is to generate revenue from the number of hits to his website. The more provocative his language & analysis, the more hits he gets & the more money he has for the next camera purchase. I think the hyperbolic language & tongue in cheek analysis is quite deliberate.Some of his testing is actually quite informative (see some of his lens comparatives...) but when he tests & compares camera bodies, he limits his comparatives to jpeg defaults. His resulting bold and sweeping conclusions from a single example only serves his greater interest of generating controversy, hits & more revenue. This is not a serious test by any measure & can be dismissed as inconclusive at best.No doubt, the D3x is a great camera, but this is not a definitive comparative. Why not test it against the 1DsIII; the real rival to the D3x?I suggest that we wait for a serious review & avoid feeding the hit count on KR's site.-evan

Why is the 5D2/DIGIC4 robbing so much detail from the image at the base ISO of 100?It doesn't and in fact his results seem to differ from just about any other test done. Take a look at the comparisons on Imaging Resource, Club Alpha or Luminous Landscape and many others posted. There is little or no difference in resolving power of these sensors. In fact there is now so little difference that it hardly matters at all at the print level.So what went wrong for Ken Rockwell? Focus issue, lens issue, in camera settings? Who knows? But in any case the results are not consistent with almost any other findings. Is he biased? Maybe, but I will leave that to others...

I guess this time it's Nikon who clicked the donation button on his page

noctwicewrote:Why is the 5D2/DIGIC4 robbing so much detail from the image at the base ISO of 100?It doesn't and in fact his results seem to differ from just about any other test done. Take a look at the comparisons on Imaging Resource, Club Alpha or Luminous Landscape and many others posted. There is little or no difference in resolving power of these sensors. In fact there is now so little difference that it hardly matters at all at the print level.I think you didn't compare the two jpegs at ISO 100 from D3x and 5DII from IR carefully. Check the fabrics, especially the red colored one, but also the gray, the gree and the black ones. The effect is not as harsh as in these KR shots, but the blurring of the red fabric as compared to the D3x or even to the old 5D is quite visible.Foliage and distant flowers are a very though subject, likely the worse for a camera to give good rendering, so the effect may have just been enhanced. Now, from some RAW comparison, it seems the two cameras are pretty close in overall IQ. So it must be the jpeg engine, at least in part (KR may have screwed somehow the shooting, but I think he, like anyone shooting a tripod test, has taken a series of shots to guarantee no shake or focus issue was present).

Randfeewrote:see here: http://www.kenrockwell.com/nikon/d3/sharpness-1.htmnow, 1 year later that Nikon has the edge in this wannabe test he says: The Nikon looks great shot as JPG. The Canon looks crappyThat was there when he said the 5D was sharper than the D3, and it's still there now. I don't see the issue.

... this guy is obviously biased in nikons favor. There is nothing wrong with that, but he should notice himself doing that and mention itHe played the advantage of the 5D vs. D3 down to being "unimportant" for photographers, since those would never care about such tiny difference. Now, that there is a Nikon that wins, he exaggerates the other way and claims, that it is a "clear advantage".... thus: Biased

Barnettwrote:I see lots of halos. The Nikon image clearly has had more sharpening than the Canon image. See my post below for what the Canon image looks like after sharpening.While I see what you are talking about, I also think you inadvertantly proved my point. You had to zoom into like 800% to see that. I am used to seeing halos on Canon files at only 100%...

p0tempkinwrote:...in fact, he's been pretty consistent about his views on photography.LOL - this is the same guy who said Lens sharpness does not matter anyway but went on to compare sharpness of several lenses/cameras. http://www.kenrockwell.com/tech/lens-sharpness.htmHe prefers shooting in-camera JPEG. This was true years ago, and it's still true today.OK to shoot jpeg at least tell us what in-camera settings were being used? But no he did not and EXIF data were stripped. Quenstionable.Here's the MILLION DOLLAR QUESTION:Why is the 5D2/DIGIC4 robbing so much detail from the image at the base ISO of 100?Thereshouldbe very little noise in the first place. Is Canon trying to steal Sony's crown of having the worst in-camera JPEG processing? In their $2700 semi-professional DSLR?The fact is that the 5D2 is a double-edged sword. DIGIC4 gave it great high ISO image quality, but they had to sacrifice a good amount of detail in the process.There are tonnes of ISO 100 samples out there showing incredible details but you seem to believe only this particular unsubstantiated sample.Also, what is the aperture used? Did he shoot wide open with the canon gear and expose the lens weaknesses so that readers think it is the 5D2's fault? And why compare unscaled image (108%) with 100% crop? Very questionable.As for the D3x being better, it shouldn't be a surprise. It's $8000. Of course it'll be better. Canon is fine right now, but they might need to rethink their priorities once the D700x (aka D900) gets announced. When Nikon crams the D3x sensor into a D700 body in late 2009, the 5D2 (and it's 3-year-old AF system) is finished.That's if they can beat the 5D2 price..Cheers!

I can buy three 5D2's for the price of one D3x. If the differences in image quality is barely noticeable, I don't see the justification of spending that much money unless you are looking for more fps. I'm sure the D3x is an exellent camera but for that price I would not buy it. --http://www.pbase.com I exist, therefore I take pictures.

Considering that Ken Rockwell is shooting in JPEG, it stands to reason that he had the camera at Full Auto and the picture might well have been taken wide open. I own this lens and stays on my camera longer than any other lens, but I do know it's limits.Anyway, it's plain absurd and stupid to compare 2 cameras based on 1 image and 1 opinion, unless that opinion is your own and you can back it up with a solid body of evidence.Grow up people!SK

I understand his point about the variety of RAW converters but comparing two JPEGs straight from the camera is more a judgment of respective image processing decisions on the part of the manufacturer. His 'pals' might all be pros who shoot JPEGs but, outside of journos, I don't know any.The conclusion is that Canon's default JPEG rendering is rubbish, which is... "uhh, ok" but says nothing about image quality.Calling the Nikon a more professional camera is also a no-brainer, it's right there in the price. If the only advantage the Nikon had going for it was less default JPEG NR that would be worth complaining about.

...as I own the 1Ds3 which has a very, very similar sensor I can comfortably say with a Canon 50mm F1.4 lens at optimum aperture I can get an image better than even his D3X. He clearly botched up the 5DII shot. -- Waldo Nell

Are you suggesting that Rockwell accepted a bribe to come up with his conclusions? I suggest you think again and desist from making slanderous statements to the world at large. If nothing else, Rockwell speaks his mind without fear or favour, and in this case based on his own testing. I suggest you examine your own dubious motivations for suggesting such a thing. Tony

Why is he comparing cameras that are not in the same class so to speak? Can someone help me to understand this? 5D2: $2700, D3x $8000. Shouldn't the 5D2 be compared to the D 700? --http://www.pbase.com I exist, therefore I take pictures.

kevm14wrote:Barnettwrote:I see lots of halos. The Nikon image clearly has had more sharpening than the Canon image. See my post below for what the Canon image looks like after sharpening.While I see what you are talking about, I also think you inadvertantly proved my point. You had to zoom into like 800% to see that. I am used to seeing halos on Canon files at only 100%...No, I could see quite clearly by looking at the 100% images that the Canon image had much less sharpening than the Nikon image and furthermore that the Nikon image was a bit over sharpened. The 400% crop was only for your benefit.Barnett

Rockwell says:"Each image came directly from each camera. I did not use any external raw processing software, as each and every piece of software (Adobe Camera Raw, Phase One, Capture NX, etc. etc.) uses different processing."He does not know what he is talking about, otherwise he would know that JPEGs out of the cameras apply different processing as well, each company has their own processing preferences, Nikons are visibly better than the ones of 5dII, but I would not mind for one second, as I shoot RAW exclusively."If I shot raw, we'd be comparing the variations in how any given piece of software processed images from different cameras instead of getting a clear view of what each camera actually does when processing an entire image to completion as a JPG."Furthermore he does not understand that a RAW file just gives the fullest, most complete potential of the camera + lens. He does as if processing a RAW file in the best possible way (better than the image pipeline in-camera) is somehow unethical.He also fails to remind that JPEG is a heavily lossy format which compresses the real image data to 15% of the original content."If you want to twiddle with raw software, you'll undoubtedly get different results. The Nikon looks great shot as JPG. The Canon looks crappy and might benefit greatly if you have nothing else better to do than screw in front of a computer just to get the images you need. "He just likes simple truths and populist views. In the past it was normal to process a negative in the dark room, nowadays it can be done easily in front of a computer. This can be as simple as pushing the file through a RAW converter of choice with the camera preferences that work best, which will lead to better results than any JPEG engine.I even think that Rockwell knows this, but somehow he prefers to provoque with the easy solutions.And this has nothing to do with Nikon or Canon, it is about minimum knowledge in modern photography.bernie

Evan Effawrote:Why not test it against the 1DsIII; the real rival to the D3x?This would be valid for speed, AF etc. For pure image quality - in RAW of course - I am quite certain that the 5d II will be very close to the Nikon.The lenses (Nikon versus Canon) will probably make a bigger difference than the sensor and electronics.Even Canon have admitted that the newer cameras (5dII) will always tend to deliver better IQ than past models, even the top models.regards Bernie

oh, just stop this..anyone, who actually used 5dII or 1dsIII (me included) will tell you that 5dII sample is very bad and it's not presents true picture. i don't care what went wrong (nr, bad focus, bad lense sample etc), but it's clearly faulty. and if you want to test resolution of the sensor(s), do it with SAME lens on both cameras, in raw format and zero sharpening (again, on both cameras) using liveview!!we all know by now nikon is using more in-camera colour noise reduction and sharpening (in very efficient way, i must admit) than canon, but i will not go there.all i'm saying is: ken, be more careful next time, there's lot of other (done right) tests on the net, that prove you wrong.

Waldo Nellwrote:...as I own the 1Ds3 which has a very, very similar sensor I can comfortably say with a Canon 50mm F1.4 lens at optimum aperture I can get an image better than even his D3X. He clearly botched up the 5DII shot. -- Waldo NellI agree. The 5DII example looks really weird, but then again I shoot always RAW. I would easily get much better sharpness per pixel with my 50/1.4.What he obviously misses that even with RAW, image processing can be automatized so that there's no need to fiddle with every image.He also ignores the fact that it possible to open RAW files from new cameras with older Photoshop version. Just convert the files with the DNG Converter first.What is really amusing is that he prefers to use a 8000$ camera and then want to shoot JPG with it. I just won't get it.

it's very bad sample "even" for jpeg shot.