The High ISO Battle of the Full Frame Heavy Weights – Canon 6D vs Canon 5D Mark III

by planetMitch19 Comments

On the “Canon5DMarkSeries” yahoo group, a member (Don Marcotte) has been very interested in finding out about the high ISO performance of the Canon EOS 6D – so he managed to visit his local store and they let him do some “dark” tests – by placing the body cap or lens cap on the camera and taking a couple of stills to see how much noise displayed in the images. He compared the Canon EOS 6D against the Canon EOS 5D Mark III and the Canon EOS 5D Mark II. He's graciously let planet5D post his results.

Please be sure to read Don's process and realize that these results are exaggerated! Don's process exaggerated the noise to pull out the differences! Here's Don's disclaimer:

Please remember that this test was designed to determine the best camera for a very specific night photography project. I would expect that no one could detect a noise difference between the 6D and 5D3 for the vast majority of daylight photography. Only in low light conditions would a 6D deliver observably better results.


Also in the Canon EOS 6D news… the folks at DxO have rated the 6D sensor above the 5D3 and 5D2!

I hope to get my hands on a Canon EOS 6D again soon and will do some video tests – and Don tells me that he'll possibly be adding more images to his test results – so keep an eye on his gallery

Don Marcotte describes the tests

First of all, I chose 15 second, ISO 3200 exposures because that is a very important feature for my night photography with terrestrial objects in the foreground and the night sky in the background. It allows me to capture enough star light without the stars beginning to trail, with low noise. I am working towards a large print of the Milky Way sinking into an old graveyard at my cottage.

I had all images taken with either a lens cap or body cap on. That allowed me to isolate any noise generated by the camera. In the astro world, these are known as “darks”.

I tried as best I could to keep my comparison process standard for all cameras. I use only CR2 files as input. I convert them to 16 bit TIFFs using DPP. Then, using CS5, I take a center 400x400px crop, apply a gamma value of 3 and save it as a JPEG. Then I construct the matrix and label it.

All images were shot or converted to 5200°K, neutral Picture Style, noise reduction off, etc.

For the 5D Mk III and the 6D, the noise was so much less that it looked like a blank black wall. To make the noise more obvious, I applied a very large stretch to the each image using the gamma function in the CS5 Levels feature. I had to go to g=3 which is much, much higher than one would ever use in normal editing.

Comments, potential errors in my methodology or suggestions for process improvements are welcome.

As much as I would like to believe that my process was rigorous, I am having trouble believing that the 6D is significantly better than the Mk III. But I am very confident that it is a low noise camera. You can tell by the lack of noise when you zoom to 100% pixels. As someone who has worked to remove noise from astro images for the past 5-6 years, I am quite familiar with what noise looks like.

When I first saw the 6D my impression was that it had less features than my 60D! The dials and things seemed woefully lacking! However, the compactness and light weight are a real plus for my aging hands/arms. Because it lacks a swivel LCD, it likely not be as popular as a 60D or T4i for astro imaging. That swivel LCD is a God Send for helping focus a telescope that is pointed up high in the sky.

I have seen some astro images taken by a modified 6D though and they are impressive. Perhaps that will snare a few astrophotographers. However, the price will be a major factor too.

I hope you find this interesting and maybe even helpful.


Don's dark tests

Look at the 30 sec results and you might be moved to tears .

It's hard to tell whether most camera owners will need ISO 3200 and 6400. That being said, there is a significant cost benefit in low light – you don't need the fastest lenses. Another benefit is that you can use faster shutter speeds for fast moving objects.

But the 6D has less function. That will likely be important for many shooters.

The 15 second test (original is here):

15 second Canon EOS 6D and Canon EOS 5D Mark III High ISO test by Don Marcotte
Shot using camera store demo units at room temperature. RAW images converted to 16 bit TIFFs by Canon Digital Photo Professional. Created 400px X 400px crops and applied Gamma=3.0 in CS5.

The 30 second test (original is here):

30 second Canon EOS 6D and Canon EOS 5D Mark III High ISO test by Don Marcotte – images stretched to display noise more clearly.

What about the Canon EOS 5D Mark II?

I have recreated the 6D-5D3-5D2 15 second noise comparison but did it using a common process this time. This should be an apple-to-apples comparison.

(original is here):

The 6D is the clear winner at 15 and 30 seconds for ISO 3200 and 6400.

Please remember that this test was designed to determine the best camera for a very specific night photography project. I would expect that no one could detect a noise difference between the 6D and 5D3 for the vast majority of daylight photography. Only in low light conditions would a 6D deliver observably better results.

I would also like to remind you that I stretched those images aggressively and the actual results will be much less noisy.
For example, the original 6D 15 sec ISO 3200 RAW image looks like it has no noise whatsoever.

I have seen 5-10 minute astro images, taken by an astro modified 6D camera, with little or no noise and those were exceptional results.

Even though the 6D is apparently the best candidate, I may still get a 5D Mk III because of the extra features that it has. I have a decision to make.


(planetMitch again:) I'd like to thank Don for posting these images and sharing with the planet5D readers!

(cover photo credit: snap from the test by Don Marcotte)


  1. Pingback: Canon 6D and 5DMk3 Noise Comparison for High-ISO Long Exposures

  2. The 6D does not have an articulating screen, but it does have Wi-Fi and an app widely available for most cell phones that turns your phone into a live view screen. Even better then an articulating screen in my opinion. I own the 6D.

  3. Pingback: Canon EOS 6D Sensor Less Noisy Than EOS 5D Mark III | CanonWatch CanonWatch

  4. Pingback: Rapidinhas do mundo da fotografia – Parte V « Meio Bit

  5. Pingback: Cool Stuff – Week of December 16, 2012 | The Blog @

  6. Would really like to see this comparison with the Nikon D600 and D800e added in. DxO tests put the Nikons as quite a bit better in sensor performance. Would like to see if it holds true in this low light noise. Yes, my intended use is night and sky photography – so tests are worth looking at. Wish they started at 800 ISO.

  7. Pingback: OFFICIAL: CANON EOS 6D - User Thread - Part 2 - Page 39

  8. Is this test even relevant? The 5d3 seems better since in such low light environment it picks up something – unlike the 6d… 🙂

    Please test an actual situation of low light in a dark room etc… no way the 5d will produce worse results… would like to see a direct comparison…

    1. Understanding the necessity of real world testing in lowlight, why is this test irrelevant? Are we to conclude that the 6D sensor is technically better in lowlight (as a test like this certainly suggests) but that autofocus limitations make it inferior from a practical standpoint?

  9. honestly, I think even in practicalty in extreme low light , the 6D wins over the 5D3 because it has -3 EV AF albeit only the center point.

    I have the D800, the D600 and EOS6D and I currently shooting Nikon main but I’ve had too many quality control issues with Nikons last year.
    I tried the 6D and testing it through its paces and so far I love it.

    it is mall and light and fast in operation. unlike my D600 it has no dust on the sensor , it does not freeze or lock up like my D800 or D700 was.

    so, I may sell my Nikons and get the 5D3 for action , to back up my EOS6D.
    the 6D is an excellent in real life use with super high ISO capability and excellent lowlight AF.

    but if you are interested in video more than still like me , you might want to stick with the 5D3 , the 6D as with the D600 and the D800/e has severe moire issue.

  10. Pingback: NP101: Fundamentals of Nightscape Photography | Mike Salway

  11. This is exactly the type of comparison I was looking for. While a test with the lens cap on might seem irrelevant, I also do astrophotography, so was wondering how high-ISO performance compares between the 5D3 and 6D. Good to see it on such an apples-to-apples basis. But like you, I might go for the 5D3 despite the 6D’s better noise performance, simply because of the other features (the 11 cross-type AF points on the 5D3 is the biggest plus for me).

  12. @Jay The 5DIII is not better. The 5DIII is noisier, which is worse for astrophotography.

  13. @Ingemar But, autofocus is NOT used for astrophotography. Therefore, autofocus irrelevant issue.

  14. @John Thurston Low Light NOISE, or the lack thereof,  is extremely important for astrophotography.

  15. Articulated screen such as on (70d) or WiFi app is not necessary as USB cable and phone works just the same for 5d3 or any other camera with live view.I went for a good quality used 5d3 (20k actuations )for same money as 6d I’m happy with initial results using manual samyang lense. I use the camera for more than astro so required improved AF just waiting for motorsport season to start. I may miss my old 1d shutters speed thou.

    Would like to know which tracking mounts are good for astro.

  16. Pingback: Sensor Canon EOS 6D con menos ruido que el de la EOS 5D Mark III

Leave a Comment