How much detail can you get from an EOS 5D Mark III?
A lot more than you may think, I would say. And not significantly less than what you would get with a Nikon D800. Photographer Ron Martinsen got a brand new D800 and did a quick and interesting comparison with the EOS 5D Mark III (while reviewing the former). Read more after the jump (crops provided).
There is no discussion about the resolution capabilities of the D800: 36MP are a lot, really a lot of pixels, and there are situations where you need this. Period. However, as has been discussed here and on other sites, the resolution galore comes with some drawbacks, the most notable being a not-so-good high ISO performance (which is way better in the 5D Mark III). Mr. Martinsen confirms this finding when he writes that «in good light this thing [the D800] is impressive» and that he does not «feel that the ISO performance is worth a hoot after ISO 3200». This confirms what we already know about Nikon’s D800: exceptionally good resolution, below the average high ISO performance. However, what really surprised me (again), is the amount of details that the 22MP sensor of the 5D Mark III is able to render. I provide two crops below (courtesy of Mr. Martinsen, JPGs out of the cam), the first one shot with the D800 the second one with the 5D Mark III. Judge by yourself and then tell me if you are still thinking you are missing something because you do not have a 36MP sensor. Regarding the in-camera-processing:
Canon’s in-camera JPEG is certainly more satisfying to my eyes, and in this case gives me an image I could technically call done and do nothing else. With that said, it does add more in-camera sharpening and seems to expose better so it gives this image an unrealistic advantage over what is possible with the D800 image with similar processing (I’ll do that later in another review).
Here are the crops. Not only does the JPG from the 5D Mark III look better (better colors, in my opinion), but I can hardly spot more details in the D800 crop. What I am questioning is the following: Is a 36MP sensor a substantial advantage in real world shootings? Ok, this are JPGs, and only RAW can tell the whole story. But then again: do you prefer to have 36MP or a camera that performs very good in high ISO and is a rugged all-rounder camera (the 5D3)?

100% crop from EOS 5D Mark III shot – Credits: R. Martinsen

100% crop from Nikon D800 shot – Credits: R. Martinsen
There is also an EOS 5D Mark III review by Ron Martinsen, with lots of pictures and samples. Don’t miss it. From the conclusion:
I’m very pleased with the high ISO performance and features of the camera that allow me to get the shots that in the past would have been lost while I went searching for my flash. I’m feeling very comfortable shooting in manual and letting the ISO dictate the exposure, but I do find it tends to shoot one stop darker than I like with no exposure compensation. As a result I find myself dialing in the necessary ISO as needed.
I feel totally comfortable with using shots up to ISO 25,600 with anything at or below 6400 being as solid as ISO 1600 and below on my 5D Mark II.
EOS 5D Mark III automatic price check: Amazon, B&H, Adorama, Canon USA, J&R, Digitalrev, Borrow Lenses, eBay.







ben
3 months ago |The problem with this comparison begans with jpegs only because Nikon jpegs are not nice and not sharp out of camera. The canon sample is always sharpened in camera. I do not care for jpeg tests, they are lame to began with, and they do not reflect professional use.
h.j.
2 years ago |Want to see good comparison? Check this one:
http://www.fredmiranda.com/5DIII-D800/index_controlled-tests.html
Frank
2 years ago |Yeah, thanks for remembering me my duties.
Mr Hipsta
2 years ago |”I can hardly spot more details in the D800 crop”
Actually, I can’t either. Probably because the images are “normalized”, the D800 downsized to the same size for easier comparison. Not the way I would do it if I was to compare resolution, he-he. And as h.j. says, RAW files should be used.
//5D2 user
Rushmore
2 years ago |” I can hardly spot more details in the D800 crop”
really?
May be you need glasses?
5D3 image looks totally artificial and blown-up.
5D3 is superior camera in other areas
but when it comes to detail/resolution D800 wins by big margin. Stop this non sense.
h.j.
2 years ago |Measuring the sensor performance by just checking the JPG out of the box is simply not wise approach. But if you intend to do so tell me why the 100% crops looks the same? Both of the sensors differ in resolutions thus 100% crops must show different magnification… Now the question is: did you down sampled D800 image to match 5Dmk3 one? If yes, I do understand that you do not see any difference in this example. Try upsizing the 5Dmk3 to match D800 resolution – then the comparison will be reliable.
Mandrake
2 years ago |An overly PP’ed eyeball does not make for a good resolution test.
reto
2 years ago |+20 sharpening in LR for the D800 file and you can actually see the resolution advantage.
Musouka
2 years ago |IMHO, Canon has always had better colors than Nikon out of the camera. However, I understand that’s a subjective thing.
When I decided to buy a DSLR back in late 2007, I sat down and reviewed my options. On the one hand, there was the Canon 40D. On the the other hand, the almighty Nikon D300. Body-wise, Nikon had the superior body and a better flash system. However, looking at the sample images on DPR, the Canon colors looked more natural to me. The lens selection of the Nikon didn’t suit my taste as much as Canon’s. In the end, I decided to go with Canon… I managed to secure 400D Kit with a low-shutter count for $450! Two years later, I got the 7D and I’m happy with it. Both systems are powerful… the photographer is what makes the photo great.
BTW, you mentioned doing an article comparing the 5DIII and 7D and see if it was worth the upgrade. Should we expect this soon?