5DII vs. 7D test, part 1

Discussion in 'Canon EOS' started by jackm|1, Nov 4, 2010.

  1. As an owner of both these bodies your test confirms my own view - that the 5DII produces noticeably higher quality images. I find that even in your first sets of images of the seat and trees the 5DII is more pleasing to look at. Everyone accepts that at high ISO, for wide angle and for pixel peaking / massive prints the 5DII outperforms the 7D. Your images show exactly what I find which is that the 5DII images (portraits, landscapes etc...) even at low ISO and small sizes look better than those from the 7D. This is not to say that the 7D is not a great camera (mine performs at least as well as my 1DIIN - if not better).
     
  2. I like the mouse-over effect for these kinds of comparisons. Seen it in a couple of other places and it really helps.
    One comparison I'd like to see: A photo with the 5DII at 85 mm with the same camera to subject distance as the 7D at 85 mm. Then, crop the 5DII image to match the 7D field of view.
     
  3. Ed,
    Why shoot a pic with the 5D II and then crop it? What kind of comparison is that? Just to see of the background blur is different?
     
  4. David, I apologize--I should have phrased it differently. I've seen a few arguments about the "telephoto" advantage of crop cameras and was wondering if it was really that much of an advantage. Right now I own a 40D and use a 70-200 f/2.8 for some sports (non-professionally). If I'm stretched all the way out to 200 mm to get in close am I better off with a 7D or cropping a 5DII image? (I'm probably better off with the 7D for other reasons when shooting sports, but just for arguments sake, please ignore any other attributes of either camera).
    FWIW, I'm not looking to make a point but as an engineer and certified nerd I'm just curious what the two images would look like.
     
  5. David,
    Us nature/wildlife guys want to know which images come out best due to the crop factor and higher mm advantage of the 7D versus the lower crop factor and lower mm of the 5D. Many of us are looking at getting the best picture we can at the best value. My 7D, 500mm F4, and 1.4 extender is sometimes still not long enough so we explore other possibilities.
     
  6. Ed, look under the 100% crops section, there is a crop from a 7D shot at 135mm. There is no mouse-over effect on that image, but you can compare that to the crop of the 5DII shot taken at 135mm. As they are both at 100%, that should tell you what kind of resolution the 7D ultimately has.

    At the same focal length, of course the 7D resolves more detail, but that is kind of an apples-to-oranges comparison (even more-so than a FF vs. APS-C comparison already is). Basically you're talking about comparing an 8mp image to an 18mp image.

    For field sports or anywhere you are focal-length limited, the 7D is the better choice, of course. I shot youth soccer this fall with my 7D and my 70-200/2.8II and was able to cover a good portion of the field. One day when I was not "on the clock" I shot my son's team again with my 5DII just for grins, and the 70-200/2.8II wide open. When I was within reach of the players, the results were dramatic. I liked the vignetting.
     
  7. Interesting and consistant with my own experience with my 5D MkII and 7D. High ISO is where the 5D2 really separates itself from the 7D, IME. I was surprised that you didn't demonstrate that.
     
  8. I did a fairly good comparison between the 5D MkII/1Ds MkIII cropped and the 7D. My conclusions were that in real world use the difference in IQ is absolutely marginal, you need to use MLU, large tripod, very fast shutter speeds or flash, optimum apertures, cable release and static high contrast fine detailed subjects etc to realise the potential resolution advantage that the 7D has on a cropped 5D MkII image. I did not get a 7D because of this and because I found the noise, even at base iso, to be too much.
    Like Philip says, that is not saying the 7D is not a great camera, it is, but it did not realise the advantages I thought it should for my use. If you want faster AF and better framing through the viewfinder etc then the 7D bests the 5D MkII comfortably for telephoto usage, if you think you are going to realise big resolution increases because it puts so many extra pixels on the subject you will be very disappointed even with perfect technique and the right subjects.
     
  9. I find the image quality from my 5D MK I, far superior to those from my 7D. I suspect most of the issue I have is what appears to be noise reduction artifacts, more than noise itself, even though I have high ISO noise reduction turned off on the 7D.
     
  10. Scott, I'm not sure I followed your experiment, but my shots taken with the 5DII and 7D both at 135mm and the same distance to the subject show a huge amount more detail resolved by the 7D. That said, this is only meaningful if you are focal-length limited. Which I was, this past soccer season, with only a 70-200 at my disposal.
     
  11. Jack,
    I don't see what you do then. I see your two 135mm images, both cropped to 100%, but I don't see "a huge amount more detail resolved by the 7D" anywhere, crop your 5D MkII image to the same as your 7D crop and upsample it. That was the very specific aim of my little test. How much extra resolution does the 7D realistically give me over a 1Ds MkIII/5D MkII cropped to the same size.
    Most people, yourself included, seem to think that there will be a huge difference, my shots show that when cropped and sized to match each other the bigger sensor with much less pixel density actually does not give up much at all, even under optimum conditions, when real world use is taken into account (AF, handholding, panning, not optimum lens aperture, not optimum exposure, contrast, subject detail etc etc) then there is effectively no difference between the two. Bench tests can be useful, but if the only way to realise the advantages the 7D pixel density has, is to use it on a bench, then it is of no value to me. Look at the crops in my thread with a single hair in them, over 200% crops, no post processing, effectively 10 foot prints and the differences between them are tiny.
    If you have a 5D MkII and a 300mm f4 and you want extra reach, your money is far better spent getting a 400 f5.6 rather than a 7D.
     
  12. Some people have been asking to see the performance of the cameras at the same focal length and subject distance, with the 5DII image cropped and resized to "match" the 7D image. I think I've done it right, please correct me if not.

    I've added crops from both cameras at 135mm f/3.5, at the same subject distance. The 5DII image has been cropped to APS-C dimensions (3510x2340) in DPP, transferred to Photoshop CS3 as a 16 bit TIFF, upsized using Bicubic Smoother to 7D size (5184x3456), then cropped to 800x1200 and saved at JPEG 12
    Scroll down to the bold word "Upscaling":
    http://www.jmphotocraft.com/5DII_v_7D/
     
  13. Jack: I think you have done an excellent and careful job at comparing two cameras that are very difficult to directly compare. Surely, you knew that no matter how you compared them, there would be those who criticized you methods. You took the time and trouble to actually do what others usually just speculate about. Thanks for sharing your results. I find the information very useful.
     
  14. Thank you Kerry.
     
  15. I offered no criticism at all, I too find the comparison interesting, the focal length and depth of field are as clear and graphic illustrations as you could wish for.
    I tested something completely different that related to other posters interests, so linked to it. I was testing for myself and for only one thing, real world resolution. It is shocking how different Jack's results are from mine. The new addition to his post is the only result I have ever seen that puts more noise in the same area from a 5D MkII sensor when compared to a 7D sensor. This alone makes me think something has been lost in the work flow.
    Here are my crops, these are over 200% enlargements of the same area of the sensors from a 1Ds MkIII and a 7D, the 7D shows a little more detail, certainly not "a huge amount", and more noise than the 1Ds MkIII. I did absolutely no processing to either crop.
    00XcoE-298369584.jpg
     
  16. I thought your comments were fair, Scott. Don't look at the pictures of the players and fans on the newspaper, they contain a lot of noise themselves, and also the plaque has a matte surface. I draw my conclusions from the text, which I think tells the real story.
    As for the workflow, did I do it right? Crop the 5DII image to APS-C dimensions (8.2mp), then upsize that to 18mp, then take an 800x1200 crop from that? Frankly I think such a comparison is nonsensical, because I would never do that in real life. Comparing upscaled interpolated pixels to native pixels doesn't seem fair or relevant to me.
    Maybe I should redo this test with a nice crisp book instead of a plaque printed from a newspaper, and maybe I should do it in better light? I used this plaque and not an actual newspaper because there was a slight breeze.
     
  17. Anybody taking interesting photos? Just wondering...
     
  18. Sure Brad. The same 7D + 70-200 combo as my test:
    [​IMG]
    [​IMG]
     
  19. Jack - I have the same comments here that I have over at dpreview. I believe, based on past comparisons, that ACR will do a better job of resolving the small text in the 7D shot. I see the same roughness there that I saw shooting map tests against film where the 7D + DPP failed to out resolve 35mm Velvia 50 on an Imacon, but the 7D + ACR was able to match the Velvia sample on small text and beat it on color detail and texture. The difference in the resolution of tiny text on the map between DPP and ACR was surprising really, and since then I quite frankly haven't used DPP for anything.
    ACR plus appropriate LCE and sharpening should just about eliminate any differences at 100% screen view, even after resizing to 5D2 dimensions, and should leave nothing to tell them apart at 16x24".
    You mentioned having an older version of PS. If you would like I would be happy to convert the RAW files and send you the results along with all settings.
    As to crop advantage when you're focal length limited, it's not huge when you're done dealing with real world challenges (i.e. motion) and making real world prints, but it is there. Using ACR might actually shrink that gap a bit though because, again, ACR seems much better at resolving fine detail near the limit of the sensor's ability. (Which becomes the 5D2 in that particular test.)
     
  20. Isn't this telling us what we already know? Here we see where the extra bucks gets us.
    To me the 5DII images are nicer, as they should be. The real issue is whether to jump to FX and stay there. I would always say yes to that. Sports is tricky though.
    "Extra reach" is always an advantage in that category, but image quality is also very important too. Whereas news also needs reach, but image quality is not nearly as important, so I suppose thats why we see more DX bodies at interersting happenings around town, but not so much on the sideline....almost almost 1Dxxxx and D3xxxx territory.
     
  21. Like I said at DPReview, thank for doing this Jack...and for being open to the questions (and critisism) that follow. Interesting about the 1Ds3 Scott!
     
  22. This opens up another line of discussion too.
    Remember when the 5D came out (and also when the Mk2 arrived)? There was this line of opinion that the prosumer body was producing as good or even better images than the pro body. There are many out there who still maintain that the original 5D makes the best landscapes, bar none. The same thing happened when the Nikon D700 came out but not quite so much.
    Side by side comparisons down to the pixel generated all this ethusiastic discussion around sensor pixel size and density and what sun glasses the sensor had etc etc. What we learned though, is that the camera software has as much to do with it as the sensor, the settings and the lens. And sometimes a manufacturer gets the combination just right...ergo the 5D.
    Its also the case with the D700. Then along comes the firmware upgrades to keep pace with or to reject....so many variables that making comparisons is often a waste of time.
    I draw a comparison with my computer that has Vista installed. Everyone bagged vista. I never had a problem and didn't know what the fuss was about. Then I installed SP2 and now its a dog and even going back to SP1 has not fixed the problem. Same with cameras. Now the same with the iPhone 3G. Install the upgrade OS to v4 and you will be in big trouble. I think cameras are the same and I do not have a lot of faith in upgrades anymore.
    Some also say that the first rev of the 5D2 software was poor and buggy. But now, magically, that seems to have gone away. Funny about that?
    So...we also need to know what rev of software the camera is running to make comparisons valid as well. Complicated, isn't it?
     
  23. Daniel, can you post the ACR vs DPP comparison?
    What I noticed is that ACR tries to guess up more detail at the expense of artificial renditions, whereas DPP is more accurate.
     
  24. Why does that kid have green hair in the back? Just curious...
     
  25. Based on the DXO info the best camera for me (by unprocessed RAWs) is the D30. If I want more resolution then the next best is the 5D. If I want decent AF then the next best is the 1DII.
    I have fallen out of love with my 7d. Not because of anything its done. But because an internet thread told me to.
    The brillinatly sharp nature and sports pics it was taking mean nothing anymore. The D30 has less noise. The 5D (mk1) is less prone to T-stop light loss (one difference between movie cams and stills cams - handheld metering vs TTL- doh!)
    All I've got to do now is find a D30/1Dmk2 with a decent screen that hasn't been hammered (there goes the 1DII) or is made of plastic (there goes the D30) and has decent AF (there goes the 5D and the 5DII)
    Where was I?
    Oh thats right. I hate my 7D.
     
  26. Don't worry too much Paul, that silly LL and DxO bit of rubbish has been fairly comprehensively debunked and shown to be the complete nonsense that it is. Just goes to show, you can make statistics say anything.
    To illustrate that point I once sat through an aeronautics lecture that proved, totally convincingly and with very great accuracy and detail, that it was impossible for a bumble bee to fly. It was done to show us the importance of testing, however, it is a good job bumble bees don't go to aeronautics lectures :)
     
  27. Brad, I wonder if its green hair spray dye that made the boys hair green or maybe he has been messing with his mums food dye.
     
  28. It is green hair spray. This was a fad this year in youth soccer in our area.
     
  29. Thank you for taking the time to do the test and post the results. Has anyone seen any test on prototype Sigma SD 1 camera to be released early 2011
    Jim
     
  30. Paul, are you serious? Or are you pulling our legs.
    The 7D is a fine camera. Too much pixel peeping is being done...too much photoshopping and corrections going on.
    I have a pro customer who has a number of brands in his bag and he tries to stay neutral. He seems to put a good slant on all this technical stuff. He prints his images with the minimum of post processing and says this: There will always be technical differences between the results obtained. But print off an image in A4 and ask yourself these questions:
    1. Good focus and sharpness?
    2. Faithful colour reproduction?
    3. Nice, interesting composition?
    4. Would I hang on my wall?
    Do that with the 7D and be satisfied. Sure, you might get better results with a different model or brand of camera. The 5D2 renders images beautifully...but for what you are doing, will you notice the difference "up on the wall". There is far more to consider than just dots under a magnifying glass.
    Of all the brands out there, you have to start spending really serious money including lenses, to get the kinds of technical results you hope for. This dissatisfaction is common to younger people who never lived with film, where the rendition of artifacts in a shot was usually only dependent on the lens and choice of film.
    The other thing is this, and I do feel that its sometimes a factor. Camera companies now are also software companies, and as the camera software evolves via upgrades, sometimes we are not told about subtle changes that these companies make to improve image quality. They usually only tell you about tweaking extra features. This is a factor as well.
    Buying and being satisfied for 2-3 years with your purchase is one of the big challenges facing the DSLR buyer. But thats the way it is. Getting yourself all wound up and worried is no place to be.
     

Share This Page