Shared publicly  - 
Canon EOS 5D Mark III vs Nikon D800 noise!

This is the one you've been waiting for! The #Canon #5D3 vs the #Nikon #D800 noise levels compared across their entire sensitivity ranges! That's the page I've linked to below, but I have in fact completed and uploaded all of my Canon 5D Mark III main tests, so feel free to check out the other results pages too... Indeed I urge you to also look at the 5D3 vs D800 quality / resolution page as it tells the other half of the story.

So you'll see how the 5D3 compares to the D800, 5D2 and NEX-7 in terms of real-life resolution (including one against the D800 at its 'medium' setting) and against the D800 and 5D2 in terms of noise. I hope to add NEX-7 noise comparisons in the future.

As always, I shot everything in RAW+JPEG, but I'm holding back on my RAW comparisons until everything's settled down on the converter front. Adobe is still not final and Canon's own DPP is currently resolving issues. So I'm waiting for a short while on the RAW side of things.

But I think you'll find the JPEG comparisons I've made very revealing, and hopefully useful...

Let me know what you think!

PS - you might find this useful or interesting +Trey Ratcliff , +Thomas Hawk , +Jason Law , +Stefan Haworth , +Todd Sisson , +RC Concepcion , +Nicole S. Young , +sly vegas , +Frederick Van Johnson, +Robert Scoble, +Scott Kublin , +Scott Jarvie , +Karen Hutton
Gordon Laing's profile photoMike Spinak's profile photoPJ Ammidon's profile photoMichael B. Stuart's profile photo
Does Nikon has such a bad internal jpeg compression process? Well, luckily most use raw - the Nikon jpegs look already at ISO 400 horrible.
+Leif Sikorski, that looks like luminance noise rather than JPEG compression artifacts to me. My takeaway is that the 5d3 seems to be about 2 full stops better in the noise department.
Thanks for the link, +Gordon Laing. Looks like the 5D3 did a pretty good job! I'm wondering what the results would be if the Nikon D800 was down-sized in Photoshop to a 22mp file.
Wow, the 5d3 did really well in comparison to the others. (The Sony also looks a lot better than I expected.)

I think I'd be very comfortable with the 5D3 up to 3200 or 6400, and I wouldn't hesistate too much to use it at 12,800 when necessary.

I'd be interested to see the RAW results, when available.

Thank you, +Gordon Laing.
Thanks for the info and images +Gordon Laing! Really appreciate you taking the time and effort to post these and I'm also looking forward to the RAW comparisons too. Any chance you are going to do a 'night sky' comparison with these models?
any tests regarding noise patterns / hot pixels when doing long exposures say 1-3 minute exposures? : )
thanks for the reviews! love seeing how each camera out performs each other at different aspects :) but pretty much at this point, we are all stuck with the lenses we have and making a decision of which camera body to purchase next is not that difficult..
+Gordon Laing I got the Nikon D4 (I was the guy who bought it off of eBay while everyone on Nikon Rumors was bitching about the audacity of the camera store that was selling it on ebay.. hehe) ... haven't been able to get a D800 yet, but I want one. My plan was to have the 800 on the tripod and the D4 as my hip shooter. So far I've been a little disappointed in the speed of the D4 - it just takes a lot longer to take photos in low-light than the D3S did... not really sure about the quality difference though :)
+Tom Anderson .. it takes alot longer to take photos in low light?.. what do you mean exactly??.. you referring to the autofocus?.. or????.. :)
hi +Jason Law I mean when I'm doing bracketed shots for an HDR. So in other words, if I set up to do 5 shots in low light conditions the D3S will take the shots much quicker than the D4 can
+Gordon Laing : you rock. Thanks. I already bought my new baby Mark the Third - but it's still great to have all this information at my finger tips. I am kind of curious how the Canon at 50 ISO would compare to the 100, but I guess I can go do that test myself. Cheers!
so you are referring to the speed of the shots between frames when using the built in interval timer?.. it is slower on the D4 than the D3s when used to bracket? and it is faster when doing so in daylight???..

I assume you have your camera on manual focus?
+Tom Anderson are you now referring to the D3x? Well of course it will, continuous high on D3x is slower than the D3s at 5fps vs 9fps. Different purpose camera's. If your referring to the D4, I would make sure you have your settings right as the D4 can shoot 10-11 fps where the D3s was 9fps.
great review +Gordon Laing - the low-light capability of the Canon 5D3 is amazing. I wonder how long before Canon have a high-megapixel alternative to the D800?
hi +Jim Davis just edited my comment - I meant the D4, there. I don't know enough about cameras to know exactly how to answer your and +Jason Law's question. I think the time to capture enough light and get an exposure is slower on the D4 than it is on the D3S (also slower on the D3X). So if I'm doing 5 bracketed shots, on the D3S they might fire quickly 2, 4, 8, 16, 30 second exposures... Whereas on the D4 the each shot might be a longer exposure.. Does this make sense?
+Tom Anderson all things considered. under identical shooting situations ie. ISO, aperture, shutter speed, continuous high FPS, resolution etc the D4 should be slightly faster with bracketed exposures. It could likely be that your shooting under different conditions and it seems slower than the D3s. Check to see if you have noise reduction turned on, this will slow things down. These camera's have so many settings and one thing could be causing this.

My D4 has not come in yet as us Canadians have to wait our turn which is usually a month later than the US.
oh wow +Jim Davis that would be awesome... I will look into this. That's not true of the D3S and D3X is it? I've been with +Trey Ratcliff while he was using a D3X and I was on a D3S on the same settings... The D3S was definitely faster.
Yes +Tom Anderson The D3s will bracket faster than the D3x due to the higher fps on the D3s. There is a 4 fps difference on continuous high between them. How are you liking the D4? Have have it and the D800e on order, should be getting a call soon! My D3x is getting repair as we speak from being submerged under water!
I can't say I've really notice the difference, besides this speed issue, Jim. I really love the backlit buttons (because I shoot at night a lot). I got a D4 because I broke my D3S - but I haven't truly noticed a difference in quality between the two yet. I haven't used the video or any of the special features. I do appreciate the new format XQD card which is faster when transferring photos after a day of shooting.
+Gordon Laing - some great information, but the images as shown don't actually give us a good comparison between these two cameras since you used two different lenses, with the Nikon lens being nearly twice as expensive as the Canon. To do a proper comparison you'll need to use the same lens on both cameras, which is possible using a Nikon-Canon adapter, or some other modified lens. The crops from the Canon look like they were optically limited, not pixel limited. If you get a chance, that would be a truly valuable comparison to see! (make sure to use absolutely top notch glass, which the Canon 24-105 is not)
I've had the 5D3 for a few days now and I'm really impressed by it. It's a very different proposition to the D800 which is also very impressive with different advantages. Overall though I'm very pleased with the Canon which suits me perfectly. The AF is fantastic, responsive & accurate even in low light. It still hunts in very dark conditions but it's a huge improvement. It really isn't fair to call this an minor upgrade to the 5D2. I'd say it's significantly more usable all round. Very responsive and a pleasure to use
+Floris van Breugel I still think this is a fair comparison even thought the lenses are not identical. The 24-105 from Canon might be cheaper than the 24-70 from either Canon or Nikon but it's optically excellent - probably better than the 24-70s is many cases. But at F8 I doubt even the most critical pixel peepers will find a big difference between all these lenses. And after all this is a real world review for real world photographers who generally use Canon lenses on Canon bodies and Nikon lenses on Nikon bodies.
I've had the 5D3 for about a week now. I'm loving the image quality, reduced noise and tonal improvements. Here's a shot I got using the in camera HDR feature in almost darkness at 8PM at Sunset Cliffs in San Diego.
Used my 70-200mm 2.8 IS II for this shot. Didn't have a wide at the moment.
The camera captured details very nicely considering the in camera processing of HDR and the fact that it was almost completely dark out.

I've been very impressed by the 5D3 and I don't regret the investment in the least. Now I just need the new 24-70mm IS when it comes out and I'll be all set.
Interesting...... It's almost like they are two different cameras ;-)

The resolution comparisons on your site were fantastic. The D800 looked nicer shot at 20mp than 36mp somehow.

Looking forward to the raw comparisons so we can get past any in-camera jpeg processing shenanigans.

Good stuff. 
Can't say that I am surprised, in fact this is exactly what I was expecting. With the current technology, larger pixels typically means better noise control and less bleed over.
So the Canon should have better noise control.
I would suspect that the D800 has worse noise than a D700 for that matter.

Granted there is more to noise than just pixel size, but this has been a fairly accurate rule of thumb, and is based in some truths from the electronic and physical properties of today's sensors.

Thanks +Gordon Laing for the great information, I will be coming back to read the entire article!
+Brian Welch you will be very surprised. The D800 is much better at high ISO then the D700. Much better... You cannot look at the D800 images at 100% as those files are much larger than the D700 files. You need to downres the D800 files to 12 MP first. Only then the two will be comparable. You will see that the downsizing helps the D800 enormously. When downsized it is only slightly behind the D3s which is remarkable. And of course when downsized to 22 MP it blows the Mark III away. By quite a lot.
I disagree +Radek Vik. I did the same comparison a while back with some samples I downloaded from all 3 cameras and downsized to 12 MP. I found that the 5D3 was the best at high ISO - by a long way, followed by the D700 and last the D800. There wasn't much between the 2 Nikons but there was slightly more chroma noise visible in the downsized D800 images so I'd say that gives the D700 the edge. I did this for ISO 3200 - 12800. I'll admit my testing wasn't very scientific but that's what I found for what it's worth. The 'downsizing to average out noise' argument doesn't really hold water in my experience unless you're talking about downsizing to about 25%.

I think I need to get out and take more pictures. I obviously have too much time on my hands!
Well I don't know what you're seeing +Allen Corlett If done using the two cameras side by side and using the same lens, there is more or less a one stop advantage of the D800 over the D700 (from around ISO 6,400 or 12,800 up).
+Floris van Breugel Well maybe we're both goggle eyed after too much pixel peeping! It's time to get out there and take some photos and stop analysing :-)
+Allen Corlett - I'm not really into pixel peeping, but as a scientist, I do like to see experiments done with the best controls possible, otherwise, what's the point? When you're testing the resolution of a sensor you don't want to start with a system that is limited by its optics - then you'd be testing the lenses, not the sensors..
At the fork in the road: Nikon or Canon (we have a T3i currently and thinking about 5DM3 or D800)?
Please help... (I'm leaning Canon because we have the lenses)
Add a comment...