202 Comments

  1. It looks like a used 1D MK II may be much better for farless $$. I paid $600 for mine with 8000 exposures.

    High ISO performance beats the D300S by a hair, which is likely what they wanted to do.

  2. This is gonna get the measurebaters into a knot. I wish someone would create a virus that destroys the computers of all measurebaters on the net.

  3. If you compare the analysis between the 7D, 50D and 5DMKII, in most results the 50D stands in between the 7D and the 5DmkII.

  4. They request that you read this page before making any comparisons:

    http://www.dxomark.com/index.php/eng/DxOMark-Sensor/DxOMark-metrics

    Basically, the score tells you how good a photo-site on camera A is compared to one on camera B. How many of them they have and how much better the signal processing and image processing makes them appear to be in the final image are not factors. It really makes one wonder what the point of it all is.

  5. This confirms what I’ve thought since the DPreview article came out:

    The 7D is clearly Canon’s best APS-C camera ever for both low-light and dynamic range.

    I really like that DxO removes resolution from the game & levels the playing field.

    Unfortunately it gets punished a bit on colour depth, and then there’s the pesky Nikon D90 with almost 200 ISO better low-light rating. But when you consider that you can print a 7D image 50% larger than a D90 image it more than evens out I think :-).

  6. Good. It used to say:

    “…In the case of the Canon EOS 7D, a camera that can perform quite admirably in the hands of pros and serious amateurs alike, the sensor format is APS-H, which is physically larger (1.3x) than an APS-C sensor.”

  7. Their DXOmark’s look rather neck and neck. Too bad the Sony A550 is such crappy body. :/

  8. I think it’d be fair to say that the 50D has the better overall measurements – albeit marginally.

    Dynamic range is a toss-up, with the 7D having better numbers at the low end, and the 50D at the higher end. Tonal range shows the 50D with a slight edge, aside from base ISO and ditto for SNR and color sensitivity.

    Are you sure you were looking at the right color in the chart?

  9. Rather I think it brings tears to the eyes of many 7D owners, particularly since the cheaper 50D is actually BETTER than the 7D

  10. A lot of you are missing the point. DxO only considers RAW sensor performance. It doesn’t take into account other camera features.

    Can the 50D AF like the 7D?
    Can the 50D fire away like the 7D?
    Can the 50D record video?
    Can the 50D withstand extreme environments like the 7D?

    Sure the 50D has a better SNR than the 7D but take a good look at DXO’s insight page. Based on their tests, higher resolution results in less noise when printing.

  11. Not to mention per-pixel-sharpness. The 7D is one of the worst cameras I’ve ever seen with regards to this.

  12. Just wondering,

    what the noise result be if 7D image is resized down to 12MP and compare it again with Nikon D300s?

    because the current noise image result is from 18MP image size. has anyone tried to resize the image and see??

  13. They sell for between $600 and $1100. There are a few who ask a couple of hundred more, but they do not sell.

  14. The 50D is better than the 7D??

    Better how?

    More Frames per second?, Better autofocus? Better high ISO Sensitivity? Better Weather Sealing? Better movie Mode?

    The answer to all those is NO.

    What planet do you live on?

  15. After reading all these comments it’s clear I should throw my 7D on ebay and go back to film on my EOS300. Thank heavens for detailed computer analysed tonal range measurements to set me straight.

  16. If I were buying an APS-C right now, the 7D would be it for sure.

    I have two itches: Video on DSLR, and FF. I love the video implementation on the 7D, but I have been itching to go to FF digital for a while.

    The 5D II video is a little awkward, and doesn’t support 24P yet. By the time the firmware for 24P comes out, I may be better off waiting for the 5D mk III. So I might get the 7D for the video.

    In any case, this is an incredibly exciting time for Photography and Video. There are so many excellent choices, I wish I was 30 years younger and just getting started. You young guys don’t know how lucky you are!

  17. Then again, it is 500 dollars cheaper, better at lower ISOs, doesn’t produce garbage 18MP files.. etc.. etc..

  18. What are you talking about? He’s referring directly to the dxo analysis.. God, is everyone here on crazy pills?

  19. Check the upper left hand corner…
    It is there if your click the PRINT versus the standard SCREEN button. To make the comparison fair, IMHO, it should be reverse…

    OH Well,

    the fact that it states the 7D shoots at 3 FPS shows how professional they are….

  20. Yup. Last year it was the 50D and now the 7D.

    Canon’s premium 1.6x models are getting worse and worse with each new model.

    Don’t know where Canon is headed with this.

  21. I can’t truly correlate the results of dxo lab results with real life. No pictures, no samples. Just a bunch of numbers. They do admit that they are not measuring the whole camera system, just the sensor. What’s the use of that?

    I still trust reviews like dpreview more, where there are real images to substantiate. We are photographers for goodness sake, what’s the use of a bunch of numbers without real images to back them up?

  22. Judging by the people bending over backwards to defend it, the people at the Canon sales department are earning their salaries.

  23. End image quality is what matters. DX0 is meaningless when it comes to picking which camera delivers the best picture, but it gives the Nikon fanbois a way to get some clueless Canon shooters all worked up into a conniption about noise and ISO performance.

  24. There’s no level playing field by removing resolution as a factor and not measuring in any way the final image produced by the cameras. The DX0 testing is completely meaningless in terms of rating which cameras DELIVER the best image to a file, unless you plan to pull the sensor out of the camera and place it on an altar of low noise worship or something.

  25. Yep, and being that no one can really do anything with raw sensor readings, which are not the same as the RAW image file which would at least give us something reasonable to consider in comparisons, it is pretty funny to watch all the whining and crying over DX0 ratings.

    I’d guess DX0 had to put the carefully worded warning that “DxOMark Sensor is NOT an evaluation of overall camera image quality or performance” after being contacted by some manufacturers legal dept. I wouldn’t be surprised if they get sued eventually just for creating the impression they are measuring overall image quality and the massive confusion and misunderstandings about how to pick the best camera they are creating.

  26. How about because they are not even testing the actual images produced by the cameras? They are not testing 12 MP vs. 18 MP files, so there’s no way to resize them to make them even, they only test the raw sensor data, not the final image, not even the raw image file.

  27. You clearly have not been saved by the DX0 messiah, removing the sensor from the camera, placing it on altar and worshipping it for its raw data characteristics! Get saved! DX0 loves you and has a wonderful plan for life.

  28. Charles Bronson on

    That was my first and last visit to the DXO website. What a bunch of who-cares-nonsense.

    I can’t even comprehend the point of some of the debates in this thread. 50D, 5DII, 7D, 500D, ect. whatever… just buy the one that suits you and stop stressing over some friggin graph. They all take great photos, get over it.

  29. check out the-digital-picture.com Strictly Canon reviews. Actual hands on review from an actual photographer (what a concept). He pretty much has a review for every lens that Canon makes (and a lot of discontinued ones as well) along with all of the DSLR bodies and a few accessories (flashes and such). That page is my first stop when researching any new purchases. That being said, dpreview is a great site and seems to be fairly unbiased.

  30. They clearly state what they test. I can’t see why they should be critized because som peolpe don’t bother to check or don’t understand the advantages and limitation of the test methodolgy.

  31. Well… simply this is silly and I’m really bored of all the so called pro reviewers, whether DPReview or DxO mark. Everytime I’m curious to see what the new posted review says but this time it really did it for me, enough is enough.

    I see photos taken by a camera, I love them, I compare them to other photos taken by another camera, and I feel it’s MUCH better, or even MARGINALLY better, all in all, BETTER, then I find the idiots at DxO saying it sucks and that D5000 scores higher

    To my eyes, the 7D PHOTOS I see anywhere are the best out of an APS-C camera, so to hell with all this grading and scoring. I never believed really that you can give a quantity to IQ. It’s something judged by the eye. I know many pros who still make good money using a 20D or a 350D which are old “low score” cameras.

  32. Its when you zoom in to 100% and think “Oh, photos in the 18MP sensor dont look as sharp as the photo from the 10MP sensor”, without looking at the overall result.

  33. I bet not a single ‘expert’ can explain this:

    Their data (viewed either at the pixel OR image level) most certainly indicate the 50D high ISO performance is nearly equal to that of the 7D. Yet, they assign the 7D a score of 809 for low light ISO and the 50D only 696.

    What’s going on here? Seems like they do not even trust their own analysis. :D

    I am now completely done with DXOMark… they are totally unreliable.

  34. Don’t forget foos, DXO list the D90 as better than the 5D at just about everything.

    7fps listed for the D300s at 14 bit? 3fps for 7D at 14 bit? Are those foos speakin’ jibba jabba? D300s only does 2fps @ 14 bit foo.

  35. I pity the DXO foos! How about using a 30 or 50MP print instead of 8. 8 just negates resolution, regardless of whether or not these foos think DR will improve.

  36. The criticism is because of their ranking system, which gives the impression of ranking cameras by image quality. People assume a Dx0 higher ranked “sensor” automatically means that will result in the better image recorded, which isn’t true. Do you see images of the sensors in their rankings? No. You see the entire camera, something they do intentionally to create the impression they are ranking the whole camera by overall image quality, an impression backed by little bar graph ranking from the sensor tests. It seems pretty clear from all the marketing stuff on their site they are trying to become the go-to site for people deciding which is the best camera, knowing full well people are going to look at where a camera is ranked and decide which camera to get. And I’m sure they know people tend to ignore the fine print clarifications that admit their rankings have nothing to do with actual end image quality of the cameras.

  37. They rank the Pentax K20D over the 7D too, even though the Pentax only scored 639 in low light ISO to the 7D’s 809 score. Also, Dx0 had to assign their own subjective ranking of the importance of Color Depth, Dynamic Range, and Low Light ISO to come up with their overall score, which if they gave a little more importantance to one of those factors it would resort all the rankings – oops. Who are they tell the world how important dyanmic range is vs. low light ISO or vs. color depth? It’s just their own subjective opinion that results in which cameras get ranked higher by the tests.

  38. You nailed it. And the reason Dx0 doesn’t match your own observations of actual photos is because Dx0 doesn’t test the actual images, not even the raw files. Their rankings are meaningless unless you pull out the sensor from the camera and just take the sensor home leaving the rest of the camera on the store counter with the packaging. Which I’m sure some data-freak Dx0 fanatic has already done.

  39. There is no doubt that DXO is measuring only the quality of the sensor. As long as readers understand that, they should have no problem.

    The sensor is where it all starts. If you do not have a quality sensor, your image will suffer, no matter how good all the other components are.

    Conversely, if you do have a quality Sensor, the image can still be messed up by the person taking the images, or by poor processing, or by a poor or defective lens, etc.

    There are so many variables that a comparison of complete camera ststems is unlikely to be accurate.

  40. There is a good reason NOT to test a complete camera system.

    There are too many other things that could affect the image, so you do not get as good of a idea as to how the sensor is doing.

    A complete system will be affected by differences and imperfections of lenses such as distortion, mis-focus, inaccuracy of aperture. And, of course, lenses from a different manufacturer such as Nikon, Sony, and Canon have substantial differences.

    Then you get differences in the processing of the image, some manufacturers add more NR for example.

    The list goes on and on.

    So if you want to test just the sensor, you eliminate as many of the variables as possible so your result is not distorted by other factors.

    There are lots of people who do tests of complete camera systems, and the results vary considerably. This is unavoidable.

    There are also those who claim to have special powers to view a printed image and declare which camera is best. What they are really doing is rating images that are like the ones they are use to producing as the best, so naturally, Canon owners rate Canon best, Nikon owners rate Nikon best, and so on. It merely rates personal preference.

  41. 7D 50D terrible sharpness. why people are still saying 50D and 7D are great? yes they have advance technology but what we all want is QUALITY IMAGE not much how advance is ur technology is.

  42. well, that’s the point though, in reality deciding what is the best camera is not as simple as looking at raw sensor data. Dx0 rankings also have a subjective aspect in that they decide what values to ascribe to the different catagories tested that produce the overall rankings. A dynamic range rated at 12.0 verses 12.2 is not likely going to ever be noticed by a photographer in the real world. The ISO ratings are somewhat useful, but Dx0 gives much more value to color depth and DR, so a .1 difference in dynamic range can outway a 300 point difference in low light ISO, according to them – as that determination of value is THEIR OPINION. So the rankings are flawed even from a raw data standpoint.

  43. The problem with Dx0 is not their raw data testing, but in the subjective values they give to the different categories that result in the rankings, and they do not tell us to what degree that raw sensor quality gets degraded by other components/processors in the camera. So what if one camera has slightly better dynamic range in raw sensor testing if that gets reduced in the final image file! It’s really meaningless rankings without the meaningful evaluation of end image quality to compare it to.

  44. Why get so over-heated by all this?
    Look at it for what it’s worth.

    I love my 40D, but if I was buying APS-C right now I’d buy the Canon 7D. And I still might get one, not for the extra pixels, but for the HD video, and the fast performance is nice too.

    But at the end of the day, I want to go FF for DOF flexibility, and the slightly richer color.

    I’m still on the fence with the 7D, not because of image quality, but because of my desire to go FF.

  45. True, one can only research so far into a product. Graphs are an easy way to show a lot of information that is seemingly clear and straight to the point, but where exactly is the information pointing? I love reading about photography, but I also get hooked into just taking a graph at face value without truly understanding what the interpretation of the graph actually is and how does the graph related to real world shooting. The Dx0 doesn’t mean a dang thing if the images suck.
    Nikon is smart to play the game the way they are. But on the other hand Canon plays the game by increasing pixels and hoping that this will be the primary factor that the typical consumer looks at when making a purchase. Higher megapixels is one of the easiest things to look at and understand that higher is always better right? Clearly not, especially if you go into deeper studies of today’s digital photography. But that takes time.
    Final message……I am so freaking excited for the 1D mark IV! Whew-hoo!

  46. Thing is some people in forums keep saying, like you, that all Canon is doing is increasing megapixels, and that’s just not accurate. They are increasing megapixels while improving noise and low light ISO ability. There are of course people who are doing nothing but shoot for web display and overly magnifying images to look at pixels, and therefore feel they don’t need anything over 10-12 MP in a camera. That’s fine for them, but there are still plenty of photographers want to make large prints, posters, things people can mount and display on a wall, not just pixels on computer monitor, and thankfully Canon keeps that in mind in their design perameters.

  47. Surely, I’m glad that each company is not exactly the same. With the differences in philosophy/marketing strategy we the consumer have more options and can choose the product that suites us the best. I’m glad for the increase in megapixels with the additional improvements in noise compared to all earlier cameras. I’ve been having many more customers interested in 20x30s due to the drop in large print prices and seldomly can I get a shot that doesn’t need some cropping to reposition and zoom in. I’m pushing a 20×30 when I crop a little with my 5D. I’m trilled about 3 more megapixels to work with on the 1D mark IV. For pure web, go Nikon, for real prints i’m happy for a bit of extra megapixels, absolutely.

  48. kid stop being naive and behave in such childish manner. I would prefer you to examine the facts and debate back rather than “SHUT UP FOO!” wow how persuade i am.

  49. imagine if there were muliple versions of 7D: 10MP-7D, 12MP-7D, 15MP-7D, 18MP-7D; which one would you buy?

    in reality there is only an 18MP-7D, anyone who would rather buy one of the other lower MP versions but still buys the 18MP because s/he wants HD video or any of the other features of 7D is just lured by Canon to buy a camera that has one of its fundamental specs (megapixels) different from what s/he thinks is ideal.

  50. DXOmark is the best estimation available for sensor DR/color/ISO-noise potential, and canon 1.6x cameras score bad in “color depth” compared to APS-C of all other brands (Nikon, Sony, Pentax), and that may have to do with 1.6x being slightly smaller than 1.5x.

  51. um, your darwin reference tested a pre-production 7D, clearly the raw decoding had issues back then, and testing was done by a Nikon shooter…Try reading a review with some credibility:

    “The EOS 7D delivers impressive image quality across the sensitivity range.”

    “Class-leading detail and resolution at base ISO, good per-pixel sharpness.”

    “Very good low-light performance, low noise levels and good retention of detail.”

    http://www.dpreview.com/reviews/canoneos7d/page30.asp

  52. Or maybe people are lured to the 7D because it delivers great photos…with impressive image quality, sharpness, detail, and very good low light low noise ability?

    “The EOS 7D delivers impressive image quality across the sensitivity range.”

    “Class-leading detail and resolution at base ISO, good per-pixel sharpness.”

    “Very good low-light performance, low noise levels and good retention of detail.”

    http://www.dpreview.com/reviews/canoneos7d/page30.asp

    DOH!!!

  53. Stop getting your info from posers in forums who get off bashing cameras they only wish they could own and try reading a creible review:

    “The EOS 7D delivers impressive image quality across the sensitivity range.”

    “Class-leading detail and resolution at base ISO, good per-pixel sharpness.”

    “Very good low-light performance, low noise levels and good retention of detail.”

    http://www.dpreview.com/reviews/canoneos7d/page30.asp

  54. Holy Crap! That’s awesome! I hope that I don’t regret not keeping my 5D2 when opting for a 1d mark IV. Do you think that the noise should be lower in the MIV compared to the 5D2? I know it’s FF vs. 1.3 crop. Isn’t the Pixel density similar? Perhaps newer technology smaller gaps in MIV?

  55. I don’t know but I just think that 18 mp on a 1.6 crop doesn’t seem right picture quality may be affected. Oh well I will just have to wait a fried to buy one and test it my self.

  56. There’s a lot of factors to look at beyond raw sensor data scores, such as is it recording in 12 bit or 14 bit. The sensor on the D300s might show slightly better color depth rating 22.5 compared to the 7D’s 21.9, but the D300s drops down to 2.5 fps when shooting in 14 bit, where the 7D is 8 fps in 14 bit, and doesn’t need an extra battery grip to do it. So if one is buying a D300s thinking they’re getting all that color depth and 8 fps with the extra battery grip, oops, sorry, no. You have to drop it to 12 bit to get 6 fps or 8 fps with the grip…so much for the slightly higher color depth rating from Dx0! The 7D is better despite the higher ranking Dx0 misleadingly gives the D300s.

  57. Is not that tech is improving or not. It’s yes technology is improvig but in this case it’s more like physics and logic. If it is that size you, can’t go any further. E.g. if ur cup is full u can’t possible fill any further. Also, I Would find blog’s review more wordthy than offical review websites like dp. Review websites could be part of marketing purpose. Canon could have paid them to do a whole review and write up a good word about it. Just my opinion.

  58. In the first place, how reliable is DXOMark when they do not trust their very own data? Despite the 50D scoring as good as, if not better than the 7D in their various measurements (at both pixel and image level), it only received a score of 696 for low light ISO as compared to the 7D which received a score of 809. How’s that for reliability? Not a single expert who trusts DXOMark can explain this.

    Sorry, I am completely done with DXOMark… they have zero credibility. All their ‘carefully’ prepared data tell a different story than everyone else : DPReview, Imaging Resource, the-digital-picture etc etc.

  59. And the joke is that they take themselves so seriously. While DPReview shows pics that show the heavy-handed noise reduction on the Sony A550 RAW files, DXOMark happily post their results without one word on the awful effects of Sony’s on-board noise reduction schemes.

    DXOMark will like to think they are posting results on sensor performances when they are in fact putting up fake data.

  60. If DP is paid by Canon why do they give Nikon cameras good reviews too? Your view makes no sense beyond you are just believing what you want to believe regardless of the facts. What blog are you believing then? One that tested a pre-production model?

    So how about photographyblog.com’s review of the 7D?

    “the 7D is a fantastic APS-C DSLR”

    “You may think that bumping the megapixel count to 18 on such a comparatively small sensor would only degrade the resulting photos, but thankfully you’d be wrong.”

  61. I don’t know if the data is faked, but there is a big subjective factor to the rankings, and imo their rankings are misleading and meaningless in terms of deciding what is the best camera to get.

  62. I agree, how can any of their tests and conclusions ever be trusted if they don’t correctly specify its frame rate in continuous drive.

    :rolleyes:

  63. Hit “Print” when looking at one of the comparison charts, now you’re looking at resolution normalised data.

    Whoops!

  64. well, it is worth it when there’s a camera that’s very similar to another one, and you need to decide between the both of them.

  65. let’s make this the longest thread everrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr

  66. maybe they are paid by amazon, who actually bought DPR few years ago and since then the site has changed a lot, i dont see any objectivity in giving sony a550 “Recommended”, while giving D3000 “Highly Recommended”, the latter is lacking in so many important areas. You take DP as a bible, that’s fine, but don’t impose it as such on others. personally I look at image quality from many sources, including downloading RAW and developing it myself to make my own judgement of image quality. diversity of sources reduces the impact of all sorts of errors like bias, human/reviewer errors, camera copy variation, or lens/SW processing/weather affecting qulity of sample images…etc. My conclusion about the 7D after seeing so many images and reviews on different sites + some 1st hand experience is that the image quality is a bit lacking in: 1)color richness 2) dynamic range (specially obvious under bright daylight with deep shadow conditions) 3) shadow noise even in base iso. all this could have been avoided if canon chose a more sensible MP count. no need to get too argumentative or defensive of Canon on that uncalled for MP increase, i will not respond to replies of that sort.

  67. you know, technically, from its audio measurements, every audio mixing board from the 70’s was complete crap.

    therefore, why do you bother listening to songs from the 70’s?

    just go to http://pbase.com/cameras and look at the results yourself. that’s the only test that matters. nobody looking at your end results in a magazine or the internet cares what brand of camera you were using.

  68. Umm, how can the rankings be subjective when they are all based on numbers that come out of a machine?

    DxO have a machine that performs various tests to measure the way a digital camera sensor performs. Why would they want to skew their results one way or the other in favour of any particular vendor?

    I’m pretty sure that if Canon cameras were producing better numbers than Nikon, then the discussion here would be completely different.

    I’m sure they would rather press their case like this to licence/sell their equipment to Canon, etc, rather than appease crowds here.

  69. Gusto wrote: “Despite the 50D scoring as good as, if not better than the 7D in their various measurements (at both pixel and image level), it only received a score of 696 for low light ISO as compared to the 7D which received a score of 809”.

    go read their methodology before making such UNINFORMED comments. They have nomalized and non-normalized comparsions and that confuses many people like you, but they make very clear which is which. The first tab with four scores (DxO sensor mark, color depth, dynamic range, hi ISO noise) is normalized, ie based on image-to-image comparison. All or most of the other tabs are based on pixel-to-pixel comparisons, lower MP cameras like 50D (relative to 7D) may win the pixel-to-pixel comparisons but lose when you normalize by reducing the images from the two cameras to a the same megapixel size (whatever that size is), the point is if shoot under identical shooting settings by 50D & 7D the same scene, the 7D image is linerly “magnified” 1.09x compared to 50D due to 18 vs 15MP, this magnification factor magnifies noise in the 7D image among other things as long as you are looking at pixel level of 15&18MP images, it is unfair comparison so just consider the 4 scores of the first tab, 7D is slightly better than 50D.

  70. “Umm, how can the rankings be subjective when they are all based on numbers that come out of a machine?”

    Good point.

    Now explain this to me: despite the 50D scoring as good as, if not better than the 7D in their various measurements (at both pixel and image level), it only received a score of 696 for low light ISO as compared to the 7D which received a score of 809.

    How’s that for objectivity in the scores?

  71. You failed to understand the difference between the two, otherwise you wouldn’t need to exclaim or use any question marks as you did above in the post i was responding to.

  72. Do you know you can choose between the screen (pixel level) and print (image level at 8MP) tabs for each subheading (SNR18%, dynamic range etc etc)?

    You yourself said: “7D is slightly better than 50D.”

    The key word here is SLIGHTLY.

    But a score difference of 809 vs 696 is HUGE. And THAT is something you cannot justify with the data they have.

    I do not understand why people worship DXOMark as though it is the only metric to evaluate camera image capability. Some even get all defensive about it.

  73. Has anyone seen the just released 7D test results from Popular Photography (January edition)? According to their tests, the 40D focuses faster in low light (and in good light) and the 40D has less noise then the 7D for ISO 1600 and lower. I believe this is one of those “per pixel” tests. Therefore, the net noise on a printed image is left up to our interpretations. Even though all these tests “might” not mean too much in the final printed image, they are highly regarded tests that affect some peoples view of the 7D…and they certainly have gained our interests.

  74. I never said DP is a bible, although can you show me another 31 page review of the 7D or another 27 page review of the D3000? Even if Amazon owns them, where’s the bias? They point out pros and cons on all the cameras they test and do a very detailed review including looking at RAW & JPG output.

    I also never said I only read DP reviews, I read lots of reviews, talk to friends and other pros using it, and then shoot with it myself, all of which I’ve done with the 7D, including comparing it to my 5D2 raw files, and I do not agree with you, or you megapixel bashing bias.

  75. Slightly better overall DxOMark (64.9 vs 62.9) but not in low-light ISO (809 vs 696)… and don’t take my own arguments and use them against me, i don’t worship or bible any site on the contrary as i said above i look at many sources including 1st hand experience and make my own judgment, what is special about DxO is that for many people like me, we find their results to confirm our judgment (which we always form before they publish their results about a given new camera)… anyway you mentioned imaging resource, here is a 50D-vs-7D ISO noise comparison based on the their test shots, in this case it is image-to-image comparison and it is very clear how the difference is considerable not slight in higher ISO settings, no argument about that except from people with whom discussion is pointless.

    ttp://forums.dpreview.com/forums/read.asp?forum=1019&message=32879822

  76. That’s right, don’t respond to this Catastrophile, your post shows your bias when you say the difference between 22.5 vs. 21.9 color depth or 12.2 vs. 11.7 is Canon scoring bad, yet you know very well you would never be able to see that difference in a RAW file comparison, and you don’t want to consider the fact Nikon makes you drop down to 12 bit shooting to get more than 2.5 fps out of the camera. Why don’t you compare a 12 bit raw file to a 14 bit file when you bash Canon and favor Nikon? It’s a much bigger difference than 22.5 and 21.9.

    OK, so you tell us, is 12 bit Nikon’s 8 FPS as good as Canon’s 14 bit 8 FPS:

    …………………..12-bit…14-bit

    Brightest stop………2,048….8,192

    Second brightest stop..1,024….4,096

    Third stop………….512……2,048

    Fourth stop…………256……1,024

    Fifth stop………….128……512

    Everything else……..128……512

    Total distinct values..4,096….16,384

    Seems 14 bit is WAY FRIGGIN BETTER THAN 12 bit. So much for Dx0’s raw sensor testing and camera ranking if you want to shoot at more than 2.5 fps!!! HAHAHAHAHAHA

  77. >> imagine if there were muliple versions of 7D: 10MP-7D, 12MP-7D, 15MP-7D, 18MP-7D

    Yes, that would ideal.

    Maybe not four different versions but just two: 12MP-7D and 18MP-7D.

    Of course the assumption here is that the 12MP-7D will have better ISO than 18MP-7D.

    In that case I’ll surely buy the 12MP-7D.

    Right now I’m not buying a 7D for sure – despite the fact that I like so much the 7D’s advanced AF-system, 100% viewfinder, and advanced metering system.

  78. David, by repeating the same quote over and over you are not really convincing anyone – just embarrasing yourself.

  79. >> that may have to do with 1.6x being slightly smaller than 1.5x.

    No.

    The 7D’s bad color sensitivity comes from a trick that Canon is applying on the sensor.

    To combat noise, Canon made the 7D sensor with ‘unequal green response’.

    What that means is that one of the green channels on the Bayer filter is diluted, so that it allows not just green light to hit the sensor but light from other wavelengths as well.

    The overall result of this trick is that more light hits the sensor – hence, noise is minimized.

    On the flip side, the sensor is more color blind, as confirmed by DxO.

    The so called ‘mazing’ effect seen on 7D images is also result from the ‘unequal green response’.

    Of course you see the mazing only with RAW converters that are not coded to compensate for that.

    The Adobe converter had the issue but they already fixed it.

  80. Hey if you don’t like what legit reviews say about the 7D that’s your problem. It’s funny how some people just want to dislike the 7D because of their stoggy outdated nikon propaganda opinions that it has too many megapixels, but again, that’s your problem not mine.

  81. The lower the number, the better the camera. That’s why the 1 is the only way to go. 7 is higher than 50. 7 is better than 50. No matter what all these stupid reviewers say.

  82. I actually did mean the 40D. I was looking at the old Pop Photo test of the 40D and compared the numbers. With all the hoopla about the new focus system, the 40D still focuses faster…at least for static objects.

  83. I think we agree the 7D is really MUCH better than the 50D at high ISO. And our observations are confirmed by Imaging Resource, DPReview, the-digital-picture etc etc.

    The issue is that the DXOMark data under both ‘screen’ and ‘print’ tabs do NOT agree at all with our observations. That is my problem with DXOMark.

    Another example: compare DXOMark data of D200 vs 40D. In this case, their actual data make sense BUT their assigned scores do NOT. This is just the opposite of what we’re seeing in the case of 50D vs 7D.

    Now, do you see what I mean?

  84. “dxo doesnt like canon too much…”

    Apparently it is Canon, not DxO.
    Nobody expects better signal to noise ratio from smaller pixels.

  85. Well, in DxO, there must be some truth in it. Have you seen any quality jump in compacts lately with small sensors?

  86. How is it that so many reviews can find fault with such a great camera? The 7D has shown digital noise can be beautiful. Pattern noise was the enemy in previous APC-C cameras, and the 7D goes a long way to solving the biggest issue I had with digital. Again, looking at the whole picture easily shows the bigger picture, noise at the per pixel level is becoming meaningless to analyse.

    Thats something these pixel peeps forget (they also forget to use a good raw converter). Look at images at 1:1 without conversion and they are positively nasty looking even at low ISO. Until you use the RAW converter.

    Ive shot quite a few great shots at ISO 3200, and in many cases the noise is improving the photo, not detracting from it.

    Maybe some of you guys should rent a 7D if youre hesitating, seriously you’ll be blown away by real world photos.

    By the next decade we will be at 100MP, approaching photo sites sensitive to single photons.. And there will be the same peeps complaining about “why doesnt Canon release a 50MP version” or “Diffraction ate my dog”.

  87. I’m not saying the raw sensor data from Dx0 is not accurate, just that their ranking whole cameras based on just those sensor readings is unfair and misleading because it doesn’t evaluate the raw image file created or factor in things like 12 bit, 14 bit, or 24 bit images, or image size – so any ranking of cameras based on just Dx0’s measurements is meaningless as to which camera is better than another.

    A clear example is the D300s and the 7D, where they rate the D300s over the 7D, but they don’t tell you the D300s can only record 12 bit images to get 6-8fps, and 14 bit is only 2.5 fps, unlike the 7D’s 14 bit 8 fps ability. In fact, pull up a comparison between the 7D and the D300s in Dx0mark and it shows both cameras at 14 bit, but the fps of the 7D as 3 fps and the fps of the D300s as 7 fps, both are wrong. The D300s is only 2.5 fps in 14 bit mode where the 7D is 8 fps. The 7D even bests the D300s in low light ISO performance. They rank the D300s as #19, and the 7D way down at #31, even though the 7D is arguably better at anything over 2.5 fps, and produces a bigger image. It’s this sort of ranking with incorrect and misleading information that shows their clear anti-Canon bias.

  88. Also, their silly print/screen tabs, where they “normalize” the image down to an 8 MP size image, is not the same as looking at the actual raw file or making prints from it. I certainly do not trust their method of “normalizing.”

  89. There’s no whoops about it. Those “print” tabs are after they “normalize” the image down to 8 MP, whch is complete and total BS designed to mislead and trick the viewer into believing their rankings rather than looking at the actual RAW files and images made from them with the correct and up to date softer from the manufacturer, not Adobe who often tkaes a long time to get their raw decoding correct.

  90. If you will shoot all your shots from all cameras you use and have them “normalized” to 8 MP by Dx0 labs before you do anything with them for the rest of your life, then I will take it back. Of course only a fool would do such a thing, and it’s no different to consider such a process to accurately represent the quality capable from a camera’s RAW file.

    RAW files have to be decoded properly to get proper results, not “normalized” by some third party. This is why some of the blogs that test RAW files from pre-production cameras usually show crappy results, and credible review sites wait until the full release firmware AND RAW file decoding software are released, and typically that requires using the manufacturers software, not Adobe or some programming Dx0 is using, or the results are not accurate. Even Adobe RAW is known to produce crappy results at times compared to the manufacturer’s RAW software, so it can certainly effect the outcome on a test, which makes Dx0’s “print” button complete and total misleading BS, just like their camera ranking system.

  91. yes seriously u kept on spamming everyone about the reviews and stuff. since it’s a pre-production model when did u get information on now the current ones are still not the pre ones? how accurate is that? I’m trying to be MP Bias here it’s common sense IF it’s full u cant fill it anymore.

  92. Thx Karl, I had forgotten about the noise pattern issues.

    I’m still on the fence with the 7D. I have the 40D now. I’m sure I’d love the 7D, but I had been planning to go FF with my next camera. Now the 7D has me second guessing myself. I love the video implementation on the 7D, and the high speed performance, but I would also like to have the DOF options that the FF gives.

    I have decent crop lenses: canon 17-55 2.8IS, tokina 50-135 2.8 (works at 135 on FF), and I just bought the tokina 11-16 2.8 (couldn’t resist since I’ve heard it’s very good and will work on FF). The fact that I have these lenses is also making me consider the 7D. But I still can’t get FF out of my mind.

  93. Im rotten at taking video. However I find myself still taking a huge amount of it, and this is a case of the camera making me look better than I am.

    Anyone who whines about 18MP RAW wasting storage space hasn’t had to deal with terabytes of video :)

    The 17-55 2.8 IS & 100 2.8L Macro IS are the two lenses I have which work great for video. The IS is particularly helpful, and I can get decent hand held footage. Probably a video tripod head & monopod would help alot, but ive already spent enough on the camera.

    Youve invested in the 17-55 as well. And its my equal favorite lens with the 100mm Macro for the 7D.

    Personally ive not used FF, so cannot really comment on its DOF, however I do find that at 2.8 the depth of field on the 18-55 is already so narrow to be sufficient for my own needs.

    Regarding DOF & Video, if you havent used video on a DSLR, you should give it a try, you have to manually focus. If you combine that with narrow depth of field, you have to become REALLY, REALLY good at chasing & anticipating moving subjects (or not change focus at all). I would have chucked out half my video if it wasn’t of loved ones… I find this even more difficult using liveview while recording during harsh daylight.

    Generally I use F 1:4.0 or 1:5.6 and still achieve very nice effect for isolating human subjects. One also has to consider that usually a shutter speed of between 30-100 at 24p is the most pleasant for video, and so during daylight I find myself either going for narrower apertures anyway, or getting out an ND filter.

  94. You’re using circular logic now (“it’s crap because it is”). Would you mind addressing the question?

  95. They are changing the file and therefore not testing the actual image capability of the camera. It’s very simple. You don’t get to convert a 18 MP RAW file image to an 8 MP “normalized” file and then say you are testing the RAW file capability of the camera. If you don’t understand the importance of proper RAW file decoding to evaluate the camera’s imaging ability there’s nothing more to say about it. They might as well be comparing low rez JPEGS, it’s just as meaningless as downgrading the file 8 MP and changing it to what they consider to be “normalized.”

  96. I’d like the option for shallower DOF for photography. There often seems to be ugly backgrounds when I want to capture a photo of people, so I want to blur the background to oblivion.

    I figure the 7D for video would be better than the 5D II, but I’d prefer to have both in one camera (shallow DOF, and good video implementation). One option I’m considering is getting a decent video camera, and keep using my 40D until I see what the next 5D (III) has to offer.

    I am totally new to video, but the 7D is tempting me.

  97. Clear explanation.

    Something I don’t quite understand.

    Is this color blind sensor also more susceptible to chroma noise? I think that must be the reason Canon 50D/7D RAW files appear to have higher chroma noise than those from the D300s/D90/D5000.

    This is what I understand so far about Nikon’s sensor design strategy. Let me know if I’ve got it right:

    – chromatically accurate filters to reduce chroma noise

    – 100% microlens coverage to improve light gathering capabilities (Canon has this now)

    – Sony A/D architecture which allows for higher dynamic range at base ISO. Drawbacks include slightly poorer high ISO dynamic range (e.g. in D300s) as well as limited video (e.g. Sony DSLRs cannot record video, or only 720p in Nikon DSLRs) and 14-bit capabilities (e.g. dramatic drop in D300s frame rate at 14 bit)

    – clip the black levels slightly to improve dynamic range numbers (all current Nikon DSLRs do that). Who cares about astro-photography? :)

    – use on-board noise-reduction schemes on RAW files (e.g. D90/D5000). Makes drastic improvements in dynamic range of D90/D5000. The drawback is the slow down of operation speed and this is why it is not implemented in the D300s.

  98. yes i do know. a man who behave like a kid! and has no sense of manners quote “SHUT UP FOO!”- Mr T

  99. You showed us all again, you don’t know who Mr. T is. Do yourself a favour and go hire The A Team.

    Only they can help you from the jibba jabba!

  100. i couldn’t be bothered who you are. you think your some kind high and mighty LION? no sense of manners and trying to boast. like I’m so free to look for “The A Team”. What sort of argument is this? not even clear referring to what?

  101. wow im like so scared~~ what cya gonna do? take my camera away? hahaha do i look like i care who the hell are you? haha save it for some kid. Kid!

  102. “- 100% microlens coverage” – not true. Should be “gapless”. Microlenses are round so in gapless arrangement they cover PI/4 of total area, i.e. 78.5%.

  103. Shoot them both, decide yourself, send one back.

    Don’t be a measurebater.. They are sad.. Very very sad.

  104. LOL!
    nah wait for 5D mark III and get familiar with G11 1st :)
    I have the G10 and it’s the closest thing to SLR (should you consider upgrading)and its been great! (for me anyway).

    though G10 is not as flexible (in operations) as my 450D… what bugs me is that my 450D doesnt have video and the G10 has! Vid Q is quite good but you will have problem with isolating a subject due to massive dof in p&s camera.

    Anyway, I recommend G11 should you want to get a versatile camera with easy operations that is similar to a DSLR… and save up to buy 5D mark III and a nice lens in the future:)

    PS. DSLR will still be superior in terms of noise… my G10 (14.6 MP) is unusable past iso 400. The reduced MP count in G11 should improve its ISO performance but should still trails that of DSLR ISO performance by miles. So, if you can put up with noise at high ISO… its a good buy In my opinion.

    Why not S90? because its lens are slow at the max optical magnification, f/2.0 doesnt really add to dof in that focal length, too thin and uncomfy for my hand… I was considering these two to replace my p&s. But I decided to save up for a 17-55 IS or 24-70 IS XD and decided to stick with my G10 and trusty 430 EX II for everyday use :)

    Happy deciding, RichardR!

  105. Is there a good comparison between 7D and 5D mark II?
    I have never used 7D personally… so, If I cant wait for a 5D mark III and decided to to opt for the 7D instead… will it be a worthy purchase?

    Now, all my friends are using 5D mark II and the results (for assignments) were really great… so I’m really torn about IQ between these two cams.

    All in all… I will stick with my 450D and save for a nice lens haha.

    But I’m really curious about the pic quality… should there is a comparison please advice me as all I can see is that people tend to just saying FF is better without a legitimate proof. (I;m looking for dynamic range and harshness between shadows and lights, not sharpness… of course the 5D II will be sharper :P)

    Concur with CB! my 450D takes great pics too :) (though my friend 5D II takes great pics more XD)

  106. If you’re completely new to video you might find that the video capabilities of DSLRs frustrating to use.

  107. Lol.

    Even my 40D smokes the D300s in terms of RAW shooting speed @ 14-bit. :) At least my 40D can sustain 6.5fps (6.3fps whatever) when shooting RAW and always in 14-bit.

  108. Does the time it takes for a car to go from 0-100km/h (or 0-60mph) have any bearing on how long it takes to go from 100km/h-0 (or 60mph-0)?

    No.

    What does this mean in this instance?

    Each measurement that DxO do is a measurement of something specific and how a camera performs in one does not necessarily relate to how it performs in another.

    The 50D results seem suspect because if you compare them with the 7D, it looks like ISO100 is missing from all of the tests. They’ve already corrected one goofup with the 50D, maybe another is in order.

  109. What surprises me is how many people comment on the camera without even owning or at least renting it. I went out and upgraded from an XSI. Many people say the xsi is one of the best for dollar purchases from canon, especially since its at 12mp which for some reason people think is the holy grail of mp counts. However, why do we pixel peep? If your shooting wild life (for cropping purposes) or your printing billboard sizes i could understand. Beyond that whats the point? I print max 8*11 and have beautiful prints from the 7D. Go out and try it before allowing others to mold your opinion.

Leave A Reply