and changed it here to compare it at ISO 6400, with Andies Camera, hinted to earlier, the Canon 5D mkII and the current best high ISO performer the Nikon D3S with 12mp, to get this:

    http://dl.dropbox.com/u/22026080/comparing.jpg [Broken]

    Note that Canon had something comparable to Andies camera and that the high pixel pitch of the NEX induces such a noise level that it disqualifies it at high ISO.

    So why would I use 6400 ISO? Because the bride and groom would be very disappointed if the candle light church ceremony would look like pea soup, or you can't get the indoor soccer shot sharp.
    Last edited by a moderator: May 5, 2017
  10. Mar 3, 2012 #9
    Indoor soccer in a candle-lit church during a wedding? I don't hang out with the right kind of people...

    The D3S' performance is impressive.
  11. Mar 3, 2012 #10

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    I don't have the NEX-7, I have the a850, which has the same sensor etc as the a900, also reviewed on dpreview.
  12. Mar 3, 2012 #11
    Sorry I was misleading, the NEX-7 is obviously only on the market for a few months, so I intended the crop on the upper right of the A900 to compare with the lower right.
  13. Mar 3, 2012 #12
    Another element of interest is file size. Look at the crop comparising shot again, but now at the file sizes.

    Note that the JPG (highest quality) of the NEX-7 is actually larger (28.3Mb) than the Raw file (24.0Mb), simply because all that noise doesn't compress very well, especially compared to the smoot result of the D3S (8.6Mb - 24.8Mb), resulting in larger processing and storage times, filling up the buffer earlier, etc etc.

    A 36MP Nikon 800D noisier file may be expected to have that same problem compared to the (probably) smoother 22MP 5D mkIII files.
  14. Mar 3, 2012 #13

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    I see what you mean, and there are some other aspects of the comparison I don't quite understand (I haven't read the detailed descriptions)- for example, the use of any in-camera noise reduction, sharpening, etc. Also, the FOV of the 100% crops don't appear as I expect, based on the size of the pixels.

    I *definitely* want to see some quantitative images from these new models...
  15. Mar 4, 2012 #14
    Meanwhile here is a series of improvised shots with a preproduction model of the 5D mkIII

    And a crop from the 6400 ISO shot:

    http://dl.dropbox.com/u/22026080/mkiii-6400.jpg [Broken]

    Comparing this (processed jpg) with the raw crops shown earlier is apples with oranges, obviously. However it sure looks good.
    Last edited by a moderator: May 5, 2017
  16. Mar 4, 2012 #15
    My wife is going to kill me if I ask for 3.5k$ camera.

    What I really want to know is how well the AF performs on the 5D3.

    This was taken with a 5D2 at 3200 iso (EF 50mm f/1.4 at f/2.8, 1/350sec, no flash, RAW processed with Adobe Lightroom v2):

    IMG_2447 by carsten de, on Flickr
  17. Mar 4, 2012 #16
    Maybe if you tell her that it beats the current high ISO leader hands down at ISO 102400?

    http://dl.dropbox.com/u/22026080/mkiii-102400.jpg [Broken]

    Ah well never mind.

    It uses the same system as the new top model 1DX which should be a significant improvement over the 1DmkIV which is clearly better than the 7D, which is already awesome in my experience.

    I should take a job and save for it.
    Last edited by a moderator: May 5, 2017
  18. Mar 4, 2012 #17
    I guess that's an argument. It actually saves 1500$ compared to a 1Dmk4 which B&H now lists around 5k$....
  19. Mar 23, 2012 #18


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I am not looking forward to paying 3K for a camera either. I have had my eye on purchasing a D7000 as my first digital SLR. It is closer to $1500 and has good reviews. Would that be a good camera for the price? I would rather stick with a Nikon because I already have several lenses from a film camera.
  20. Mar 23, 2012 #19
  21. Mar 23, 2012 #20


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Thanks, Andre. I will check out the 5100 also.

    I have previously verified the compatibility on the lenses on another lens compatibility site. I even took the lenses that I have to the store to make absolutely sure that they would fit. I know that they would only be good for manual mode but, I would hate for them to go to waste.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook